nimbus-eth1/tests/test_clique/pool.nim

450 lines
16 KiB
Nim
Raw Normal View History

# Nimbus
# Copyright (c) 2018-2019 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.
import
std/[random, sequtils, strformat, strutils, tables, times],
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
../../nimbus/[config, chain_config, constants, genesis],
../../nimbus/db/db_chain,
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/p2p/[chain,
clique,
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
clique/clique_desc,
clique/clique_genvote,
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_helpers,
clique/clique_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
clique/snapshot/snapshot_desc],
./voter_samples as vs,
Feature/goerli replay clique poa (#743) * extract unused clique/mining support into separate file why: mining is currently unsupported by nimbus * Replay first 51840 transactions from Goerli block chain why: Currently Goerli is loaded but the block headers are not verified. Replaying allows real data PoA development. details: Simple stupid gzipped dump/undump layer for debugging based on the zlib module (no nim-faststream support.) This is a replay running against p2p/chain.persistBlocks() where the data were captured from. * prepare stubs for PoA engine * split executor source into sup-modules why: make room for updates, clique integration should go into executor/update_poastate.nim * Simplify p2p/executor.processBlock() function prototype why: vmState argument always wraps basicChainDB * split processBlock() into sub-functions why: isolate the part where it will support clique/poa * provided additional processTransaction() function prototype without _fork_ argument why: with the exception of some tests, the _fork_ argument is always derived from the other prototype argument _vmState_ details: similar situation with makeReceipt() * provide new processBlock() version explicitly supporting PoA details: The new processBlock() version supporting PoA is the general one also supporting non-PoA networks, it needs an additional _Clique_ descriptor function argument for PoA state (if any.) The old processBlock() function without the _Clique_ descriptor argument retorns an error on PoA networgs (e.g. Goerli.) * re-implemented Clique descriptor as _ref object_ why: gives more flexibility when moving around the descriptor object details: also cleaned up a bit the clique sources * comments for clarifying handling of Clique/PoA state descriptor
2021-07-06 13:14:45 +00:00
eth/[common, keys, p2p, rlp, trie/db],
ethash,
secp256k1_abi,
stew/objects
export
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
vs, snapshot_desc
const
prngSeed = 42
type
XSealKey = array[EXTRA_SEAL,byte]
XSealValue = object
blockNumber: uint64
account: string
TesterPool* = ref object ## Pool to maintain currently active tester accounts,
## mapped from textual names used in the tests below
## to actual Ethereum private keys capable of signing
## transactions.
prng: Rand
accounts: Table[string,PrivateKey] ## accounts table
boot: CustomNetwork ## imported Genesis configuration
batch: seq[seq[BlockHeader]] ## collect header chains
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
chain: Chain
names: Table[EthAddress,string] ## reverse lookup for debugging
xSeals: Table[XSealKey,XSealValue] ## collect signatures for debugging
# ------------------------------------------------------------------------------
# Private Helpers
# ------------------------------------------------------------------------------
proc getBlockHeader(ap: TesterPool; number: BlockNumber): BlockHeader =
## Shortcut => db/db_chain.getBlockHeader()
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
doAssert ap.chain.clique.db.getBlockHeader(number, result)
proc getBlockHeader(ap: TesterPool; hash: Hash256): BlockHeader =
## Shortcut => db/db_chain.getBlockHeader()
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
doAssert ap.chain.clique.db.getBlockHeader(hash, result)
proc isZero(a: openArray[byte]): bool =
result = true
for w in a:
if w != 0:
return false
proc rand(ap: TesterPool): byte =
ap.prng.rand(255).byte
proc newPrivateKey(ap: TesterPool): PrivateKey =
## Roughly modelled after `random(PrivateKey,getRng()[])` with
## non-secure but reproducible PRNG
var data{.noinit.}: array[SkRawSecretKeySize,byte]
