nimbus-eth2/beacon_chain/nimbus_beacon_node.nim

2130 lines
78 KiB
Nim
Raw Normal View History

# beacon_chain
# Copyright (c) 2018-2023 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.
2022-07-29 10:53:42 +00:00
when (NimMajor, NimMinor) < (1, 4):
{.push raises: [Defect].}
else:
{.push raises: [].}
import
std/[os, random, sequtils, terminal, times],
2022-06-21 08:29:16 +00:00
chronos, chronicles, chronicles/chronos_tools,
metrics, metrics/chronos_httpserver,
stew/[byteutils, io2],
eth/p2p/discoveryv5/[enr, random2],
eth/keys,
./consensus_object_pools/vanity_logs/vanity_logs,
./networking/topic_params,
./rpc/[rest_api, state_ttl_cache],
./spec/datatypes/[altair, bellatrix, phase0],
./spec/[deposit_snapshots, engine_authentication, weak_subjectivity],
./validators/[keystore_management, validator_duties],
"."/[
beacon_node, beacon_node_light_client, deposits, interop,
nimbus_binary_common, statusbar, trusted_node_sync, wallets]
when defined(posix):
import system/ansi_c
from ./spec/datatypes/eip4844 import SignedBeaconBlock
from
libp2p/protocols/pubsub/gossipsub
import
TopicParams, validateParameters, init
when defined(windows):
import winlean
type
LPCSTR* = cstring
LPSTR* = cstring
SERVICE_STATUS* {.final, pure.} = object
dwServiceType*: DWORD
dwCurrentState*: DWORD
dwControlsAccepted*: DWORD
dwWin32ExitCode*: DWORD
dwServiceSpecificExitCode*: DWORD
dwCheckPoint*: DWORD
dwWaitHint*: DWORD
SERVICE_STATUS_HANDLE* = DWORD
LPSERVICE_STATUS* = ptr SERVICE_STATUS
LPSERVICE_MAIN_FUNCTION* = proc (para1: DWORD, para2: LPSTR) {.stdcall.}
SERVICE_TABLE_ENTRY* {.final, pure.} = object
lpServiceName*: LPSTR
lpServiceProc*: LPSERVICE_MAIN_FUNCTION
LPSERVICE_TABLE_ENTRY* = ptr SERVICE_TABLE_ENTRY
LPHANDLER_FUNCTION* = proc (para1: DWORD): WINBOOL{.stdcall.}
const
SERVICE_WIN32_OWN_PROCESS = 16
SERVICE_RUNNING = 4
SERVICE_STOPPED = 1
SERVICE_START_PENDING = 2
SERVICE_STOP_PENDING = 3
SERVICE_CONTROL_STOP = 1
SERVICE_CONTROL_PAUSE = 2
SERVICE_CONTROL_CONTINUE = 3
SERVICE_CONTROL_INTERROGATE = 4
SERVICE_ACCEPT_STOP = 1
NO_ERROR = 0
SERVICE_NAME = LPCSTR "NIMBUS_BEACON_NODE"
var
gSvcStatusHandle: SERVICE_STATUS_HANDLE
gSvcStatus: SERVICE_STATUS
proc reportServiceStatus*(dwCurrentState, dwWin32ExitCode, dwWaitHint: DWORD) {.gcsafe.}
proc StartServiceCtrlDispatcher*(lpServiceStartTable: LPSERVICE_TABLE_ENTRY): WINBOOL{.
stdcall, dynlib: "advapi32", importc: "StartServiceCtrlDispatcherA".}
proc SetServiceStatus*(hServiceStatus: SERVICE_STATUS_HANDLE,
lpServiceStatus: LPSERVICE_STATUS): WINBOOL{.stdcall,
dynlib: "advapi32", importc: "SetServiceStatus".}
proc RegisterServiceCtrlHandler*(lpServiceName: LPCSTR,
lpHandlerProc: LPHANDLER_FUNCTION): SERVICE_STATUS_HANDLE{.
