nimbus-eth1/nimbus/sync/beacon/skeleton_db.nim

212 lines
6.5 KiB
Nim
Raw Normal View History

# Nimbus
# Copyright (c) 2023-2024 Status Research & Development GmbH
# 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.
import
eth/rlp,
eth/common/eth_types_rlp,
./skeleton_desc,
./skeleton_utils,
../../db/storage_types,
../../utils/utils,
../../core/chain
export
eth_types_rlp.blockHash
{.push gcsafe, raises: [].}
logScope:
topics = "skeleton"
# ------------------------------------------------------------------------------
# Private helpers
# ------------------------------------------------------------------------------
template get(sk: SkeletonRef, key: untyped): untyped =
sk.db.newKvt().get(key.toOpenArray).valueOr: EmptyBlob
template put(sk: SkeletonRef, key, val: untyped): untyped =
let rc = sk.db.newKvt().put(key.toOpenArray, val)
if rc.isErr:
raiseAssert "put() failed: " & $$rc.error
template del(sk: SkeletonRef, key: untyped): untyped =
discard sk.db.newKvt().del(key.toOpenArray)
proc append(w: var RlpWriter, s: Segment) =
w.startList(3)
w.append(s.head)
w.append(s.tail)
w.append(s.next)
proc append(w: var RlpWriter, p: Progress) =
w.startList(3)
w.append(p.segments)
w.append(p.linked)
w.append(p.canonicalHeadReset)
proc readImpl(rlp: var Rlp, T: type Segment): Segment {.raises: [RlpError].} =
rlp.tryEnterList()
Segment(
head: rlp.read(uint64),
tail: rlp.read(uint64),
next: rlp.read(Hash256),
)
proc readImpl(rlp: var Rlp, T: type Progress): Progress {.raises: [RlpError].} =
rlp.tryEnterList()
Progress(
segments: rlp.read(seq[Segment]),
linked : rlp.read(bool),
canonicalHeadReset: rlp.read(bool),
)
# ------------------------------------------------------------------------------
# Public functions
# ------------------------------------------------------------------------------
proc getHeader*(sk: SkeletonRef,
number: uint64,
onlySkeleton: bool = false): Result[Opt[BlockHeader], string] =
## Gets a block from the skeleton or canonical db by number.
try:
let rawHeader = sk.get(skeletonHeaderKey(number.BlockNumber))
if rawHeader.len != 0:
let output = rlp.decode(rawHeader, BlockHeader)
return ok(Opt.some output)
if onlySkeleton:
return ok(Opt.none BlockHeader)
# As a fallback, try to get the block from the canonical chain
# in case it is available there
var output: BlockHeader
if sk.db.getBlockHeader(number.BlockNumber, output):
return ok(Opt.some output)
ok(Opt.none BlockHeader)
except RlpError as ex:
err(ex.msg)
proc getHeader*(sk: SkeletonRef,
blockHash: Hash256,
onlySkeleton: bool = false):
Result[Opt[BlockHeader], string] =
## Gets a skeleton block from the db by hash
try:
let rawNumber = sk.get(skeletonBlockHashToNumberKey(blockHash))
if rawNumber.len != 0:
var output: BlockHeader
let number = rlp.decode(rawNumber, BlockNumber)
if sk.db.getBlockHeader(number, output):
return ok(Opt.some output)
if onlySkeleton:
return ok(Opt.none BlockHeader)
# As a fallback, try to get the block from the canonical chain
# in case it is available there
var output: BlockHeader
if sk.db.getBlockHeader(blockHash, output):
return ok(Opt.some output)
ok(Opt.none BlockHeader)
except RlpError as ex:
err(ex.msg)
proc putHeader*(sk: SkeletonRef, header: BlockHeader) =
## Writes a skeleton block header to the db by number
let encodedHeader = rlp.encode(header)
sk.put(skeletonHeaderKey(header.number), encodedHeader)
sk.put(
skeletonBlockHashToNumberKey(header.blockHash),
rlp.encode(header.number)
)
proc putBody*(sk: SkeletonRef, header: BlockHeader, body: BlockBody): Result[void, string] =
## Writes block body to db
try:
let
encodedBody = rlp.encode(body)
bodyHash = sumHash(body)
headerHash = header.blockHash
keyHash = sumHash(headerHash, bodyHash)
sk.put(skeletonBodyKey(keyHash), encodedBody)
ok()
