nimbus-eth1/nimbus/p2p/clique/clique_snapshot.nim

423 lines
14 KiB
Nim
Raw Normal View History

Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# Nimbus
# Copyright (c) 2018 Status Research & Development GmbH
# Licensed under either of
# * Apache License, version 2.0, ([LICENSE-APACHE](LICENSE-APACHE) or
# http://www.apache.org/licenses/LICENSE-2.0)
# * MIT license ([LICENSE-MIT](LICENSE-MIT) or
# http://opensource.org/licenses/MIT)
# at your option. This file may not be copied, modified, or distributed except
# according to those terms.
##
## Snapshot for Clique PoA Consensus Protocol
## ==========================================
##
## For details see
## `EIP-225 <https://github.com/ethereum/EIPs/blob/master/EIPS/eip-225.md>`_
## and
## `go-ethereum <https://github.com/ethereum/EIPs/blob/master/EIPS/eip-225.md>`_
##
import
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
std/[sequtils, strformat, strutils],
../../constants,
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
../../db/db_chain,
./clique_cfg,
./clique_defs,
./clique_desc,
./snapshot/[lru_snaps, snapshot_apply, snapshot_desc],
chronicles,
eth/[common, keys],
nimcrypto,
stew/results,
stint
type
# Internal sub-descriptor for `LocalSnapsDesc`
LocalPivot = object
header: BlockHeader
hash: Hash256
# Internal sub-descriptor for `LocalSnapsDesc`
LocalPath = object
snaps: Snapshot ## snapshot for given hash
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
chain: seq[BlockHeader] ## header chain towards snapshot
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
error: CliqueError ## error message
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
# Internal sub-descriptor for `LocalSnapsDesc`
LocalSubChain = object
first: int ## first chain[] element to be used
top: int ## length of chain starting at position 0
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
LocalSnaps = object
c: Clique
start: LocalPivot ## start here searching for checkpoints
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
trail: LocalPath ## snapshot location
subChn: LocalSubChain ## chain[] sub-range
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
parents: seq[BlockHeader] ## explicit parents
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
{.push raises: [Defect].}
logScope:
topics = "clique PoA snapshot"
# ------------------------------------------------------------------------------
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
# Private debugging functions, pretty printing
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# ------------------------------------------------------------------------------
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
proc say(d: var LocalSnaps; v: varargs[string,`$`]) {.inline.} =
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
discard
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
# uncomment body to enable
#d.c.cfg.say v
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
proc pp(q: openArray[BlockHeader]; n: int): string {.inline.} =
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
result = "["
if 5 < n:
result &= toSeq(q[0 .. 2]).mapIt("#" & $it.blockNumber).join(", ")
result &= " .." & $n & ".. #" & $q[n-1].blockNumber
else:
result &= toSeq(q[0 ..< n]).mapIt("#" & $it.blockNumber).join(", ")
result &= "]"
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
proc pp(b: BlockNumber, q: openArray[BlockHeader]; n: int): string {.inline.} =
"#" & $b & " + " & q.pp(n)
proc pp(q: openArray[BlockHeader]): string {.inline.} =
q.pp(q.len)
proc pp(b: BlockNumber, q: openArray[BlockHeader]): string {.inline.} =
b.pp(q, q.len)
proc pp(h: BlockHeader, q: openArray[BlockHeader]; n: int): string {.inline.} =
"headers=(" & h.blockNumber.pp(q,n) & ")"
proc pp(h: BlockHeader, q: openArray[BlockHeader]): string {.inline.} =
h.pp(q,q.len)
proc pp(t: var LocalPath; w: var LocalSubChain): string {.inline.} =
var (a, b) = (w.first, w.top)
if a == 0 and b == 0: b = t.chain.len
"trail=(#" & $t.snaps.blockNumber & " + " & t.chain[a ..< b].pp & ")"
proc pp(t: var LocalPath): string {.inline.} =
var w = LocalSubChain()
t.pp(w)
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# ------------------------------------------------------------------------------
# Private helpers
# ------------------------------------------------------------------------------
proc maxCheckPointLe(d: var LocalSnaps;
number: BlockNumber): BlockNumber {.inline.} =
let epc = number mod d.c.cfg.ckpInterval
if epc < number:
number - epc
else:
# epc == number => number < ckpInterval
0.u256
proc isCheckPoint(d: var LocalSnaps;
number: BlockNumber): bool {.inline.} =
(number mod d.c.cfg.ckpInterval) == 0
proc isEpoch(d: var LocalSnaps;
number: BlockNumber): bool {.inline.} =
(number mod d.c.cfg.epoch) == 0
proc isSnapshotPosition(d: var LocalSnaps;