stdcall, dynlib: "advapi32", importc: "RegisterServiceCtrlHandlerA".}
type
RpcServer = RpcHttpServer
2020-06-05 15:08:50 +00:00
template init(T: type RpcHttpServer, ip: ValidIpAddress, port: Port): T =
newRpcHttpServer([initTAddress(ip, port)])
# https://github.com/ethereum/eth2.0-metrics/blob/master/metrics.md#interop-metrics
declareGauge beacon_slot, "Latest slot of the beacon chain state"
declareGauge beacon_current_epoch, "Current epoch"
# Finalization tracking
declareGauge finalization_delay,
"Epoch delay between scheduled epoch and finalized epoch"
declareGauge ticks_delay,
"How long does to take to run the onSecond loop"
declareGauge next_action_wait,
"Seconds until the next attestation will be sent"
logScope: topics = "beacnde"
func getVanityLogs(stdoutKind: StdoutLogKind): VanityLogs =
case stdoutKind
of StdoutLogKind.Auto: raiseAssert "inadmissable here"
of StdoutLogKind.Colors:
VanityLogs(
onMergeTransitionBlock: color🐼,
onFinalizedMergeTransitionBlock: blink🐼,
onUpgradeToCapella: color🦉)
of StdoutLogKind.NoColors:
VanityLogs(
onMergeTransitionBlock: mono🐼,
onFinalizedMergeTransitionBlock: mono🐼,
onUpgradeToCapella: mono🦉)
of StdoutLogKind.Json, StdoutLogKind.None:
VanityLogs(
onMergeTransitionBlock:
(proc() = notice "🐼 Proof of Stake Activated 🐼"),
onFinalizedMergeTransitionBlock:
(proc() = notice "🐼 Proof of Stake Finalized 🐼"),
onUpgradeToCapella:
(proc() = notice "🦉 Withdrowls now available 🦉"))
proc loadChainDag(
config: BeaconNodeConf,
cfg: RuntimeConfig,
db: BeaconChainDB,
eventBus: EventBus,
validatorMonitor: ref ValidatorMonitor,
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
networkGenesisValidatorsRoot: Opt[Eth2Digest],
shouldEnableTestFeatures: bool): ChainDAGRef =
info "Loading block DAG from database", path = config.databaseDir
var dag: ChainDAGRef
proc onLightClientFinalityUpdate(data: ForkedLightClientFinalityUpdate) =
if dag == nil: return
withForkyFinalityUpdate(data):
when lcDataFork >= LightClientDataFork.Altair:
let contextFork =
dag.cfg.stateForkAtEpoch(forkyFinalityUpdate.contextEpoch)
eventBus.finUpdateQueue.emit(
RestVersioned[ForkedLightClientFinalityUpdate](
data: data,
jsonVersion: contextFork,
sszContext: dag.forkDigests[].atStateFork(contextFork)))
proc onLightClientOptimisticUpdate(data: ForkedLightClientOptimisticUpdate) =
if dag == nil: return
withForkyOptimisticUpdate(data):
when lcDataFork >= LightClientDataFork.Altair:
let contextFork =
dag.cfg.stateForkAtEpoch(forkyOptimisticUpdate.contextEpoch)
eventBus.optUpdateQueue.emit(
RestVersioned[ForkedLightClientOptimisticUpdate](
data: data,
jsonVersion: contextFork,
sszContext: dag.forkDigests[].atStateFork(contextFork)))
let
extraFlags =
if shouldEnableTestFeatures: {enableTestFeatures, lowParticipation}
else: {enableTestFeatures}
chainDagFlags =
if config.strictVerification: {strictVerification}
else: {}
onLightClientFinalityUpdateCb =
if config.lightClientDataServe: onLightClientFinalityUpdate
else: nil
onLightClientOptimisticUpdateCb =
if config.lightClientDataServe: onLightClientOptimisticUpdate
else: nil
dag = ChainDAGRef.init(
cfg, db, validatorMonitor, extraFlags + chainDagFlags, config.eraDir,
vanityLogs = getVanityLogs(detectTTY(config.logStdout)),
lcDataConfig = LightClientDataConfig(
serve: config.lightClientDataServe,
importMode: config.lightClientDataImportMode,
maxPeriods: config.lightClientDataMaxPeriods,
onLightClientFinalityUpdate: onLightClientFinalityUpdateCb,
onLightClientOptimisticUpdate: onLightClientOptimisticUpdateCb))
if networkGenesisValidatorsRoot.isSome:
let databaseGenesisValidatorsRoot =
getStateField(dag.headState, genesis_validators_root)
if networkGenesisValidatorsRoot.get != databaseGenesisValidatorsRoot:
fatal "The specified --data-dir contains data for a different network",
networkGenesisValidatorsRoot = networkGenesisValidatorsRoot.get,
databaseGenesisValidatorsRoot,
dataDir = config.dataDir
quit 1
# The first pruning after restart may take a while..
if config.historyMode == HistoryMode.Prune:
dag.pruneHistory(true)
dag
proc checkWeakSubjectivityCheckpoint(
dag: ChainDAGRef,
wsCheckpoint: Checkpoint,
beaconClock: BeaconClock) =
let
currentSlot = beaconClock.now.slotOrZero
isCheckpointStale = not is_within_weak_subjectivity_period(
dag.cfg, currentSlot, dag.headState, wsCheckpoint)
if isCheckpointStale:
error "Weak subjectivity checkpoint is stale",
currentSlot, checkpoint = wsCheckpoint,
headStateSlot = getStateField(dag.headState, slot)
quit 1
proc initFullNode(
node: BeaconNode,
2022-06-21 08:29:16 +00:00
rng: ref HmacDrbgContext,
dag: ChainDAGRef,
taskpool: TaskPoolPtr,
getBeaconTime: GetBeaconTimeFn) =
template config(): auto = node.config
proc onAttestationReceived(data: Attestation) =
node.eventBus.attestQueue.emit(data)
proc onSyncContribution(data: SignedContributionAndProof) =
node.eventBus.contribQueue.emit(data)
proc onVoluntaryExitAdded(data: SignedVoluntaryExit) =
node.eventBus.exitQueue.emit(data)
proc onBlockAdded(data: ForkedTrustedSignedBeaconBlock) =
let optimistic =
if node.currentSlot().epoch() >= dag.cfg.BELLATRIX_FORK_EPOCH:
some node.dag.is_optimistic(data.root)
else:
none[bool]()
node.eventBus.blocksQueue.emit(
EventBeaconBlockObject.init(data, optimistic))
proc onHeadChanged(data: HeadChangeInfoObject) =
let eventData =
if node.currentSlot().epoch() >= dag.cfg.BELLATRIX_FORK_EPOCH:
var res = data
res.optimistic = some node.dag.is_optimistic(data.block_root)
res
else:
data
node.eventBus.headQueue.emit(eventData)
proc onChainReorg(data: ReorgInfoObject) =
let eventData =
if node.currentSlot().epoch() >= dag.cfg.BELLATRIX_FORK_EPOCH:
var res = data
res.optimistic = some node.dag.is_optimistic(data.new_head_block)
res
else:
data
node.eventBus.reorgQueue.emit(eventData)
proc makeOnFinalizationCb(
# This `nimcall` functions helps for keeping track of what
# needs to be captured by the onFinalization closure.
eventBus: EventBus,
eth1Monitor: Eth1Monitor): OnFinalizedCallback {.nimcall.} =
static: doAssert (eth1Monitor is ref)
return proc(dag: ChainDAGRef, data: FinalizationInfoObject) =
if eth1Monitor != nil:
let finalizedEpochRef = dag.getFinalizedEpochRef()
discard trackFinalizedState(eth1Monitor,
finalizedEpochRef.eth1_data,
finalizedEpochRef.eth1_deposit_index)
node.updateLightClientFromDag()
let eventData =
if node.currentSlot().epoch() >= dag.cfg.BELLATRIX_FORK_EPOCH:
var res = data
res.optimistic = some node.dag.is_optimistic(data.block_root)
res
else:
data
eventBus.finalQueue.emit(eventData)
func getLocalHeadSlot(): Slot =
dag.head.slot
proc getLocalWallSlot(): Slot =
node.beaconClock.now.slotOrZero
func getFirstSlotAtFinalizedEpoch(): Slot =
dag.finalizedHead.slot
func getBackfillSlot(): Slot =
dag.backfill.slot
era: load blocks and states (#3394) * era: load blocks and states Era files contain finalized history and can be thought of as an alternative source for block and state data that allows clients to avoid syncing this information from the P2P network - the P2P network is then used to "top up" the client with the most recent data. They can be freely shared in the community via whatever means (http, torrent, etc) and serve as a permanent cold store of consensus data (and, after the merge, execution data) for history buffs and bean counters alike. This PR gently introduces support for loading blocks and states in two cases: block requests from rest/p2p and frontfilling when doing checkpoint sync. The era files are used as a secondary source if the information is not found in the database - compared to the database, there are a few key differences: * the database stores the block indexed by block root while the era file indexes by slot - the former is used only in rest, while the latter is used both by p2p and rest. * when loading blocks from era files, the root is no longer trivially available - if it is needed, it must either be computed (slow) or cached (messy) - the good news is that for p2p requests, it is not needed * in era files, "framed" snappy encoding is used while in the database we store unframed snappy - for p2p2 requests, the latter requires recompression while the former could avoid it * front-filling is the process of using era files to replace backfilling - in theory this front-filling could happen from any block and front-fills with gaps could also be entertained, but our backfilling algorithm cannot take advantage of this because there's no (simple) way to tell it to "skip" a range. * front-filling, as implemented, is a bit slow (10s to load mainnet): we load the full BeaconState for every era to grab the roots of the blocks - it would be better to partially load the state - as such, it would also be good to be able to partially decompress snappy blobs * lookups from REST via root are served by first looking up a block summary in the database, then using the slot to load the block data from the era file - however, there needs to be an option to create the summary table from era files to fully support historical queries To test this, `ncli_db` has an era file exporter: the files it creates should be placed in an `era` folder next to `db` in the data directory. What's interesting in particular about this setup