for n in 0 ..< data.len:
data[n] = ap.rand
# verify generated key, see keys.random(PrivateKey) from eth/keys.nim
var dataPtr0 = cast[ptr cuchar](unsafeAddr data[0])
doAssert secp256k1_ec_seckey_verify(
secp256k1_context_no_precomp, dataPtr0) == 1
# Convert to PrivateKey
PrivateKey.fromRaw(data).value
proc privateKey(ap: TesterPool; account: string): PrivateKey =
## Return private key for given tester `account`
if account != "":
if account in ap.accounts:
result = ap.accounts[account]
else:
result = ap.newPrivateKey
ap.accounts[account] = result
let address = result.toPublicKey.toCanonicalAddress
ap.names[address] = account
# ------------------------------------------------------------------------------
# Private pretty printer call backs
# ------------------------------------------------------------------------------
proc findName(ap: TesterPool; address: EthAddress): string =
## Find name for a particular address
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 address notin ap.names:
ap.names[address] = &"X{ap.names.len+1}"
ap.names[address]
proc findSignature(ap: TesterPool; sig: openArray[byte]): XSealValue =
## Find a previusly registered signature
if sig.len == XSealKey.len:
let key = toArray(XSealKey.len,sig)
if key in ap.xSeals:
result = ap.xSeals[key]
proc ppNonce(ap: TesterPool; v: BlockNonce): string =
## Pretty print nonce
if v == NONCE_AUTH:
"AUTH"
elif v == NONCE_DROP:
"DROP"
else:
&"0x{v.toHex}"
proc ppAddress(ap: TesterPool; v: EthAddress): string =
## Pretty print address
if v.isZero:
result = "@0"
else:
let a = ap.findName(v)
if a == "":
result = &"@{v}"
else:
result = &"@{a}"
proc ppExtraData(ap: TesterPool; v: Blob): string =
## Visualise `extraData` field
if v.len < EXTRA_VANITY + EXTRA_SEAL or
((v.len - (EXTRA_VANITY + EXTRA_SEAL)) mod EthAddress.len) != 0:
result = &"0x{v.toHex}[{v.len}]"
else:
var data = v
#
# extra vanity prefix
let vanity = data[0 ..< EXTRA_VANITY]
data = data[EXTRA_VANITY ..< data.len]
result = if vanity.isZero: "0u256+" else: &"{vanity.toHex}+"
#
# list of addresses
if EthAddress.len + EXTRA_SEAL <= data.len:
var glue = "["
while EthAddress.len + EXTRA_SEAL <= data.len:
let address = toArray(EthAddress.len,data[0 ..< EthAddress.len])
data = data[EthAddress.len ..< data.len]
result &= &"{glue}{ap.ppAddress(address)}"
glue = ","
result &= "]+"
#
# signature
let val = ap.findSignature(data)
if val.account != "":
result &= &"<#{val.blockNumber},{val.account}>"
elif data.isZero:
result &= &"<0>"
else:
let sig = SkSignature.fromRaw(data)
if sig.isOk:
result &= &"<{sig.value.toHex}>"
else:
result &= &"0x{data.toHex}[{data.len}]"
proc ppBlockHeader(ap: TesterPool; v: BlockHeader; delim: string): string =
## Pretty print block header
let sep = if 0 < delim.len: delim else: ";"
&"(blockNumber=#{v.blockNumber.truncate(uint64)}" &
Feature/goerli replay clique poa (#743) * extract unused clique/mining support into separate file why: mining is currently unsupported by nimbus * Replay first 51840 transactions from Goerli block chain why: Currently Goerli is loaded but the block headers are not verified. Replaying allows real data PoA development. details: Simple stupid gzipped dump/undump layer for debugging based on the zlib module (no nim-faststream support.) This is a replay running against p2p/chain.persistBlocks() where the data were captured from. * prepare stubs for PoA engine * split executor source into sup-modules why: make room for updates, clique integration should go into executor/update_poastate.nim * Simplify p2p/executor.processBlock() function prototype why: vmState argument always wraps basicChainDB * split processBlock() into sub-functions why: isolate the part where it will support clique/poa * provided additional processTransaction() function prototype without _fork_ argument why: with the exception of some tests, the _fork_ argument is always derived from the other prototype argument _vmState_ details: similar situation with makeReceipt() * provide new processBlock() version explicitly supporting PoA details: The new processBlock() version supporting PoA is the general one also supporting non-PoA networks, it needs an additional _Clique_ descriptor function argument for PoA state (if any.) The old processBlock() function without the _Clique_ descriptor argument retorns an error on PoA networgs (e.g. Goerli.) * re-implemented Clique descriptor as _ref object_ why: gives more flexibility when moving around the descriptor object details: also cleaned up a bit the clique sources * comments for clarifying handling of Clique/PoA state descriptor