number: BlockNumber): bool {.inline.} =
# clique/clique.go(394): if number == 0 || (number%c.config.Epoch [..]
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
if d.isEpoch(number):
if number.isZero:
# At the genesis => snapshot the initial state.
return true
if not d.c.applySnapsMinBacklog:
return true
if d.c.cfg.roThreshold < d.trail.chain.len:
# We have piled up more headers than allowed to be re-orged (chain
# reinit from a freezer), regard checkpoint trusted and snapshot it.
return true
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# ------------------------------------------------------------------------------
# Private functions
# ------------------------------------------------------------------------------
proc findSnapshot(d: var LocalSnaps): bool
{.inline, gcsafe, raises: [Defect,CatchableError].} =
## Search for a snapshot starting at current header starting at the pivot
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
## value `d.start`. The snapshot returned in `trail` is a clone of the
## cached snapshot and can be modified later.
var
(header, hash) = (d.start.header, d.start.hash)
parentsLen = d.parents.len
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
# For convenience, ignore the current header as top parents list entry
if 0 < parentsLen and d.parents[^1] == header:
parentsLen.dec
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
while true:
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
#d.say "findSnapshot ", header.pp(d.parents, parentsLen),
# " trail=", d.trail.chain.pp
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
let number = header.blockNumber
# Check whether the snapshot was recently visited and cahed
if d.c.recents.hasLruSnaps(hash):
let rc = d.c.recents.getLruSnaps(hash)
if rc.isOK:
# we made sure that this is not a blind entry (currently no reason
# why there should be any, though)
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
d.trail.snaps = rc.value.cloneSnapshot
# d.say "findSnapshot cached ", d.trail.pp
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
debug "Found recently cached voting snapshot",
blockNumber = number,
blockHash = hash
return true
# If an on-disk checkpoint snapshot can be found, use that
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
if d.isCheckPoint(number):
let rc = d.c.cfg.loadSnapshot(hash)
if rc.isOk:
d.trail.snaps = rc.value.cloneSnapshot
d.say "findSnapshot disked ", d.trail.pp
trace "Loaded voting snapshot from disk",
blockNumber = number,
blockHash = hash
# clique/clique.go(386): snap = s
return true
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# Note that epoch is a restart and sync point. Eip-225 requires that the
# epoch header contains the full list of currently authorised signers.
if d.isSnapshotPosition(number):
# clique/clique.go(395): checkpoint := chain.GetHeaderByNumber [..]
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
d.trail.snaps = d.c.cfg.newSnapshot(header)
if d.trail.snaps.storeSnapshot.isOK:
d.say "findSnapshot <epoch> ", d.trail.pp
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
info "Stored voting snapshot to disk",
blockNumber = number,
blockHash = hash
return true
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
# No snapshot for this header, get the parent header and move backward
hash = header.parentHash
# Add to batch (reversed list order, biggest block number comes first)
d.trail.chain.add header
# Assign parent header
if 0 < parentslen:
# If we have explicit parents, pop it from the parents list
parentsLen.dec
header = d.parents[parentsLen]
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# clique/clique.go(416): if header.Hash() != hash [..]
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
if header.blockHash != hash:
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
d.trail.error = (errUnknownAncestor,"")
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
return false
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
# No explicit parents (or no more parents left), reach out to the database
elif not d.c.cfg.db.getBlockHeader(hash, header):
d.trail.error = (errUnknownAncestor,"")
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
return false
# => while loop
# notreached
raiseAssert "findSnapshot(): wrong exit from forever-loop"
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
proc applyTrail(d: var LocalSnaps): CliqueOkResult
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
{.inline, gcsafe, raises: [Defect,CatchableError].} =
## Apply any `trail` headers on top of the snapshot `snap`
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
if d.subChn.first < d.subChn.top:
block:
# clique/clique.go(434): snap, err := snap.apply(headers)
d.say "applyTrail ", d.trail.pp(d.subChn)
let rc = d.trail.snaps.snapshotApplySeq(
d.trail.chain, d.subChn.top-1, d.subChn.first)
if rc.isErr:
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
d.say "applyTrail snaps=#",d.trail.snaps.blockNumber, " err=",$rc.error
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
return err(rc.error)
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
d.say "applyTrail snaps=#", d.trail.snaps.blockNumber
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
# If we've generated a new checkpoint snapshot, save to disk
if d.isCheckPoint(d.trail.snaps.blockNumber):
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
var rc = d.trail.snaps.storeSnapshot
if rc.isErr:
return err(rc.error)
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
d.say "updateSnapshot <disk> chechkpoint #", d.trail.snaps.blockNumber
trace "Stored voting snapshot to disk",
blockNumber = d.trail.snaps.blockNumber,
blockHash = d.trail.snaps.blockHash
ok()