is that `db` remains as the source of truth for security purposes - it stores the latest synced head root which in turn determines where a node "starts" its consensus participation - the era directory however can be freely shared between nodes / people without any (significant) security implications, assuming the era files are consistent / not broken. There's lots of future improvements to be had: * we can drop the in-memory `BlockRef` index almost entirely - at this point, resident memory usage of Nimbus should drop to a cool 500-600 mb * we could serve era files via REST trivially: this would drop backfill times to whatever time it takes to download the files - unlike the current implementation that downloads block by block, downloading an era at a time almost entirely cuts out request overhead * we can "reasonably" recreate detailed state history from almost any point in time, turning an O(slot) process into O(1) effectively - we'll still need caches and indices to do this with sufficient efficiency for the rest api, but at least it cuts the whole process down to minutes instead of hours, for arbitrary points in time * CI: ignore failures with Nim-1.6 (temporary) * test fixes Co-authored-by: Ștefan Talpalaru <stefantalpalaru@yahoo.com>
2022-03-23 08:58:17 +00:00
func getFrontfillSlot(): Slot =
max(dag.frontfill.get(BlockId()).slot, dag.horizon)
era: load blocks and states (#3394) * era: load blocks and states Era files contain finalized history and can be thought of as an alternative source for block and state data that allows clients to avoid syncing this information from the P2P network - the P2P network is then used to "top up" the client with the most recent data. They can be freely shared in the community via whatever means (http, torrent, etc) and serve as a permanent cold store of consensus data (and, after the merge, execution data) for history buffs and bean counters alike. This PR gently introduces support for loading blocks and states in two cases: block requests from rest/p2p and frontfilling when doing checkpoint sync. The era files are used as a secondary source if the information is not found in the database - compared to the database, there are a few key differences: * the database stores the block indexed by block root while the era file indexes by slot - the former is used only in rest, while the latter is used both by p2p and rest. * when loading blocks from era files, the root is no longer trivially available - if it is needed, it must either be computed (slow) or cached (messy) - the good news is that for p2p requests, it is not needed * in era files, "framed" snappy encoding is used while in the database we store unframed snappy - for p2p2 requests, the latter requires recompression while the former could avoid it * front-filling is the process of using era files to replace backfilling - in theory this front-filling could happen from any block and front-fills with gaps could also be entertained, but our backfilling algorithm cannot take advantage of this because there's no (simple) way to tell it to "skip" a range. * front-filling, as implemented, is a bit slow (10s to load mainnet): we load the full BeaconState for every era to grab the roots of the blocks - it would be better to partially load the state - as such, it would also be good to be able to partially decompress snappy blobs * lookups from REST via root are served by first looking up a block summary in the database, then using the slot to load the block data from the era file - however, there needs to be an option to create the summary table from era files to fully support historical queries To test this, `ncli_db` has an era file exporter: the files it creates should be placed in an `era` folder next to `db` in the data directory. What's interesting in particular about this setup is that `db` remains as the source of truth for security purposes - it stores the latest synced head root which in turn determines where a node "starts" its consensus participation - the era directory however can be freely shared between nodes / people without any (significant) security implications, assuming the era files are consistent / not broken. There's lots of future improvements to be had: * we can drop the in-memory `BlockRef` index almost entirely - at this point, resident memory usage of Nimbus should drop to a cool 500-600 mb * we could serve era files via REST trivially: this would drop backfill times to whatever time it takes to download the files - unlike the current implementation that downloads block by block, downloading an era at a time almost entirely cuts out request overhead * we can "reasonably" recreate detailed state history from almost any point in time, turning an O(slot) process into O(1) effectively - we'll still need caches and indices to do this with sufficient efficiency for the rest api, but at least it cuts the whole process down to minutes instead of hours, for arbitrary points in time * CI: ignore failures with Nim-1.6 (temporary) * test fixes Co-authored-by: Ștefan Talpalaru <stefantalpalaru@yahoo.com>
2022-03-23 08:58:17 +00:00
let
quarantine = newClone(
Quarantine.init())
attestationPool = newClone(
AttestationPool.init(dag, quarantine, onAttestationReceived))
syncCommitteeMsgPool = newClone(
SyncCommitteeMsgPool.init(rng, onSyncContribution))
lightClientPool = newClone(
LightClientPool())
exitPool = newClone(
ExitPool.init(dag, attestationPool, onVoluntaryExitAdded))