2021-07-06 13:14:45 +00:00
&"{sep}parentHash={v.parentHash}" &
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
&"{sep}selfHash={v.blockHash}" &
Feature/goerli replay clique poa (#743) * extract unused clique/mining support into separate file why: mining is currently unsupported by nimbus * Replay first 51840 transactions from Goerli block chain why: Currently Goerli is loaded but the block headers are not verified. Replaying allows real data PoA development. details: Simple stupid gzipped dump/undump layer for debugging based on the zlib module (no nim-faststream support.) This is a replay running against p2p/chain.persistBlocks() where the data were captured from. * prepare stubs for PoA engine * split executor source into sup-modules why: make room for updates, clique integration should go into executor/update_poastate.nim * Simplify p2p/executor.processBlock() function prototype why: vmState argument always wraps basicChainDB * split processBlock() into sub-functions why: isolate the part where it will support clique/poa * provided additional processTransaction() function prototype without _fork_ argument why: with the exception of some tests, the _fork_ argument is always derived from the other prototype argument _vmState_ details: similar situation with makeReceipt() * provide new processBlock() version explicitly supporting PoA details: The new processBlock() version supporting PoA is the general one also supporting non-PoA networks, it needs an additional _Clique_ descriptor function argument for PoA state (if any.) The old processBlock() function without the _Clique_ descriptor argument retorns an error on PoA networgs (e.g. Goerli.) * re-implemented Clique descriptor as _ref object_ why: gives more flexibility when moving around the descriptor object details: also cleaned up a bit the clique sources * comments for clarifying handling of Clique/PoA state descriptor
2021-07-06 13:14:45 +00:00
&"{sep}stateRoot={v.stateRoot}" &
&"{sep}coinbase={ap.ppAddress(v.coinbase)}" &
&"{sep}nonce={ap.ppNonce(v.nonce)}" &
&"{sep}extraData={ap.ppExtraData(v.extraData)})"
# ------------------------------------------------------------------------------
# Private: Constructor helpers
# ------------------------------------------------------------------------------
proc initPrettyPrinters(pp: var PrettyPrinters; ap: TesterPool) =
pp.nonce = proc(v:BlockNonce): string = ap.ppNonce(v)
pp.address = proc(v:EthAddress): string = ap.ppAddress(v)
pp.extraData = proc(v:Blob): string = ap.ppExtraData(v)
pp.blockHeader = proc(v:BlockHeader; d:string): string = ap.ppBlockHeader(v,d)
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 resetChainDb(ap: TesterPool; extraData: Blob; debug = false) =
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
## Setup new block chain with bespoke genesis
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
ap.chain = BaseChainDB(
db: newMemoryDb(),
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
config: ap.boot.config).newChain
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
ap.chain.clique.db.populateProgress
# new genesis block
var g = ap.boot.genesis
if 0 < extraData.len:
g.extraData = extraData
g.commit(ap.chain.clique.db)
# fine tune Clique descriptor
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
ap.chain.clique.cfg.debug = debug
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
ap.chain.clique.cfg.prettyPrint.initPrettyPrinters(ap)
proc initTesterPool(ap: TesterPool): TesterPool {.discardable.} =
result = ap
result.prng = initRand(prngSeed)
result.batch = @[newSeq[BlockHeader]()]
result.accounts = initTable[string,PrivateKey]()
result.xSeals = initTable[XSealKey,XSealValue]()
result.names = initTable[EthAddress,string]()
result.resetChainDb(@[])
# ------------------------------------------------------------------------------
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: pretty printer support