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
proc updateSnapshot(d: var LocalSnaps): SnapshotResult
{.gcsafe, raises: [Defect,CatchableError].} =
## Find snapshot for header `d.start.header` and assign it to the LRU cache.
## This function was expects thet the LRU cache already has a slot allocated
## for the snapshot having run `getLruSnaps()`.
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
d.say "updateSnapshot begin ", d.start.header.blockNumber.pp(d.parents)
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# Search for previous snapshots
if not d.findSnapshot:
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
return err(d.trail.error)
# Initialise range for header chain[] to be applied to `d.trail.snaps`
d.subChn.top = d.trail.chain.len
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# Previous snapshot found, apply any pending trail headers on top of it
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
if 0 < d.subChn.top:
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
let
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
first = d.trail.chain[^1].blockNumber
last = d.trail.chain[0].blockNumber
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
ckpt = d.maxCheckPointLe(last)
# If there is at least one checkpoint part of the trail sequence, make sure
# that we can store the latest one. This will be done by the `applyTrail()`
# handler for the largest block number in the sequence (note that the trail
# block numbers are in reverse order.)
if first <= ckpt and ckpt < last:
# Split the trail sequence so that the first one has the checkpoint
# entry with largest block number.
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
let inx = (last - ckpt).truncate(int)
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
# First part (note reverse block numbers.)
d.subChn.first = inx
let rc = d.applyTrail
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
if rc.isErr:
return err(rc.error)
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
# Second part (note reverse block numbers.)
d.subChn.first = 0
d.subChn.top = inx
var rc = d.applyTrail
if rc.isErr:
return err(rc.error)
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# clique/clique.go(438): c.recents.Add(snap.Hash, snap)
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
if not d.c.recents.setLruSnaps(d.trail.snaps):
# Someting went seriously wrong, most probably this function was called
# before checking the LRU cache first -- lol
return err((errSetLruSnaps, &"block #{d.trail.snaps.blockNumber}"))
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
if 1 < d.trail.chain.len:
d.say "updateSnapshot ok #", d.trail.snaps.blockNumber,
" trail.len=", d.trail.chain.len
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
ok(d.trail.snaps)
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# ------------------------------------------------------------------------------
# Public functions
# ------------------------------------------------------------------------------
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
proc cliqueSnapshotSeq*(c: Clique; header: Blockheader;
parents: var seq[Blockheader]): SnapshotResult
{.gcsafe, raises: [Defect,CatchableError].} =
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
## Create authorisation state snapshot of a given point in the block chain
## and store it in the `Clique` descriptor to be retrievable as `c.snapshot`
## if successful.
##
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
## If the `parents[]` argument list top element (if any) is the same as the
## `header` argument, this top element is silently ignored.
##
## If this function is successful, the compiled `Snapshot` will also be
## stored in the `Clique` descriptor which can be retrieved later
## via `c.snapshot`.
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
let rc1 = c.recents.getLruSnaps(header.blockHash)
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
if rc1.isOk:
c.snapshot = rc1.value
return ok(rc1.value)
# Avoid deep copy, sequence will not be changed by `updateSnapshot()`
parents.shallow
var snaps = LocalSnaps(
c: c,
parents: parents,
start: LocalPivot(
header: header,
Fearture/poa clique tuning (#765) * Provide API details: API is bundled via clique.nim. * Set extraValidation as default for PoA chains why: This triggers consensus verification and an update of the list of authorised signers. These signers are integral part of the PoA block chain. todo: Option argument to control validation for the nimbus binary. * Fix snapshot state block number why: Using sub-sequence here, so the len() function was wrong. * Optional start where block verification begins why: Can speed up time building loading initial parts of block chain. For PoA, this allows to prove & test that authorised signers can be (correctly) calculated starting at any point on the block chain. todo: On Goerli around blocks #193537..