consensusManager = ConsensusManager.new(
dag, attestationPool, quarantine, node.eth1Monitor,
ActionTracker.init(rng, config.subscribeAllSubnets),
node.dynamicFeeRecipientsStore, config.validatorsDir,
config.defaultFeeRecipient)
blockProcessor = BlockProcessor.new(
config.dumpEnabled, config.dumpDirInvalid, config.dumpDirIncoming,
rng, taskpool, consensusManager, node.validatorMonitor, getBeaconTime)
blockVerifier = proc(signedBlock: ForkedSignedBeaconBlock):
Future[Result[void, VerifierError]] =
# The design with a callback for block verification is unusual compared
# to the rest of the application, but fits with the general approach
# taken in the sync/request managers - this is an architectural compromise
# that should probably be reimagined more holistically in the future.
let resfut = newFuture[Result[void, VerifierError]]("blockVerifier")
blockProcessor[].addBlock(MsgSource.gossip, signedBlock,
Opt.none(eip4844.BlobsSidecar), resfut)
resfut
processor = Eth2Processor.new(
config.doppelgangerDetection,
blockProcessor, node.validatorMonitor, dag, attestationPool, exitPool,
node.attachedValidators, syncCommitteeMsgPool, lightClientPool,
quarantine, rng, getBeaconTime, taskpool)
syncManager = newSyncManager[Peer, PeerId](
node.network.peerPool, SyncQueueKind.Forward, getLocalHeadSlot,
getLocalWallSlot, getFirstSlotAtFinalizedEpoch, getBackfillSlot,
era: load blocks and states (#3394) * era: load blocks and states Era files contain finalized history and can be thought of as an alternative source for block and state data that allows clients to avoid syncing this information from the P2P network - the P2P network is then used to "top up" the client with the most recent data. They can be freely shared in the community via whatever means (http, torrent, etc) and serve as a permanent cold store of consensus data (and, after the merge, execution data) for history buffs and bean counters alike. This PR gently introduces support for loading blocks and states in two cases: block requests from rest/p2p and frontfilling when doing checkpoint sync. The era files are used as a secondary source if the information is not found in the database - compared to the database, there are a few key differences: * the database stores the block indexed by block root while the era file indexes by slot - the former is used only in rest, while the latter is used both by p2p and rest. * when loading blocks from era files, the root is no longer trivially available - if it is needed, it must either be computed (slow) or cached (messy) - the good news is that for p2p requests, it is not needed * in era files, "framed" snappy encoding is used while in the database we store unframed snappy - for p2p2 requests, the latter requires recompression while the former could avoid it * front-filling is the process of using era files to replace backfilling - in theory this front-filling could happen from any block and front-fills with gaps could also be entertained, but our backfilling algorithm cannot take advantage of this because there's no (simple) way to tell it to "skip" a range. * front-filling, as implemented, is a bit slow (10s to load mainnet): we load the full BeaconState for every era to grab the roots of the blocks - it would be better to partially load the state - as such, it would also be good to be able to partially decompress snappy blobs * lookups from REST via root are served by first looking up a block summary in the database, then using the slot to load the block data from the era file - however, there needs to be an option to create the summary table from era files to fully support historical queries To test this, `ncli_db` has an era file exporter: the files it creates should be placed in an `era` folder next to `db` in the data directory. What's interesting in particular about this setup is that `db` remains as the source of truth for security purposes - it stores the latest synced head root which in turn determines where a node "starts" its consensus participation - the era directory however can be freely shared between nodes / people without any (significant) security implications, assuming the era files are consistent / not broken. There's lots of future improvements to be had: * we can drop the in-memory `BlockRef` index almost entirely - at this point, resident memory usage of Nimbus should drop to a cool 500-600 mb * we could serve era files via REST trivially: this would drop backfill times to whatever time it takes to download the files - unlike the current implementation that downloads block by block, downloading an era at a time almost entirely cuts out request overhead * we can "reasonably" recreate detailed state history from almost any point in time, turning an O(slot) process into O(1) effectively - we'll still need caches and indices to do this with sufficient efficiency for the rest api, but at least it cuts the whole process down to minutes instead of hours, for arbitrary points in time * CI: ignore failures with Nim-1.6 (temporary) * test fixes Co-authored-by: Ștefan Talpalaru <stefantalpalaru@yahoo.com>
2022-03-23 08:58:17 +00:00
getFrontfillSlot, dag.tail.slot, blockVerifier)
backfiller = newSyncManager[Peer, PeerId](
node.network.peerPool, SyncQueueKind.Backward, getLocalHeadSlot,
getLocalWallSlot, getFirstSlotAtFinalizedEpoch, getBackfillSlot,