# ------------------------------------------------------------------------------
proc getPrettyPrinters*(t: TesterPool): var PrettyPrinters =
## Mixin for pretty printers, see `clique/clique_cfg.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
t.chain.clique.cfg.prettyPrint
proc say*(t: TesterPool; v: varargs[string,`$`]) =
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 t.chain.clique.cfg.debug:
stderr.write v.join & "\n"
proc sayHeaderChain*(ap: TesterPool; indent = 0): TesterPool {.discardable.} =
result = ap
let pfx = ' '.repeat(indent)
var top = if 0 < ap.batch[^1].len: ap.batch[^1][^1]
else: ap.getBlockHeader(0.u256)
ap.say pfx, " top header: " & ap.pp(top, 16+indent)
while not top.blockNumber.isZero:
top = ap.getBlockHeader(top.parentHash)
ap.say pfx, "parent header: " & ap.pp(top, 16+indent)
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: Constructor
# ------------------------------------------------------------------------------
proc newVoterPool*(networkId = GoerliNet): TesterPool =
TesterPool(
boot: CustomNetwork(
genesis: genesisBlockForNetwork(networkId, CustomNetwork()),
config: chainConfig(networkId, CustomNetwork()))).initTesterPool
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: getter
# ------------------------------------------------------------------------------
proc chain*(ap: TesterPool): auto {.inline.} =
## Getter
ap.chain
proc clique*(ap: TesterPool): auto {.inline.} =
## Getter
ap.chain.clique
proc db*(ap: TesterPool): auto {.inline.} =
## Getter
ap.clique.db
proc debug*(ap: TesterPool): auto {.inline.} =
## Getter
ap.clique.cfg.debug
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 cliqueSigners*(ap: TesterPool): auto {.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
## Getter
ap.clique.cliqueSigners
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 cliqueSignersLen*(ap: TesterPool): auto {.inline.} =
## Getter
ap.clique.cliqueSignersLen
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 snapshot*(ap: TesterPool): auto {.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
## Getter
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
ap.clique.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
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 failed*(ap: TesterPool): CliqueFailed {.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
## Getter
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
ap.clique.failed
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: setter
# ------------------------------------------------------------------------------
proc `debug=`*(ap: TesterPool; debug: bool) {.inline,} =
## Set debugging mode on/off
ap.clique.cfg.debug = debug
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 `verifyFrom=`*(ap: TesterPool; verifyFrom: uint64) {.inline.} =
## Setter, block number where `Clique` should start
ap.chain.verifyFrom = verifyFrom
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
# ------------------------------------------------------------------------------
# clique/snapshot_test.go(62): func (ap *testerAccountPool) address(account [..]
proc address*(ap: TesterPool; account: string): EthAddress =
## retrieves the Ethereum address of a tester account by label, creating
## a new account if no previous one exists yet.
if account != "":
result = ap.privateKey(account).toPublicKey.toCanonicalAddress
# ------------------------------------------------------------------------------
# Public: set up & manage voter database
# ------------------------------------------------------------------------------
proc resetVoterChain*(ap: TesterPool; signers: openArray[string];
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
epoch = 0; runBack = true): TesterPool {.discardable.} =
## Reset the batch list for voter headers and update genesis block
result = ap
ap.batch = @[newSeq[BlockHeader]()]
# clique/snapshot_test.go(384): signers := make([]common.Address, [..]
let signers = signers.mapIt(ap.address(it)).sorted(EthAscending)
var extraData = 0.byte.repeat(EXTRA_VANITY)
# clique/snapshot_test.go(399): for j, signer := range signers {
for signer in signers:
extraData.add signer.toSeq
# clique/snapshot_test.go(397):
extraData.add 0.byte.repeat(EXTRA_SEAL)
# store modified genesis block and epoch
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
ap.resetChainDb(extraData, ap.debug )
