nimbus-eth1/nimbus/sync/legacy.nim

1131 lines
36 KiB
Nim
Raw Normal View History

Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
# nim-eth
# Copyright (c) 2018-2022 Status Research & Development GmbH
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
# Licensed and distributed under either of
# * MIT license (license terms in the root directory or at
# https://opensource.org/licenses/MIT).
# * Apache v2 license (license terms in the root directory or at
# https://www.apache.org/licenses/LICENSE-2.0).
# at your option. This file may not be copied, modified, or distributed
# except according to those terms.
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
import
std/[sets, options, random,
hashes, sequtils, math, tables, times],
chronicles,
chronos,
2022-12-02 04:39:12 +00:00
eth/p2p,
eth/p2p/[private/p2p_types, peer_pool],
stew/byteutils,
"."/[protocol, types],
../core/[chain, clique/clique_sealer, eip4844, gaslimit, withdrawals],
2022-12-02 04:39:12 +00:00
../core/pow/difficulty,
../constants,
../utils/utils,
../common/common
{.push raises:[].}
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
Snap sync simplify object inheritance (#1098) * Reorg SnapPeerBase descriptor, notably start/stop flags details: Instead of using three boolean flags startedFetch, stopped, and stopThisState a single enum type is used with values SyncRunningOk, SyncStopRequest, and SyncStopped. * Restricting snap to eth66 and later why: Id-tracked request/response wire protocol can handle overlapped responses when requests are sent in row. * Align function names with source code file names why: Easier to reconcile when following the implemented logic. * Update trace logging (want file locations) why: The macros previously used hid the relevant file location (when `chroniclesLineNumbers` turned on.) It rather printed the file location of the template that was wrapping `trace`. * Use KeyedQueue table instead of sequence why: Quick access, easy configuration as LRU or FIFO with max entries (currently LRU.) * Dissolve `SnapPeerEx` object extension into `SnapPeer` why; It is logically cleaner and more obvious not to inherit from `SnapPeerBase` but to specify opaque field object references of the merged `SnapPeer` object. These can then be locally inherited. * Dissolve `SnapSyncEx` object extension into `SnapSync` why; It is logically cleaner and more obvious not to inherit from `SnapSyncEx` but to specify opaque field object references of the `SnapPeer` object. These can then be locally inherited. Also, in the re-factored code here the interface descriptor `SnapSyncCtx` inherited `SnapSyncEx` which was sub-optimal (OO inheritance makes it easier to work with call back functions.)
2022-05-23 16:53:19 +00:00
logScope:
topics = "legacy-sync"
Snap sync simplify object inheritance (#1098) * Reorg SnapPeerBase descriptor, notably start/stop flags details: Instead of using three boolean flags startedFetch, stopped, and stopThisState a single enum type is used with values SyncRunningOk, SyncStopRequest, and SyncStopped. * Restricting snap to eth66 and later why: Id-tracked request/response wire protocol can handle overlapped responses when requests are sent in row. * Align function names with source code file names why: Easier to reconcile when following the implemented logic. * Update trace logging (want file locations) why: The macros previously used hid the relevant file location (when `chroniclesLineNumbers` turned on.) It rather printed the file location of the template that was wrapping `trace`. * Use KeyedQueue table instead of sequence why: Quick access, easy configuration as LRU or FIFO with max entries (currently LRU.) * Dissolve `SnapPeerEx` object extension into `SnapPeer` why; It is logically cleaner and more obvious not to inherit from `SnapPeerBase` but to specify opaque field object references of the merged `SnapPeer` object. These can then be locally inherited. * Dissolve `SnapSyncEx` object extension into `SnapSync` why; It is logically cleaner and more obvious not to inherit from `SnapSyncEx` but to specify opaque field object references of the `SnapPeer` object. These can then be locally inherited. Also, in the re-factored code here the interface descriptor `SnapSyncCtx` inherited `SnapSyncEx` which was sub-optimal (OO inheritance makes it easier to work with call back functions.)
2022-05-23 16:53:19 +00:00
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
const
minPeersToStartSync* = 2 # Wait for consensus of at least this
# number of peers before syncing
CleanupInterval = initDuration(minutes = 20)
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
type
#SyncStatus = enum
# syncSuccess
# syncNotEnoughPeers
# syncTimeOut
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
HashToTime = TableRef[Hash256, Time]
BlockchainSyncDefect* = object of Defect
## Catch and relay exception
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
WantedBlocksState = enum
Initial,
Requested,
Received,
Persisted
WantedBlocks = object
isHash: bool
hash: Hash256
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
startIndex: BlockNumber
numBlocks: uint
state: WantedBlocksState
headers: seq[BlockHeader]
bodies: seq[BlockBody]
LegacySyncRef* = ref object
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
workQueue: seq[WantedBlocks]
2022-12-02 04:39:12 +00:00
chain: ChainRef
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
peerPool: PeerPool
trustedPeers: HashSet[Peer]
hasOutOfOrderBlocks: bool
busyPeers: HashSet[Peer]
knownByPeer: Table[Peer, HashToTime]
lastCleanup: Time
# ------------------------------------------------------------------------------
# Private functions: sync progress
# ------------------------------------------------------------------------------
template endBlockNumber(ctx: LegacySyncRef): BlockNumber =
ctx.chain.com.syncHighest
template `endBlockNumber=`(ctx: LegacySyncRef, number: BlockNumber) =