era: load blocks and states (#3394) * era: load blocks and states Era files contain finalized history and can be thought of as an alternative source for block and state data that allows clients to avoid syncing this information from the P2P network - the P2P network is then used to "top up" the client with the most recent data. They can be freely shared in the community via whatever means (http, torrent, etc) and serve as a permanent cold store of consensus data (and, after the merge, execution data) for history buffs and bean counters alike. This PR gently introduces support for loading blocks and states in two cases: block requests from rest/p2p and frontfilling when doing checkpoint sync. The era files are used as a secondary source if the information is not found in the database - compared to the database, there are a few key differences: * the database stores the block indexed by block root while the era file indexes by slot - the former is used only in rest, while the latter is used both by p2p and rest. * when loading blocks from era files, the root is no longer trivially available - if it is needed, it must either be computed (slow) or cached (messy) - the good news is that for p2p requests, it is not needed * in era files, "framed" snappy encoding is used while in the database we store unframed snappy - for p2p2 requests, the latter requires recompression while the former could avoid it * front-filling is the process of using era files to replace backfilling - in theory this front-filling could happen from any block and front-fills with gaps could also be entertained, but our backfilling algorithm cannot take advantage of this because there's no (simple) way to tell it to "skip" a range. * front-filling, as implemented, is a bit slow (10s to load mainnet): we load the full BeaconState for every era to grab the roots of the blocks - it would be better to partially load the state - as such, it would also be good to be able to partially decompress snappy blobs * lookups from REST via root are served by first looking up a block summary in the database, then using the slot to load the block data from the era file - however, there needs to be an option to create the summary table from era files to fully support historical queries To test this, `ncli_db` has an era file exporter: the files it creates should be placed in an `era` folder next to `db` in the data directory. What's interesting in particular about this setup is that `db` remains as the source of truth for security purposes - it stores the latest synced head root which in turn determines where a node "starts" its consensus participation - the era directory however can be freely shared between nodes / people without any (significant) security implications, assuming the era files are consistent / not broken. There's lots of future improvements to be had: * we can drop the in-memory `BlockRef` index almost entirely - at this point, resident memory usage of Nimbus should drop to a cool 500-600 mb * we could serve era files via REST trivially: this would drop backfill times to whatever time it takes to download the files - unlike the current implementation that downloads block by block, downloading an era at a time almost entirely cuts out request overhead * we can "reasonably" recreate detailed state history from almost any point in time, turning an O(slot) process into O(1) effectively - we'll still need caches and indices to do this with sufficient efficiency for the rest api, but at least it cuts the whole process down to minutes instead of hours, for arbitrary points in time * CI: ignore failures with Nim-1.6 (temporary) * test fixes Co-authored-by: Ștefan Talpalaru <stefantalpalaru@yahoo.com>
2022-03-23 08:58:17 +00:00
getFrontfillSlot, dag.backfill.slot, blockVerifier, maxHeadAge = 0)
router = (ref MessageRouter)(
processor: processor,
network: node.network)
if node.config.lightClientDataServe:
proc scheduleSendingLightClientUpdates(slot: Slot) =
if node.lightClientPool[].broadcastGossipFut != nil:
return
if slot <= node.lightClientPool[].latestBroadcastedSlot:
return
node.lightClientPool[].latestBroadcastedSlot = slot
template fut(): auto = node.lightClientPool[].broadcastGossipFut
fut = node.handleLightClientUpdates(slot)
fut.addCallback do (p: pointer) {.gcsafe.}:
fut = nil
router.onSyncCommitteeMessage = scheduleSendingLightClientUpdates
dag.setFinalizationCb makeOnFinalizationCb(node.eventBus, node.eth1Monitor)
dag.setBlockCb(onBlockAdded)
dag.setHeadCb(onHeadChanged)
dag.setReorgCb(onChainReorg)
node.dag = dag
node.quarantine = quarantine
node.attestationPool = attestationPool
node.syncCommitteeMsgPool = syncCommitteeMsgPool
node.lightClientPool = lightClientPool
node.exitPool = exitPool
node.processor = processor
node.blockProcessor = blockProcessor
node.consensusManager = consensusManager
node.requestManager = RequestManager.init(node.network, blockVerifier)
node.syncManager = syncManager
node.backfiller = backfiller
node.router = router
node.addValidators()
block:
# Add in-process validators to the list of "known" validators such that
# we start with a reasonable ENR
let wallSlot = node.beaconClock.now().slotOrZero()
for validator in node.attachedValidators[].validators.values():
if config.validatorMonitorAuto:
node.validatorMonitor[].addMonitor(validator.pubkey, validator.index)
if validator.index.isSome():
node.consensusManager[].actionTracker.knownValidators[
validator.index.get()] = wallSlot
let stabilitySubnets =
node.consensusManager[].actionTracker.stabilitySubnets(wallSlot)