#197568, processing time increases disproportionally -- needs to be understand * For Clique test, get old grouping back (7 transactions per log entry) why: Forgot to change back after troubleshooting * Fix field/function/module-name misunderstanding why: Make compilation work * Use eth_types.blockHash() rather than utils.hash() in Clique modules why: Prefer lib module * Dissolve snapshot_misc.nim details: .. into clique_verify.nim (the other source file clique_unused.nim is inactive) * Hide unused AsyncLock in Clique descriptor details: Unused here but was part of the Go reference implementation * Remove fakeDiff flag from Clique descriptor details: This flag was a kludge in the Go reference implementation used for the canonical tests. The tests have been adapted so there is no need for the fakeDiff flag and its implementation. * Not observing minimum distance from epoch sync point why: For compiling PoA state, the go implementation will walk back to the epoch header with at least 90000 blocks apart from the current header in the absence of other synchronisation points. Here just the nearest epoch header is used. The assumption is that all the checkpoints before have been vetted already regardless of the current branch. details: The behaviour of using the nearest vs the minimum distance epoch is controlled by a flag and can be changed at run time. * Analysing processing time (patch adds some debugging/visualisation support) why: At the first half million blocks of the Goerli replay, blocks on the interval #194854..#196224 take exceptionally long to process, but not due to PoA processing. details: It turns out that much time is spent in p2p/excecutor.processBlock() where the elapsed transaction execution time is significantly greater for many of these blocks. Between the 1371 blocks #194854..#196224 there are 223 blocks with more than 1/2 seconds execution time whereas there are only 4 such blocks before and 13 such after this range up to #504192. * fix debugging symbol in clique_desc (causes CI failing) * Fixing canonical reference tests why: Two errors were introduced earlier but ovelooked: 1. "Remove fakeDiff flag .." patch was incomplete 2. "Not observing minimum distance .." introduced problem w/tests 23/24 details: Fixing 2. needed to revert the behaviour by setting the applySnapsMinBacklog flag for the Clique descriptor. Also a new test was added to lock the new behaviour. * Remove cruft why: Clique/PoA processing was intended to take place somewhere in executor/process_block.processBlock() but was decided later to run from chain/persist_block.persistBlock() instead. * Update API comment * ditto
2021-07-30 14:06:51 +00:00
hash: header.blockHash))
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
let rc2 = snaps.updateSnapshot
if rc2.isOk:
c.snapshot = rc2.value
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
rc2
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
proc cliqueSnapshotSeq*(c: Clique; hash: Hash256;
parents: var seq[Blockheader]): SnapshotResult
{.gcsafe,raises: [Defect,CatchableError].} =
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
## Create authorisation state snapshot of a given point in the block chain
## and store it in the `Clique` descriptor to be retrievable as `c.snapshot`
## if successful.
##
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
## If the `parents[]` argument list top element (if any) is the same as the
## `header` argument, this top element is silently ignored.
##
## If this function is successful, the compiled `Snapshot` will also be
## stored in the `Clique` descriptor which can be retrieved later
## via `c.snapshot`.
let rc1 = c.recents.getLruSnaps(hash)
if rc1.isOk:
c.snapshot = rc1.value
return ok(rc1.value)
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
var header: BlockHeader
if not c.cfg.db.getBlockHeader(hash, header):
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
return err((errUnknownHash,""))
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
# Avoid deep copy, sequence will not be changed by `updateSnapshot()`
parents.shallow
var snaps = LocalSnaps(
c: c,
parents: parents,
start: LocalPivot(
header: header,
hash: hash))
let rc2 = snaps.updateSnapshot
if rc2.isOk:
c.snapshot = rc2.value
rc2
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
# clique/clique.go(369): func (c *Clique) snapshot(chain [..]
proc cliqueSnapshot*(c: Clique; header: Blockheader;
parents: var seq[Blockheader]): SnapshotResult
{.gcsafe, raises: [Defect,CatchableError].} =
var list = toSeq(parents)
c.cliqueSnapshotSeq(header,list)
proc cliqueSnapshot*(c: Clique;hash: Hash256;
parents: openArray[Blockheader]): SnapshotResult
{.gcsafe, raises: [Defect,CatchableError].} =
var list = toSeq(parents)
c.cliqueSnapshotSeq(hash,list)
proc cliqueSnapshot*(c: Clique; header: Blockheader): SnapshotResult
{.inline,gcsafe,raises: [Defect,CatchableError].} =
## Short for `cliqueSnapshot(c,header,@[])`
var blind: seq[Blockheader]
c.cliqueSnapshotSeq(header, blind)
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
proc cliqueSnapshot*(c: Clique; hash: Hash256): SnapshotResult
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
{.gcsafe,raises: [Defect,CatchableError].} =