ctx.chain.com.syncHighest = number
# Block which was downloaded and verified
template finalizedBlock(ctx: LegacySyncRef): BlockNumber =
ctx.chain.com.syncCurrent
template `finalizedBlock=`(ctx: LegacySyncRef, number: BlockNumber) =
ctx.chain.com.syncCurrent = number
# ------------------------------------------------------------------------------
# Private functions: peers related functions
# ------------------------------------------------------------------------------
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
proc hash*(p: Peer): Hash = hash(cast[pointer](p))
proc cleanupKnownByPeer(ctx: LegacySyncRef) =
let now = getTime()
var tmp = initHashSet[Hash256]()
for _, map in ctx.knownByPeer:
for hash, time in map:
if time - now >= CleanupInterval:
tmp.incl hash
for hash in tmp:
map.del(hash)
tmp.clear()
var tmpPeer = initHashSet[Peer]()
for peer, map in ctx.knownByPeer:
if map.len == 0:
tmpPeer.incl peer
for peer in tmpPeer:
ctx.knownByPeer.del peer
ctx.lastCleanup = now
proc addToKnownByPeer(ctx: LegacySyncRef,
blockHash: Hash256,
peer: Peer): bool =
var map: HashToTime
ctx.knownByPeer.withValue(peer, val) do:
map = val[]
result = true
do:
map = newTable[Hash256, Time]()
ctx.knownByPeer[peer] = map
result = false
map[blockHash] = getTime()
proc getPeers(ctx: LegacySyncRef, thisPeer: Peer): seq[Peer] =
# do not send back block/blockhash to thisPeer
for peer in peers(ctx.peerPool):
if peer != thisPeer:
result.add peer
proc handleLostPeer(ctx: LegacySyncRef) =
# TODO: ask the PeerPool for new connections and then call
# `obtainBlocksFromPeer`
discard
# ------------------------------------------------------------------------------
# Private functions: validators
# ------------------------------------------------------------------------------
proc validateDifficulty(ctx: LegacySyncRef,
2022-12-02 04:39:12 +00:00
header, parentHeader: BlockHeader,
consensusType: ConsensusType): bool =
try:
2022-12-02 04:39:12 +00:00
let com = ctx.chain.com
2022-12-02 04:39:12 +00:00
case consensusType
of ConsensusType.POA:
let rc = ctx.chain.clique.calcDifficulty(parentHeader)
if rc.isErr:
return false
if header.difficulty < rc.get():
trace "provided header difficulty is too low",
expect=rc.get(), get=header.difficulty
return false
2022-12-02 04:39:12 +00:00
of ConsensusType.POW:
let calcDiffc = com.calcDifficulty(header.timestamp, parentHeader)
if header.difficulty < calcDiffc:
trace "provided header difficulty is too low",
expect=calcDiffc, get=header.difficulty
return false
2022-12-02 04:39:12 +00:00
of ConsensusType.POS:
if header.difficulty != 0.u256:
trace "invalid difficulty",
expect=0, get=header.difficulty
return false
return true
except CatchableError as e:
error "Exception in FastSync.validateDifficulty()",
exc = e.name, err = e.msg
return false
proc validateHeader(ctx: LegacySyncRef, header: BlockHeader,
2022-12-02 04:39:12 +00:00
height = none(BlockNumber)): bool
{.raises: [CatchableError].} =
if header.parentHash == GENESIS_PARENT_HASH:
return true
let
2022-12-02 04:39:12 +00:00
db = ctx.chain.db
com = ctx.chain.com
var parentHeader: BlockHeader
if not db.getBlockHeader(header.parentHash, parentHeader):
error "can't get parentHeader",
hash=header.parentHash, number=header.blockNumber
return false
if header.blockNumber != parentHeader.blockNumber + 1.toBlockNumber:
trace "invalid block number",
expect=parentHeader.blockNumber + 1.toBlockNumber,
get=header.blockNumber
return false
if header.timestamp <= parentHeader.timestamp:
trace "invalid timestamp",
parent=parentHeader.timestamp,
header=header.timestamp
return false
2022-12-02 04:39:12 +00:00
let consensusType = com.consensus(header)
if not ctx.validateDifficulty(header, parentHeader, consensusType):
return false
2022-12-02 04:39:12 +00:00
if consensusType == ConsensusType.POA:
let period = initDuration(seconds = com.cliquePeriod)
# Timestamp diff between blocks is lower than PERIOD (clique)
if parentHeader.timestamp + period > header.timestamp:
trace "invalid timestamp diff (lower than period)",
parent=parentHeader.timestamp,
header=header.timestamp,
period
return false
2022-12-02 04:39:12 +00:00
var res = com.validateGasLimitOrBaseFee(header, parentHeader)
if res.isErr:
trace "validate gaslimit error",
msg=res.error
return false
if height.isSome:
let dif = height.get() - parentHeader.blockNumber
if not (dif < 8.toBlockNumber and dif > 1.toBlockNumber):
trace "uncle block has a parent that is too old or too young",
dif=dif,
height=height.get(),
parentNumber=parentHeader.blockNumber
return false
2022-12-02 04:39:12 +00:00
res = com.validateWithdrawals(header)
if res.isErr:
trace "validate withdrawals error",
msg=res.error
return false
res = com.validateEip4844Header(header)
if res.isErr:
trace "validate eip4844 error",
msg=res.error
return false
return true
# ------------------------------------------------------------------------------
# Private functions: sync worker
# ------------------------------------------------------------------------------
proc broadcastBlockHash(ctx: LegacySyncRef, hashes: seq[NewBlockHashesAnnounce], peers: seq[Peer]) {.async.} =
try:
var bha = newSeqOfCap[NewBlockHashesAnnounce](hashes.len)
for peer in peers:
for val in hashes:
let alreadyKnownByPeer = ctx.addToKnownByPeer(val.hash, peer)
if not alreadyKnownByPeer:
bha.add val
if bha.len > 0:
trace trEthSendNewBlockHashes, numHashes=bha.len, peer
await peer.newBlockHashes(bha)
bha.setLen(0)
except TransportError:
debug "Transport got closed during broadcastBlockHash"
except CatchableError as e:
debug "Exception in broadcastBlockHash", exc = e.name, err = e.msg
proc broadcastBlock(ctx: LegacySyncRef, blk: EthBlock, peers: seq[Peer]) {.async.} =
try:
let
db = ctx.chain.db
blockHash = blk.header.blockHash
td = db.getScore(blk.header.parentHash) +
blk.header.difficulty
for peer in peers:
let alreadyKnownByPeer = ctx.addToKnownByPeer(blockHash, peer)
if not alreadyKnownByPeer:
trace trEthSendNewBlock,