# Here, we also set the correct ENR should we be in all subnets mode!
node.network.updateStabilitySubnetMetadata(stabilitySubnets)
node.network.initBeaconSync(dag, getBeaconTime)
node.updateValidatorMetrics()
2021-05-19 06:38:13 +00:00
const SlashingDbName = "slashing_protection"
# changing this requires physical file rename as well or history is lost.
proc init*(T: type BeaconNode,
2022-06-21 08:29:16 +00:00
rng: ref HmacDrbgContext,
config: BeaconNodeConf,
metadata: Eth2NetworkMetadata): BeaconNode
{.raises: [Defect, CatchableError].} =
var taskpool: TaskPoolPtr
template cfg: auto = metadata.cfg
template eth1Network: auto = metadata.eth1Network
try:
if config.numThreads < 0:
fatal "The number of threads --numThreads cannot be negative."
quit 1
elif config.numThreads == 0:
taskpool = TaskPoolPtr.new(numThreads = min(countProcessors(), 16))
else:
taskpool = TaskPoolPtr.new(numThreads = config.numThreads)
info "Threadpool started", numThreads = taskpool.numThreads
except Exception as exc:
raise newException(Defect, "Failure in taskpool initialization.")
let
eventBus = EventBus(
blocksQueue: newAsyncEventQueue[EventBeaconBlockObject](),
headQueue: newAsyncEventQueue[HeadChangeInfoObject](),
reorgQueue: newAsyncEventQueue[ReorgInfoObject](),
finUpdateQueue: newAsyncEventQueue[
RestVersioned[ForkedLightClientFinalityUpdate]](),
optUpdateQueue: newAsyncEventQueue[
RestVersioned[ForkedLightClientOptimisticUpdate]](),
attestQueue: newAsyncEventQueue[Attestation](),
contribQueue: newAsyncEventQueue[SignedContributionAndProof](),
exitQueue: newAsyncEventQueue[SignedVoluntaryExit](),
finalQueue: newAsyncEventQueue[FinalizationInfoObject]()
)
db = BeaconChainDB.new(config.databaseDir, cfg, inMemory = false)
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
if config.finalizedCheckpointBlock.isSome:
warn "--finalized-checkpoint-block has been deprecated, ignoring"
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
let checkpointState = if config.finalizedCheckpointState.isSome:
let checkpointStatePath = config.finalizedCheckpointState.get.string
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
let tmp = try:
newClone(readSszForkedHashedBeaconState(
cfg, readAllBytes(checkpointStatePath).tryGet()))
except SszError as err:
fatal "Checkpoint state deserialization failed",
err = formatMsg(err, checkpointStatePath)
quit 1
except CatchableError as err:
fatal "Failed to read checkpoint state file", err = err.msg
quit 1
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
if not getStateField(tmp[], slot).is_epoch:
fatal "--finalized-checkpoint-state must point to a state for an epoch slot",
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
slot = getStateField(tmp[], slot)
quit 1
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
tmp
else:
nil
let optJwtSecret = rng[].loadJwtSecret(config, allowCreate = false)
2022-03-31 14:43:05 +00:00
if config.web3Urls.len() == 0:
notice "Running without execution client - validator features disabled (see https://nimbus.guide/eth1.html)"
var eth1Monitor: Eth1Monitor
var genesisState =
if metadata.genesisData.len > 0:
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
try:
newClone readSszForkedHashedBeaconState(
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
cfg,
metadata.genesisData.toOpenArrayByte(0, metadata.genesisData.high))
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
except CatchableError as err:
raiseAssert "Invalid baked-in state: " & err.msg
else:
nil
if not ChainDAGRef.isInitialized(db).isOk():
if genesisState == nil and checkpointState == nil:
2023-01-12 00:49:43 +00:00
fatal "No database and no genesis snapshot found. Please supply a genesis.ssz " &
"with the network configuration"
quit 1
2019-10-25 14:53:31 +00:00
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
if not genesisState.isNil and not checkpointState.isNil:
if getStateField(genesisState[], genesis_validators_root) !=