except CatchableError as ex:
err(ex.msg)
proc getBody*(sk: SkeletonRef, header: BlockHeader): Result[Opt[BlockBody], string] =
## Reads block body from db
## sumHash is the hash of [txRoot, ommersHash, wdRoot]
try:
let
bodyHash = header.sumHash
headerHash = header.blockHash
keyHash = sumHash(headerHash, bodyHash)
rawBody = sk.get(skeletonBodyKey(keyHash))
if rawBody.len > 0:
return ok(Opt.some rlp.decode(rawBody, BlockBody))
ok(Opt.none BlockBody)
except RlpError as ex:
err(ex.msg)
proc writeProgress*(sk: SkeletonRef) =
## Writes the progress to db
for sub in sk.subchains:
debug "Writing sync progress subchains", sub
let encodedProgress = rlp.encode(sk.progress)
sk.put(skeletonProgressKey(), encodedProgress)
proc readProgress*(sk: SkeletonRef): Result[void, string] =
## Reads the SkeletonProgress from db
try:
let rawProgress = sk.get(skeletonProgressKey())
if rawProgress.len == 0:
return ok()
sk.progress = rlp.decode(rawProgress, Progress)
ok()
except RlpError as ex:
err(ex.msg)
proc deleteHeaderAndBody*(sk: SkeletonRef, header: BlockHeader) =
## Deletes a skeleton block from the db by number
sk.del(skeletonHeaderKey(header.number))
sk.del(skeletonBlockHashToNumberKey(header.blockHash))
sk.del(skeletonBodyKey(header.sumHash))
proc canonicalHead*(sk: SkeletonRef): Result[BlockHeader, string] =
## Returns Opt.some or error, never returns Opt.none
try:
ok(sk.db.getCanonicalHead())
except CatchableError as ex:
err(ex.msg)
proc resetCanonicalHead*(sk: SkeletonRef, newHead, oldHead: uint64) =
debug "RESET CANONICAL", newHead, oldHead
sk.chain.com.syncCurrent = newHead.BlockNumber
proc insertBlocks*(sk: SkeletonRef,
Consolidate block type for block processing (#2325) This PR consolidates the split header-body sequences into a single EthBlock sequence and cleans up the fallout from that which significantly reduces block processing overhead during import thanks to less garbage collection and fewer copies of things all around. Notably, since the number of headers must always match the number of bodies, we also get rid of a pointless degree of freedom that in the future could introduce unnecessary bugs. * only read header and body from era file * avoid several unnecessary copies along the block processing way * simplify signatures, cleaning up unused arguemnts and returns * use `stew/assign2` in a few strategic places where the generated nim assignent is slow and add a few `move` to work around poor analysis in nim 1.6 (will need to be revisited for 2.0) ``` stats-20240607_2223-a814aa0b.csv vs stats-20240608_0714-21c1d0a9.csv bps_x bps_y tps_x tps_y bpsd tpsd timed block_number (498305, 713245] 1,540.52 1,809.73 2,361.58 2775.340189 17.63% 17.63% -14.92% (713245, 928185] 730.36 865.26 1,715.90 2028.973852 18.01% 18.01% -15.21% (928185, 1143126] 663.03 789.10 2,529.26 3032.490771 19.79% 19.79% -16.28% (1143126, 1358066] 393.46 508.05 2,152.50 2777.578119 29.13% 29.13% -22.50% (1358066, 1573007] 370.88 440.72 2,351.31 2791.896052 18.81% 18.81% -15.80% (1573007, 1787947] 283.65 335.11 2,068.93 2441.373402 17.60% 17.60% -14.91% (1787947, 2002888] 287.29 342.11 2,078.39 2474.179448 18.99% 18.99% -15.91% (2002888, 2217828] 293.38 343.16 2,208.83 2584.77457 17.16% 17.16% -14.61% (2217828, 2432769] 140.09 167.86 1,081.87 1296.336926 18.82% 18.82% -15.80% blocks: 1934464, baseline: 3h13m1s, contender: 2h43m47s bpsd (mean): 19.55% tpsd (mean): 19.55% Time (total): -29m13s, -15.14% ```
2024-06-09 14:32:20 +00:00
blocks: openArray[EthBlock],
fromEngine: bool): Result[uint64, string] =