number=blk.header.blockNumber, td=td,
hash=short(blockHash), peer
await peer.newBlock(blk, td)
except TransportError:
debug "Transport got closed during broadcastBlock"
except CatchableError as e:
debug "Exception in broadcastBlock", exc = e.name, err = e.msg
proc broadcastBlockHash(ctx: LegacySyncRef, blk: EthBlock, peers: seq[Peer]) {.async.} =
try:
let bha = NewBlockHashesAnnounce(
number: blk.header.blockNumber,
hash: blk.header.blockHash
)
for peer in peers:
let alreadyKnownByPeer = ctx.addToKnownByPeer(bha.hash, peer)
if not alreadyKnownByPeer:
trace trEthSendNewBlockHashes,
number=bha.number, hash=short(bha.hash), peer
await peer.newBlockHashes([bha])
except TransportError:
debug "Transport got closed during broadcastBlockHash"
except CatchableError as e:
debug "Exception in broadcastBlockHash", exc = e.name, err = e.msg
proc sendBlockOrHash(ctx: LegacySyncRef, peer: Peer) {.async.} =
# because peer TD is lower than us,
# it become our recipient of block and block hashes
# instead of we download from it
try:
let
db = ctx.chain.db
peerBlockHash = peer.state(eth).bestBlockHash
var peerBlockHeader: BlockHeader
if not db.getBlockHeader(peerBlockHash, peerBlockHeader):
error "can't get block header", hash=short(peerBlockHash)
return
let
dist = (ctx.finalizedBlock - peerBlockHeader.blockNumber).truncate(int)
start = peerBlockHeader.blockNumber + 1.toBlockNumber
if dist == 1:
# only one block apart, send NewBlock
let number = ctx.finalizedBlock
var header: BlockHeader
var body: BlockBody
if not db.getBlockHeader(number, header):
error "can't get block header", number=number
return
let blockHash = header.blockHash
if not db.getBlockBody(blockHash, body):
error "can't get block body", number=number
return
let
ourTD = db.getScore(blockHash)
newBlock = EthBlock(
header: header,
txs: body.transactions,
uncles: body.uncles)
trace "send newBlock",
number = header.blockNumber,
hash = short(header.blockHash),
td = ourTD, peer
await peer.newBlock(newBlock, ourTD)
return
if dist > maxHeadersFetch:
# distance is too far, ignore this peer
return
# send hashes in batch
var
hash: Hash256
number = 0
hashes = newSeqOfCap[NewBlockHashesAnnounce](maxHeadersFetch)
while number < dist:
let blockNumber = start + number.toBlockNumber
if not db.getBlockHash(blockNumber, hash):
error "failed to get block hash", number=blockNumber
return
hashes.add(NewBlockHashesAnnounce(
number: blockNumber,
hash: hash))
if hashes.len == maxHeadersFetch:
trace "send newBlockHashes(batch)", numHashes=hashes.len, peer
await peer.newBlockHashes(hashes)
hashes.setLen(0)
inc number
# send the rest of hashes if available
if hashes.len > 0:
trace "send newBlockHashes(remaining)", numHashes=hashes.len, peer
await peer.newBlockHashes(hashes)
except TransportError:
debug "Transport got closed during obtainBlocksFromPeer"
except CatchableError as e:
debug "Exception in getBestBlockNumber()", exc = e.name, err = e.msg
# no need to exit here, because the context might still have blocks to fetch
# from this peer
discard e
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
proc endIndex(b: WantedBlocks): BlockNumber =
result = b.startIndex
result += (b.numBlocks - 1).toBlockNumber
proc availableWorkItem(ctx: LegacySyncRef): int =
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
var maxPendingBlock = ctx.finalizedBlock # the last downloaded & processed
trace "queue len", length = ctx.workQueue.len
result = -1
for i in 0 .. ctx.workQueue.high:
case ctx.workQueue[i].state
of Initial:
# When there is a work item at Initial state, immediatly use this one.
# This usually means a previous work item that failed somewhere in the
# process, and thus can be reused to work on.
return i
of Persisted:
# In case of Persisted, we can reset this work item to a new one.
result = i
# No break here to give work items in Initial state priority and to
# calculate endBlock.
else:
discard
# Check all endBlocks of all workqueue items to decide on next range of
# blocks to collect & process.
let endBlock = ctx.workQueue[i].endIndex
if endBlock > maxPendingBlock:
maxPendingBlock = endBlock
let nextRequestedBlock = maxPendingBlock + 1
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
# If this next block doesn't exist yet according to any of our peers, don't
# return a work item (and sync will be stopped).
if nextRequestedBlock > ctx.endBlockNumber:
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
return -1
# Increase queue when there are no free (Initial / Persisted) work items in
# the queue. At start, queue will be empty.
if result == -1:
result = ctx.workQueue.len
ctx.workQueue.setLen(result + 1)
# Create new work item when queue was increased, reset when selected work item
# is at Persisted state.
var numBlocks = (ctx.endBlockNumber - nextRequestedBlock).truncate(int) + 1
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
if numBlocks > maxHeadersFetch:
numBlocks = maxHeadersFetch
ctx.workQueue[result] = WantedBlocks(
startIndex: nextRequestedBlock,
numBlocks : numBlocks.uint,
state : Initial,
isHash : false)
proc appendWorkItem(ctx: LegacySyncRef, hash: Hash256,
startIndex: BlockNumber, numBlocks: uint) =
for i in 0 .. ctx.workQueue.high:
if ctx.workQueue[i].state == Persisted:
ctx.workQueue[i] = WantedBlocks(
isHash : true,
hash : hash,
startIndex: startIndex,
numBlocks : numBlocks,
state : Initial)
return
let i = ctx.workQueue.len
ctx.workQueue.setLen(i + 1)
ctx.workQueue[i] = WantedBlocks(
isHash : true,
hash : hash,
startIndex: startIndex,
numBlocks : numBlocks,
state : Initial)
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
proc persistWorkItem(ctx: LegacySyncRef, wi: var WantedBlocks): ValidationResult =
try:
result = ctx.chain.persistBlocks(wi.headers, wi.bodies)
except CatchableError as e:
error "storing persistent blocks failed",
error = $e.name, msg = e.msg
result = ValidationResult.Error
except Defect as e:
# Pass through
raise e
except Exception as e:
# Notorious case where the `Chain` reference applied to `persistBlocks()`
# has the compiler traced a possible `Exception` (i.e. `ctx.chain` could
# be uninitialised.)
error "exception while storing persistent blocks",
error = $e.name, msg = e.msg
raise (ref Defect)(msg: $e.name & ": " & e.msg)
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
case result
of ValidationResult.OK:
ctx.finalizedBlock = wi.endIndex
wi.state = Persisted
of ValidationResult.Error:
wi.state = Initial
# successful or not, we're done with these blocks
wi.headers = @[]
wi.bodies = @[]
proc persistPendingWorkItems(ctx: LegacySyncRef): (int, ValidationResult) =
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
var nextStartIndex = ctx.finalizedBlock + 1
var keepRunning = true
var hasOutOfOrderBlocks = false
trace "Looking for out of order blocks"
while keepRunning:
keepRunning = false
hasOutOfOrderBlocks = false
# Go over the full work queue and check for every work item if it is in
# Received state and has the next blocks in line to be processed.
for i in 0 ..< ctx.workQueue.len:
let start = ctx.workQueue[i].startIndex
# There should be at least 1 like this, namely the just received work item
# that initiated this call.
if ctx.workQueue[i].state == Received:
if start == nextStartIndex:
trace "Processing pending work item", number = start
result = (i, ctx.persistWorkItem(ctx.workQueue[i]))
# TODO: We can stop here on failure, but have to set
# hasOutofORderBlocks. Is this always valid?
nextStartIndex = ctx.finalizedBlock + 1
keepRunning = true
break
else:
hasOutOfOrderBlocks = true
ctx.hasOutOfOrderBlocks = hasOutOfOrderBlocks
proc returnWorkItem(ctx: LegacySyncRef, workItem: int): ValidationResult =
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
let wi = addr ctx.workQueue[workItem]
let askedBlocks = wi.numBlocks.int
let receivedBlocks = wi.headers.len
let start {.used.} = wi.startIndex
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
if askedBlocks == receivedBlocks:
trace "Work item complete",
start,
askedBlocks,
receivedBlocks
if wi.startIndex != ctx.finalizedBlock + 1:
trace "Blocks out of order", start, final = ctx.finalizedBlock
ctx.hasOutOfOrderBlocks = true
if ctx.hasOutOfOrderBlocks:
let (index, validation) = ctx.persistPendingWorkItems()
# Only report an error if it was this peer's work item that failed
2022-04-08 04:54:11 +00:00
if validation == ValidationResult.Error and index == workItem:
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
result = ValidationResult.Error
# TODO: What about failures on other peers' work items?
# In that case the peer will probably get disconnected on future erroneous
# work items, but before this occurs, several more blocks (that will fail)
# might get downloaded from this peer. This will delay the sync and this
# should be improved.
else:
trace "Processing work item", number = wi.startIndex
# Validation result needs to be returned so that higher up can be decided
# to disconnect from this peer in case of error.
result = ctx.persistWorkItem(wi[])
else:
trace "Work item complete but we got fewer blocks than requested, so we're ditching the whole thing.",
start,
askedBlocks,
receivedBlocks
return ValidationResult.Error
proc getBestBlockNumber(p: Peer): Future[BlockNumber] {.async.} =
let request = BlocksRequest(
startBlock: HashOrNum(isHash: true,
hash: p.state(eth).bestBlockHash),
maxResults: 1,
skip: 0,
reverse: true)
trace trEthSendSendingGetBlockHeaders, peer=p,
startBlock=request.startBlock.hash.toHex, max=request.maxResults
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
let latestBlock = await p.getBlockHeaders(request)
if latestBlock.isSome:
if latestBlock.get.headers.len > 0:
result = latestBlock.get.headers[0].blockNumber
trace trEthRecvReceivedBlockHeaders, peer=p,
Snap sync simplify object inheritance (#1098) * Reorg SnapPeerBase descriptor, notably start/stop flags details: Instead of using three boolean flags startedFetch, stopped, and stopThisState a single enum type is used with values SyncRunningOk, SyncStopRequest, and SyncStopped. * Restricting snap to eth66 and later why: Id-tracked request/response wire protocol can handle overlapped responses when requests are sent in row. * Align function names with source code file names why: Easier to reconcile when following the implemented logic. * Update trace logging (want file locations) why: The macros previously used hid the relevant file location (when `chroniclesLineNumbers` turned on.) It rather printed the file location of the template that was wrapping `trace`. * Use KeyedQueue table instead of sequence why: Quick access, easy configuration as LRU or FIFO with max entries (currently LRU.) * Dissolve `SnapPeerEx` object extension into `SnapPeer` why; It is logically cleaner and more obvious not to inherit from `SnapPeerBase` but to specify opaque field object references of the merged `SnapPeer` object. These can then be locally inherited. * Dissolve `SnapSyncEx` object extension into `SnapSync` why; It is logically cleaner and more obvious not to inherit from `SnapSyncEx` but to specify opaque field object references of the `SnapPeer` object. These can then be locally inherited. Also, in the re-factored code here the interface descriptor `SnapSyncCtx` inherited `SnapSyncEx` which was sub-optimal (OO inheritance makes it easier to work with call back functions.)