## Short for `cliqueSnapshot(c,hash,@[])`
Renamed source file clique_utils => clique_helpers (#762) * Renamed source file clique_utils => clique_helpers why: New name is more in line with other modules where local libraries are named similarly. * re-implemented PoA verification module as clique_verify.nim details: The verification code was ported from the go sources and provisionally stored in the clique_misc.nim source file. todo: Bring it to life. * re-design Snapshot descriptor as: ref object why: Avoids some copying descriptor objects details: The snapshot management in clique_snapshot.nim has been cleaned up. todo: There is a lot of unnecessary copying & sub-list manipulation of seq[BlockHeader] lists which needs to be simplified by managing index intervals. * optimised sequence handling for Clique/PoA why: To much ado about nothing details: * Working with shallow sequences inside PoA processing avoids unnecessary copying. * Using degenerate lists in the cliqueVerify() batch where only the parent (and no other ancestor) is needed. todo: Expose only functions that are needed, shallow sequences should be handles with care. * fix var-parameter function argument * Activate PoA engine -- currently proof of concept details: PoA engine is activated with newChain(extraValidation = true) applied to a PoA network. status and todo: The extraValidation flag on the Chain object can be set at a later state which allows to pre-load parts of the block chain without verification. Setting it later will only go back the block chain to the latest epoch checkpoint. This is inherent to the Clique protocol, needs testing though. PoA engine works in fine weather mode on Goerli replay. With the canonical eip-225 tests, there are quite a few fringe conditions that fail. These can easily fudged over to make things work but need some more work to understand and correct properly. * Make the last offending verification header available why: Makes some fringe case tests work. details: Within a failed transaction comprising several blocks, this feature help to identify the offending block if there was a PoA verification error. * Make PoA header verifier store the final snapshot why: The last snapshot needed by the verifier is the one of the parent but the list of authorised signer is derived from the current snapshot. So updating to the latest snapshot provides the latest signers list. details: Also, PoA processing has been implemented as transaction in persistBlocks() with Clique state rollback. Clique tests succeed now. * Avoiding double yields in iterator => replaced by template why: Tanks to Andri who observed it (see #762) * Calibrate logging interval and fix logging event detection why: Logging interval as copied from Go implementation was too large and needed re-calibration. Elapsed time calculation was bonkers, negative the wrong way round.
2021-07-21 13:31:52 +00:00
var blind: seq[Blockheader]
c.cliqueSnapshot(hash, blind)
Feature/implement poa processing (#748) * re-shuffled Clique functions why: Due to the port from the go-sources, the interface logic is not optimal for nimbus. The main visible function is currently snapshot() and most of the _procurement_ of this function result has been moved to a sub-directory. * run eip-225 Clique test against p2p/chain.persistBlocks() why: Previously, loading the test block chains was fugdged with the purpose only to fill the database. As it is now clear how nimbus works on Goerli, the same can be achieved with a more realistic scenario. details: Eventually these tests will be pre-cursor to the reply tests for the Goerli chain supporting TDD approach with more simple cases. * fix exception annotations for executor module why: needed for exception tracking details: main annoyance are vmState methods (in state.nim) which potentially throw a base level Exception (a proc would only throws CatchableError) * split p2p/chain into sub-modules and fix exception annotations why: make space for implementing PoA stuff * provide over-loadable Clique PRNG why: There is a PRNG provided for generating reproducible number sequences. The functions which employ the PRNG to generate time slots were ported ported from the go-implementation. They are currently unused. * implement trusted signer assembly in p2p/chain.persistBlocks() details: * PoA processing moved there at the end of a transaction. Currently, there is no action (eg. transaction rollback) if this fails. * The unit tests with staged blocks work ok. In particular, there should be tests with to-be-rejected blocks. * TODO: 1.Optimise throughput/cache handling; 2.Verify headers * fix statement cast in pool.nim * added table features to LRU cache why: Clique uses the LRU cache using a mixture of volatile online items from the LRU cache and database checkpoints for hard synchronisation. For performance, Clique needs more table like features. details: First, last, and query key added, as well as efficient random delete added. Also key-item pair iterator added for debugging. * re-factored LRU snapshot caching why: Caching was sub-optimal (aka. bonkers) in that it skipped over memory caches in many cases and so mostly rebuild the snapshot from the last on-disk checkpoint. details; The LRU snapshot toValue() handler has been moved into the module clique_snapshot. This is for the fact that toValue() is not supposed to see the whole LRU cache database. So there must be a higher layer working with the the whole LRU cache and the on-disk checkpoint database. also: some clean up todo: The code still assumes that the block headers are valid in itself. This is particular important when an epoch header (aka re-sync header) is processed as it must contain the PoA result of all previous headers. So blocks need to be verified when they come in before used for PoA processing. * fix some snapshot cache fringe cases why: Must not index empty sequences in clique_snapshot module
2021-07-14 15:13:27 +00:00
# ------------------------------------------------------------------------------
# End
# ------------------------------------------------------------------------------