Consolidate block type for block processing (#2325) This PR consolidates the split header-body sequences into a single EthBlock sequence and cleans up the fallout from that which significantly reduces block processing overhead during import thanks to less garbage collection and fewer copies of things all around. Notably, since the number of headers must always match the number of bodies, we also get rid of a pointless degree of freedom that in the future could introduce unnecessary bugs. * only read header and body from era file * avoid several unnecessary copies along the block processing way * simplify signatures, cleaning up unused arguemnts and returns * use `stew/assign2` in a few strategic places where the generated nim assignent is slow and add a few `move` to work around poor analysis in nim 1.6 (will need to be revisited for 2.0) ``` stats-20240607_2223-a814aa0b.csv vs stats-20240608_0714-21c1d0a9.csv bps_x bps_y tps_x tps_y bpsd tpsd timed block_number (498305, 713245] 1,540.52 1,809.73 2,361.58 2775.340189 17.63% 17.63% -14.92% (713245, 928185] 730.36 865.26 1,715.90 2028.973852 18.01% 18.01% -15.21% (928185, 1143126] 663.03 789.10 2,529.26 3032.490771 19.79% 19.79% -16.28% (1143126, 1358066] 393.46 508.05 2,152.50 2777.578119 29.13% 29.13% -22.50% (1358066, 1573007] 370.88 440.72 2,351.31 2791.896052 18.81% 18.81% -15.80% (1573007, 1787947] 283.65 335.11 2,068.93 2441.373402 17.60% 17.60% -14.91% (1787947, 2002888] 287.29 342.11 2,078.39 2474.179448 18.99% 18.99% -15.91% (2002888, 2217828] 293.38 343.16 2,208.83 2584.77457 17.16% 17.16% -14.61% (2217828, 2432769] 140.09 167.86 1,081.87 1296.336926 18.82% 18.82% -15.80% blocks: 1934464, baseline: 3h13m1s, contender: 2h43m47s bpsd (mean): 19.55% tpsd (mean): 19.55% Time (total): -29m13s, -15.14% ```
2024-06-09 14:32:20 +00:00
discard ? sk.chain.persistBlocks(blocks)
ok(blocks.len.uint64)
proc insertBlock*(sk: SkeletonRef,
header: BlockHeader,
fromEngine: bool): Result[uint64, string] =
let maybeBody = sk.getBody(header).valueOr:
return err(error)
if maybeBody.isNone:
return err("insertBlock: Block body not found: " & $header.u64)
Consolidate block type for block processing (#2325) This PR consolidates the split header-body sequences into a single EthBlock sequence and cleans up the fallout from that which significantly reduces block processing overhead during import thanks to less garbage collection and fewer copies of things all around. Notably, since the number of headers must always match the number of bodies, we also get rid of a pointless degree of freedom that in the future could introduce unnecessary bugs. * only read header and body from era file * avoid several unnecessary copies along the block processing way * simplify signatures, cleaning up unused arguemnts and returns * use `stew/assign2` in a few strategic places where the generated nim assignent is slow and add a few `move` to work around poor analysis in nim 1.6 (will need to be revisited for 2.0) ``` stats-20240607_2223-a814aa0b.csv vs stats-20240608_0714-21c1d0a9.csv bps_x bps_y tps_x tps_y bpsd tpsd timed block_number (498305, 713245] 1,540.52 1,809.73 2,361.58 2775.340189 17.63% 17.63% -14.92% (713245, 928185] 730.36 865.26 1,715.90 2028.973852 18.01% 18.01% -15.21% (928185, 1143126] 663.03 789.10 2,529.26 3032.490771 19.79% 19.79% -16.28% (1143126, 1358066] 393.46 508.05 2,152.50 2777.578119 29.13% 29.13% -22.50% (1358066, 1573007] 370.88 440.72 2,351.31 2791.896052 18.81% 18.81% -15.80% (1573007, 1787947] 283.65 335.11 2,068.93 2441.373402 17.60% 17.60% -14.91% (1787947, 2002888] 287.29 342.11 2,078.39 2474.179448 18.99% 18.99% -15.91% (2002888, 2217828] 293.38 343.16 2,208.83 2584.77457 17.16% 17.16% -14.61% (2217828, 2432769] 140.09 167.86 1,081.87 1296.336926 18.82% 18.82% -15.80% blocks: 1934464, baseline: 3h13m1s, contender: 2h43m47s bpsd (mean): 19.55% tpsd (mean): 19.55% Time (total): -29m13s, -15.14% ```
2024-06-09 14:32:20 +00:00
sk.insertBlocks([EthBlock.init(header, maybeBody.get)], fromEngine)