2022-05-23 16:53:19 +00:00
count=latestBlock.get.headers.len,
blockNumber=(if latestBlock.get.headers.len > 0: $result else: "missing")
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
proc toRequest(workItem: WantedBlocks): BlocksRequest =
if workItem.isHash:
BlocksRequest(
startBlock: HashOrNum(isHash: true, hash: workItem.hash),
maxResults: workItem.numBlocks,
skip: 0,
reverse: false)
else:
BlocksRequest(
startBlock: HashOrNum(isHash: false, number: workItem.startIndex),
maxResults: workItem.numBlocks,
skip: 0,
reverse: false)
type
BodyHash = object
txRoot: Hash256
uncleHash: Hash256
proc hash*(x: BodyHash): Hash =
var h: Hash = 0
h = h !& hash(x.txRoot.data)
h = h !& hash(x.uncleHash.data)
result = !$h
proc fetchBodies(ctx: LegacySyncRef, peer: Peer,
workItemIdx: int, reqBodies: seq[bool]): Future[bool] {.async.} =
template workItem: auto = ctx.workQueue[workItemIdx]
var bodies = newSeqOfCap[BlockBody](workItem.headers.len)
var hashes = newSeqOfCap[KeccakHash](maxBodiesFetch)
doAssert(reqBodies.len == workItem.headers.len)
template fetchBodies() =
trace trEthSendSendingGetBlockBodies, peer,
hashes=hashes.len
let b = await peer.getBlockBodies(hashes)
if b.isNone:
raise newException(CatchableError, "Was not able to get the block bodies")
let bodiesLen = b.get.blocks.len
trace trEthRecvReceivedBlockBodies, peer,
count=bodiesLen, requested=hashes.len
if bodiesLen == 0:
raise newException(CatchableError, "Zero block bodies received for request")
elif bodiesLen < hashes.len:
hashes.delete(0, bodiesLen - 1)
elif bodiesLen == hashes.len:
hashes.setLen(0)
else:
raise newException(CatchableError, "Too many block bodies received for request")
bodies.add(b.get.blocks)
var numRequest = 0
for i, h in workItem.headers:
if reqBodies[i]:
hashes.add(h.blockHash)
inc numRequest
if hashes.len == maxBodiesFetch:
fetchBodies()
while hashes.len != 0:
fetchBodies()
workItem.bodies = newSeqOfCap[BlockBody](workItem.headers.len)
var bodyHashes = initTable[BodyHash, int]()
for z, body in bodies:
let bodyHash = BodyHash(
txRoot: calcTxRoot(body.transactions),
uncleHash: rlpHash(body.uncles))
bodyHashes[bodyHash] = z
for i, req in reqBodies:
if req:
let bodyHash = BodyHash(
txRoot: workItem.headers[i].txRoot,
uncleHash: workItem.headers[i].ommersHash)
let z = bodyHashes.getOrDefault(bodyHash, -1)
if z == -1:
error "header missing it's body",
number=workItem.headers[i].blockNumber,
hash=workItem.headers[i].blockHash.short
return false
workItem.bodies.add bodies[z]
else:
workItem.bodies.add BlockBody()
return true
proc obtainBlocksFromPeer(ctx: LegacySyncRef, peer: Peer) {.async.} =
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
# Update our best block number
try:
if ctx.endBlockNumber <= ctx.finalizedBlock:
# endBlockNumber need update
let bestBlockNumber = await peer.getBestBlockNumber()
if bestBlockNumber > ctx.endBlockNumber:
trace "New sync end block number", number = bestBlockNumber
ctx.endBlockNumber = bestBlockNumber
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
except TransportError:
debug "Transport got closed during obtainBlocksFromPeer"
except CatchableError as e:
debug "Exception in getBestBlockNumber()", exc = e.name, err = e.msg
# no need to exit here, because the context might still have blocks to fetch
# from this peer
discard e
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
while (let workItemIdx = ctx.availableWorkItem(); workItemIdx != -1 and
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
peer.connectionState notin {Disconnecting, Disconnected}):
template workItem: auto = ctx.workQueue[workItemIdx]
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
workItem.state = Requested
trace "Requesting block headers", start = workItem.startIndex,
count = workItem.numBlocks, peer = peer.remote.node
let request = toRequest(workItem)
var dataReceived = true
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
try:
trace trEthSendSendingGetBlockHeaders, peer,
startBlock=request.startBlock.number, max=request.maxResults,
step=traceStep(request)
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
let results = await peer.getBlockHeaders(request)
if results.isSome:
trace trEthRecvReceivedBlockHeaders, peer,
count=results.get.headers.len, requested=request.maxResults
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
shallowCopy(workItem.headers, results.get.headers)
var
reqBodies = newSeqOfCap[bool](workItem.headers.len)
numRequest = 0
nextIndex = workItem.startIndex
# skip requesting empty bodies
for h in workItem.headers:
if h.blockNumber != nextIndex:
raise newException(CatchableError,
"The block numbers are not in sequence. Not processing this workItem.")
nextIndex = nextIndex + 1
let req = h.txRoot != EMPTY_ROOT_HASH or
h.ommersHash != EMPTY_UNCLE_HASH
reqBodies.add(req)
if req: inc numRequest
if numRequest > 0:
dataReceived = dataReceived and
await ctx.fetchBodies(peer, workItemIdx, reqBodies)
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
else:
# all bodies are empty
workItem.bodies.setLen(workItem.headers.len)
let peers = ctx.getPeers(peer)
if peers.len > 0:
var hashes = newSeqOfCap[NewBlockHashesAnnounce](workItem.headers.len)
for h in workItem.headers:
hashes.add NewBlockHashesAnnounce(
hash: h.blockHash,
number: h.blockNumber)
trace "broadcast block hashes", numPeers=peers.len, numHashes=hashes.len
await ctx.broadcastBlockHash(hashes, peers)
# fetchBodies can fail
dataReceived = dataReceived and true
else:
dataReceived = false
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
except TransportError:
debug "Transport got closed during obtainBlocksFromPeer",
peer
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
except CatchableError as e:
# the success case sets `dataReceived`, so we can just fall back to the
# failure path below. If we signal time-outs with exceptions such
# failures will be easier to handle.
debug "Exception in obtainBlocksFromPeer()",
exc = e.name, err = e.msg, peer
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
var giveUpOnPeer = false
if dataReceived:
trace "Finished obtaining blocks", peer, numBlocks=workItem.headers.len
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
workItem.state = Received
let res = ctx.returnWorkItem(workItemIdx)
if res != ValidationResult.OK:
trace "validation error"
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
giveUpOnPeer = true
else:
giveUpOnPeer = true
if giveUpOnPeer:
workItem.state = Initial
try:
await peer.disconnect(SubprotocolReason)
except CatchableError:
discard
ctx.handleLostPeer()
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
break
proc peersAgreeOnChain(a, b: Peer): Future[bool] {.async.} =
# Returns true if one of the peers acknowledges existence of the best block
# of another peer.
var
a = a
b = b
if a.state(eth).bestDifficulty < b.state(eth).bestDifficulty:
swap(a, b)
let request = BlocksRequest(
startBlock: HashOrNum(isHash: true,
hash: b.state(eth).bestBlockHash),
maxResults: 1,
skip: 0,
reverse: true)
trace trEthSendSendingGetBlockHeaders, peer=a,
startBlock=request.startBlock.hash.toHex, max=request.maxResults
var latestBlock: Option[blockHeadersObj]
try:
latestBlock = await a.getBlockHeaders(request)
except TransportError:
debug "Transport got closed during peersAgreeOnChain"
return
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
result = latestBlock.isSome and latestBlock.get.headers.len > 0
Snap sync simplify object inheritance (#1098) * Reorg SnapPeerBase descriptor, notably start/stop flags details: Instead of using three boolean flags startedFetch, stopped, and stopThisState a single enum type is used with values SyncRunningOk, SyncStopRequest, and SyncStopped. * Restricting snap to eth66 and later why: Id-tracked request/response wire protocol can handle overlapped responses when requests are sent in row. * Align function names with source code file names why: Easier to reconcile when following the implemented logic. * Update trace logging (want file locations) why: The macros previously used hid the relevant file location (when `chroniclesLineNumbers` turned on.) It rather printed the file location of the template that was wrapping `trace`. * Use KeyedQueue table instead of sequence why: Quick access, easy configuration as LRU or FIFO with max entries (currently LRU.) * Dissolve `SnapPeerEx` object extension into `SnapPeer` why; It is logically cleaner and more obvious not to inherit from `SnapPeerBase` but to specify opaque field object references of the merged `SnapPeer` object. These can then be locally inherited. * Dissolve `SnapSyncEx` object extension into `SnapSync` why; It is logically cleaner and more obvious not to inherit from `SnapSyncEx` but to specify opaque field object references of the `SnapPeer` object. These can then be locally inherited. Also, in the re-factored code here the interface descriptor `SnapSyncCtx` inherited `SnapSyncEx` which was sub-optimal (OO inheritance makes it easier to work with call back functions.)
2022-05-23 16:53:19 +00:00
if latestBlock.isSome:
let blockNumber {.used.} = if result: $latestBlock.get.headers[0].blockNumber
else: "missing"
trace trEthRecvReceivedBlockHeaders, peer=a,
count=latestBlock.get.headers.len, blockNumber
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
proc randomTrustedPeer(ctx: LegacySyncRef): Peer =
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
var k = rand(ctx.trustedPeers.len - 1)
var i = 0
for p in ctx.trustedPeers:
result = p
if i == k: return
inc i
proc startSyncWithPeerImpl(ctx: LegacySyncRef, peer: Peer) {.async.} =
Snap sync simplify object inheritance (#1098) * Reorg SnapPeerBase descriptor, notably start/stop flags details: Instead of using three boolean flags startedFetch, stopped, and stopThisState a single enum type is used with values SyncRunningOk, SyncStopRequest, and SyncStopped. * Restricting snap to eth66 and later why: Id-tracked request/response wire protocol can handle overlapped responses when requests are sent in row. * Align function names with source code file names why: Easier to reconcile when following the implemented logic. * Update trace logging (want file locations) why: The macros previously used hid the relevant file location (when `chroniclesLineNumbers` turned on.) It rather printed the file location of the template that was wrapping `trace`. * Use KeyedQueue table instead of sequence why: Quick access, easy configuration as LRU or FIFO with max entries (currently LRU.) * Dissolve `SnapPeerEx` object extension into `SnapPeer` why; It is logically cleaner and more obvious not to inherit from `SnapPeerBase` but to specify opaque field object references of the merged `SnapPeer` object. These can then be locally inherited. * Dissolve `SnapSyncEx` object extension into `SnapSync` why; It is logically cleaner and more obvious not to inherit from `SnapSyncEx` but to specify opaque field object references of the `SnapPeer` object. These can then be locally inherited. Also, in the re-factored code here the interface descriptor `SnapSyncCtx` inherited `SnapSyncEx` which was sub-optimal (OO inheritance makes it easier to work with call back functions.)
2022-05-23 16:53:19 +00:00
trace "Start sync", peer, trustedPeers = ctx.trustedPeers.len
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
if ctx.trustedPeers.len >= minPeersToStartSync:
# We have enough trusted peers. Validate new peer against trusted
if await peersAgreeOnChain(peer, ctx.randomTrustedPeer()):
ctx.trustedPeers.incl(peer)
asyncSpawn ctx.obtainBlocksFromPeer(peer)
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
elif ctx.trustedPeers.len == 0:
# Assume the peer is trusted, but don't start sync until we reevaluate
# it with more peers
trace "Assume trusted peer", peer
ctx.trustedPeers.incl(peer)
else:
# At this point we have some "trusted" candidates, but they are not
# "trusted" enough. We evaluate `peer` against all other candidates.
# If one of the candidates disagrees, we swap it for `peer`. If all
# candidates agree, we add `peer` to trusted set. The peers in the set
# will become "fully trusted" (and sync will start) when the set is big
# enough
var
agreeScore = 0
disagreedPeer: Peer
for tp in ctx.trustedPeers:
if await peersAgreeOnChain(peer, tp):
inc agreeScore
else:
disagreedPeer = tp
let disagreeScore = ctx.trustedPeers.len - agreeScore
if agreeScore == ctx.trustedPeers.len:
ctx.trustedPeers.incl(peer) # The best possible outcome
elif disagreeScore == 1:
trace "Peer is no longer trusted for sync", peer
ctx.trustedPeers.excl(disagreedPeer)
ctx.trustedPeers.incl(peer)
else:
trace "Peer not trusted for sync", peer
if ctx.trustedPeers.len == minPeersToStartSync:
for p in ctx.trustedPeers:
asyncSpawn ctx.obtainBlocksFromPeer(p)
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
proc startSyncWithPeer(ctx: LegacySyncRef, peer: Peer) =
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
try:
let
db = ctx.chain.db
header = db.getBlockHeader(ctx.finalizedBlock)
ourTD = db.getScore(header.blockHash)
peerTD = peer.state(eth).bestDifficulty
if peerTD <= ourTD:
# do nothing if peer have same height
if peerTD < ourTD:
trace "Peer have lower TD, become recipient",
peer, ourTD, peerTD
asyncSpawn ctx.sendBlockOrHash(peer)
return
ctx.busyPeers.incl(peer)
let f = ctx.startSyncWithPeerImpl(peer)
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
f.callback = proc(data: pointer) {.gcsafe.} =
if f.failed:
if f.error of TransportError:
debug "Transport got closed during startSyncWithPeer"
else:
error "startSyncWithPeer failed", msg = f.readError.msg, peer
ctx.busyPeers.excl(peer)
asyncSpawn f
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
except TransportError:
debug "Transport got closed during startSyncWithPeer"
except CatchableError as e:
debug "Exception in startSyncWithPeer()", exc = e.name, err = e.msg
proc startObtainBlocks(ctx: LegacySyncRef, peer: Peer) =
# simpler version of startSyncWithPeer
try:
ctx.busyPeers.incl(peer)
let f = ctx.obtainBlocksFromPeer(peer)
f.callback = proc(data: pointer) {.gcsafe.} =
if f.failed:
if f.error of TransportError:
debug "Transport got closed during startObtainBlocks"
else:
error "startObtainBlocks failed", msg = f.readError.msg, peer
ctx.busyPeers.excl(peer)
asyncSpawn f
except TransportError:
debug "Transport got closed during startObtainBlocks"
except CatchableError as e:
debug "Exception in startObtainBlocks()", exc = e.name, err = e.msg
proc onPeerConnected(ctx: LegacySyncRef, peer: Peer) =
trace "New candidate for sync", peer
ctx.startSyncWithPeer(peer)
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
proc onPeerDisconnected(ctx: LegacySyncRef, p: Peer) =
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
trace "peer disconnected ", peer = p
ctx.trustedPeers.excl(p)
ctx.busyPeers.excl(p)
ctx.knownByPeer.del(p)
# ------------------------------------------------------------------------------
# Public constructor/destructor
# ------------------------------------------------------------------------------
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
proc new*(T: type LegacySyncRef; ethNode: EthereumNode; chain: ChainRef): T
{.gcsafe, raises:[CatchableError].} =
result = LegacySyncRef(
# workQueue: n/a
# endBlockNumber: n/a
# hasOutOfOrderBlocks: n/a
chain: chain,
peerPool: ethNode.peerPool,
trustedPeers: initHashSet[Peer]())
# finalizedBlock
chain.com.syncCurrent = chain.db.getCanonicalHead().blockNumber
proc start*(ctx: LegacySyncRef) =
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
## Code for the fast blockchain sync procedure:
## <https://github.com/ethereum/wiki/wiki/Parallel-Block-Downloads>_
## <https://github.com/ethereum/go-ethereum/pull/1889__
try:
var blockHash: Hash256
let
db = ctx.chain.db
2022-12-02 04:39:12 +00:00
com = ctx.chain.com
if not db.getBlockHash(ctx.finalizedBlock, blockHash):
debug "FastSync.start: Failed to get blockHash",
number=ctx.finalizedBlock
return
2022-12-02 04:39:12 +00:00
if com.consensus == ConsensusType.POS:
debug "Fast sync is disabled after POS merge"
return
ctx.chain.com.syncStart = ctx.finalizedBlock
info "Fast Sync: start sync from",
number=ctx.chain.com.syncStart,
hash=blockHash
except CatchableError as e:
debug "Exception in FastSync.start()",
exc = e.name, err = e.msg
var po = PeerObserver(
onPeerConnected:
proc(p: Peer) {.gcsafe.} =
ctx.onPeerConnected(p),
onPeerDisconnected:
proc(p: Peer) {.gcsafe.} =
ctx.onPeerDisconnected(p))
po.setProtocol eth
ctx.peerPool.addObserver(ctx, po)
Sync: Move `blockchain_sync` code and use it with `eth/65` Move `blockchain_sync.nim` from `nim-eth` to `nimbus-eth1`. This lets `blockchain_sync` use the `eth/65` protocol to synchronise with more modern peers than before. Practically, the effect is the sync process runs more quickly and reliably than before. It finds usable peers, and they are up to date. Note, this is mostly old code, and it mostly performs "classic sync", the original Ethereum method. Here's a summary of this code: - It decides on a blockchain canonical head by sampling a few peers. - Starting from block 0 (genesis), it downloads each block header and block, mostly in order. - After it downloads each block, it executes the EVM transactions in that block and updates state trie from that, before going to the next block. - This way the database state is updated by EVM executions in block order, and new state is persisted to the trie database after each block. Even though it mentions Geth "fast sync" (comments near end of file), and has some elements, it isn't really. The most obvious missing part is this code _doesn't download a state trie_, it calculates all state from block 0. Geth "fast sync" has several parts: 1. Find an agreed common chain among several peers to treat as probably secure, and a sufficiently long suffix to provide "statistical economic consensus" when it is validated. 