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
ap.clique.cfg.epoch = 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
ap.clique.applySnapsMinBacklog = runBack
# clique/snapshot_test.go(415): blocks, _ := core.GenerateChain(&config, [..]
proc appendVoter*(ap: TesterPool;
voter: TesterVote): TesterPool {.discardable.} =
## Append a voter header to the block chain batch list
result = ap
doAssert 0 < ap.batch.len # see initTesterPool() and resetVoterChain()
let parent = if ap.batch[^1].len == 0:
ap.getBlockHeader(0.u256)
else:
ap.batch[^1][^1]
let header = ap.chain.clique.cliqueGenvote(
voter = ap.address(voter.voted),
seal = ap.privateKey(voter.signer),
parent = parent,
elapsed = initDuration(seconds = 100),
voteInOk = voter.auth,
outOfTurn = voter.noTurn,
checkPoint = voter.checkpoint.mapIt(ap.address(it)).sorted(EthAscending))
if 0 < voter.checkpoint.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
doAssert (header.blockNumber mod ap.clique.cfg.epoch) == 0
# Register for debugging
let
extraLen = header.extraData.len
extraSeal = header.extraData[extraLen - EXTRA_SEAL ..< extraLen]
ap.xSeals[toArray(XSealKey.len,extraSeal)] = XSealValue(
blockNumber: header.blockNumber.truncate(uint64),
account: voter.signer)
if voter.newbatch:
ap.batch.add @[]
ap.batch[^1].add header
proc appendVoter*(ap: TesterPool;
voters: openArray[TesterVote]): TesterPool {.discardable.} =
## Append a list of voter headers to the block chain batch list
result = ap
for voter in voters:
ap.appendVoter(voter)
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 commitVoterChain*(ap: TesterPool; postProcessOk = false;
stopFaultyHeader = false): TesterPool {.discardable.} =
## Write the headers from the voter header batch list to the block chain DB.
##
## If `postProcessOk` is set, an additional verification step is added at
## the end of each transaction.
##
## if `stopFaultyHeader` is set, the function stops immediately on error.
## Otherwise the offending block is removed, the rest of the batch is
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
## adjusted and applied again repeatedly.
result = ap
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 reChainOk = false
for n in 0 ..< ap.batch.len:
block forLoop:
var headers = ap.batch[n]
while true:
if headers.len == 0:
break forLoop # continue with for loop
ap.say &"*** transaction ({n}) list: [",
headers.mapIt(&"#{it.blockNumber}").join(", "), "]"
# Realign rest of transaction to existing block chain
if reChainOk:
var parent = ap.chain.clique.db.getCanonicalHead
for i in 0 ..< headers.len:
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
headers[i].parentHash = parent.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
headers[i].blockNumber = parent.blockNumber + 1
parent = headers[i]
# Perform transaction into the block chain
let bodies = BlockBody().repeat(headers.len)
if ap.chain.persistBlocks(headers,bodies) == ValidationResult.OK:
break
if stopFaultyHeader:
return
# If the offending block is the last one of the last transaction,
# then there is nothing to do.
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 culprit = headers.filterIt(ap.failed[0] == it.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
doAssert culprit.len == 1
let number = culprit[0].blockNumber
if n + 1 == ap.batch.len and number == headers[^1].blockNumber:
return
# Remove offending block and try again for the rest
ap.say "*** persistBlocks failed, omitting block #", culprit
let prevLen = headers.len
headers = headers.filterIt(number != it.blockNumber)
doAssert headers.len < prevLen
reChainOk = true
if ap.debug:
ap.say "*** snapshot argument: #", headers[^1].blockNumber
ap.sayHeaderChain(8)
when false: # all addresses are typically pp-mappable
ap.say " address map: ", toSeq(ap.names.pairs)
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
.mapIt(&"@{it[1]}:{it[0]}")
.sorted
.join("\n" & ' '.repeat(23))
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 postProcessOk:
discard ap.clique.cliqueSnapshot(headers[^1])
# ------------------------------------------------------------------------------
# End
# ------------------------------------------------------------------------------