2. Perform a subset of PoW calculations, skipping forward over a segment to verify some of the PoWs according to a pattern in the relevant paper. 3. Download the state trie from the block at the start of that last segment. 4. Execute only the blocks/transactions in that last segment, using the downloaded state trie, to fill out the later states and properly validate the blocks in the last segment. Some other issues with `blockchain_sync` code: - If it ever reaches the head of the chain, it doesn't follow new blocks with increasing block numbers, at least not rapidly. - If the chain undergoes a reorg, this code won't fetch a block number it has already fetched, so it can't accept the reorg. It will end up conflicted with peers. This hasn't mattered because the development focus has been on the bulk of the catching up process, not the real-time head and reorgs. - So it probably doesn't work correctly when it gets close to the head due to many small reorgs, though it might for subtle reasons. - Some of the network message handling isn't sufficiently robust, and it discards some replies that have valid data according to specification. - On rare occasions the initial query mapping block hash to number can fail (because the peer's state changes). - It makes some assumptions about the state of peers based on their responses which may not be valid (I'm not convinced they are). The method for working out "trusted" peers that agree a common chain prefix is clever. It compares peers by asking each peer if it has the header matching another peer's canonical head block by hash. But it's not clear that merely knowing about a block constitutes agreement about the canonical chain. (If it did, query by block number would give the same answer more authoritatively.) Nonetheless, being able to run this sync process on `eth/65` is useful. <# interactive rebase in progress; onto 66532e8a Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-07-22 13:36:10 +00:00
# ------------------------------------------------------------------------------
# Public procs: eth wire protocol handlers
# ------------------------------------------------------------------------------
proc handleNewBlockHashes(ctx: LegacySyncRef,
peer: Peer,
hashes: openArray[NewBlockHashesAnnounce]) =
trace trEthRecvNewBlockHashes,
numHash=hashes.len
if hashes.len == 0:
return
var number = hashes[0].number
if hashes.len > 1:
for i in 1..<hashes.len:
let val = hashes[i]
if val.number != number + 1.toBlockNumber:
error "Found a gap in block hashes"
return
number = val.number
number = hashes[^1].number
if number <= ctx.endBlockNumber:
trace "Will not set new synctarget",
newSyncHeight=number,
endBlockNumber=ctx.endBlockNumber,
peer
return
# don't send back hashes to this peer
for val in hashes:
discard ctx.addToKnownByPeer(val.hash, peer)
# set new sync target, + 1'u means including last block
let numBlocks = (number - hashes[0].number).truncate(uint) + 1'u
ctx.appendWorkItem(hashes[0].hash, hashes[0].number, numBlocks)
ctx.endBlockNumber = number
trace "New sync target height", number
if ctx.busyPeers.len > 0:
# do nothing. busy peers will keep syncing
# until new sync target reached
trace "sync using busyPeers",
len=ctx.busyPeers.len
return
if ctx.trustedPeers.len == 0:
trace "sync with this peer"
ctx.startObtainBlocks(peer)
else:
trace "sync with random peer"
let peer = ctx.randomTrustedPeer()
ctx.startSyncWithPeer(peer)
proc handleNewBlock(ctx: LegacySyncRef,
peer: Peer,
blk: EthBlock,
totalDifficulty: DifficultyInt) {.
gcsafe, raises: [CatchableError].} =
trace trEthRecvNewBlock,
number=blk.header.blockNumber,
hash=short(blk.header.blockHash)
if ctx.lastCleanup - getTime() > CleanupInterval:
ctx.cleanupKnownByPeer()
# Don't send NEW_BLOCK announcement to peer that sent original new block message
discard ctx.addToKnownByPeer(blk.header.blockHash, peer)
if blk.header.blockNumber > ctx.finalizedBlock + 1.toBlockNumber:
# If the block number exceeds one past our height we cannot validate it
trace "NewBlock got block past our height",
number=blk.header.blockNumber
return
if not ctx.validateHeader(blk.header):
error "invalid header from peer",
peer, hash=short(blk.header.blockHash)
return
# Send NEW_BLOCK to square root of total number of peers in pool
# https://github.com/ethereum/devp2p/blob/master/caps/eth.md#block-propagation
let
numPeersToShareWith = sqrt(ctx.peerPool.len.float32).int
peers = ctx.getPeers(peer)
debug "num peers to share with",
number=numPeersToShareWith,
numPeers=peers.len
if peers.len > 0 and numPeersToShareWith > 0:
asyncSpawn ctx.broadcastBlock(blk, peers[0..<numPeersToShareWith])
var parentHash: Hash256
if not ctx.chain.db.getBlockHash(ctx.finalizedBlock, parentHash):
error "failed to get parent hash",
number=ctx.finalizedBlock
return
if parentHash == blk.header.parentHash:
# If new block is child of current chain tip, insert new block into chain
let body = BlockBody(
transactions: blk.txs,
uncles: blk.uncles
)
let res = ctx.chain.persistBlocks([blk.header], [body])
# Check if new sync target height can be set
if res == ValidationResult.OK:
ctx.endBlockNumber = blk.header.blockNumber
ctx.finalizedBlock = blk.header.blockNumber
else:
# Call handleNewBlockHashes to retrieve all blocks between chain tip and new block
let newSyncHeight = NewBlockHashesAnnounce(
number: blk.header.blockNumber,
hash: blk.header.blockHash
)
ctx.handleNewBlockHashes(peer, [newSyncHeight])
if peers.len > 0 and numPeersToShareWith > 0:
# Send `NEW_BLOCK_HASHES` message for received block to all other peers
asyncSpawn ctx.broadcastBlockHash(blk, peers[numPeersToShareWith..^1])
proc newBlockHashesHandler*(arg: pointer,
peer: Peer,
hashes: openArray[NewBlockHashesAnnounce]) =
let ctx = cast[LegacySyncRef](arg)
ctx.handleNewBlockHashes(peer, hashes)
proc newBlockHandler*(arg: pointer,
peer: Peer,
blk: EthBlock,
totalDifficulty: DifficultyInt) {.
gcsafe, raises: [CatchableError].} =
let ctx = cast[LegacySyncRef](arg)
ctx.handleNewBlock(peer, blk, totalDifficulty)
# End