getStateField(checkpointState[], genesis_validators_root):
fatal "Checkpoint state does not match genesis - check the --network parameter",
rootFromGenesis = getStateField(
genesisState[], genesis_validators_root),
rootFromCheckpoint = getStateField(
checkpointState[], genesis_validators_root)
quit 1
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
try:
# Always store genesis state if we have it - this allows reindexing and
# answering genesis queries
if not genesisState.isNil:
ChainDAGRef.preInit(db, genesisState[])
if not checkpointState.isNil:
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
if genesisState.isNil or
getStateField(checkpointState[], slot) != GENESIS_SLOT:
ChainDAGRef.preInit(db, checkpointState[])
doAssert ChainDAGRef.isInitialized(db).isOk(), "preInit should have initialized db"
except CatchableError as exc:
error "Failed to initialize database", err = exc.msg
quit 1
else:
if not checkpointState.isNil:
fatal "A database already exists, cannot start from given checkpoint",
dataDir = config.dataDir
quit 1
# Doesn't use std/random directly, but dependencies might
randomize(rng[].rand(high(int)))
let
validatorMonitor = newClone(ValidatorMonitor.init(
config.validatorMonitorAuto, config.validatorMonitorTotals))
for key in config.validatorMonitorPubkeys:
validatorMonitor[].addMonitor(key, Opt.none(ValidatorIndex))
let
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
networkGenesisValidatorsRoot =
if not genesisState.isNil:
Opt.some(getStateField(genesisState[], genesis_validators_root))
else:
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
Opt.none(Eth2Digest)
dag = loadChainDag(
config, cfg, db, eventBus,
validatorMonitor, networkGenesisValidatorsRoot,
config.deploymentPhase <= DeploymentPhase.Testnet)
genesisTime = getStateField(dag.headState, genesis_time)
beaconClock = BeaconClock.init(genesisTime)
2022-03-31 14:43:05 +00:00
getBeaconTime = beaconClock.getBeaconTimeFn()
if config.weakSubjectivityCheckpoint.isSome:
dag.checkWeakSubjectivityCheckpoint(
config.weakSubjectivityCheckpoint.get, beaconClock)
if eth1Monitor.isNil and config.web3Urls.len > 0:
eth1Monitor = Eth1Monitor.init(
Implement split preset/config support (#2710) * Implement split preset/config support This is the initial bulk refactor to introduce runtime config values in a number of places, somewhat replacing the existing mechanism of loading network metadata. It still needs more work, this is the initial refactor that introduces runtime configuration in some of the places that need it. The PR changes the way presets and constants work, to match the spec. In particular, a "preset" now refers to the compile-time configuration while a "cfg" or "RuntimeConfig" is the dynamic part. A single binary can support either mainnet or minimal, but not both. Support for other presets has been removed completely (can be readded, in case there's need). There's a number of outstanding tasks: * `SECONDS_PER_SLOT` still needs fixing * loading custom runtime configs needs redoing * checking constants against YAML file * yeerongpilly support `build/nimbus_beacon_node --network=yeerongpilly --discv5:no --log-level=DEBUG` * load fork epoch from config * fix fork digest sent in status * nicer error string for request failures * fix tools * one more * fixup * fixup * fixup * use "standard" network definition folder in local testnet Files are loaded from their standard locations, including genesis etc, to conform to the format used in the `eth2-networks` repo. * fix launch scripts, allow unknown config values * fix base config of rest test * cleanups * bundle mainnet config using common loader * fix spec links and names * only include supported preset in binary * drop yeerongpilly, add altair-devnet-0, support boot_enr.yaml
2021-07-12 13:01:38 +00:00
cfg,
metadata.depositContractBlock,
metadata.depositContractBlockHash,
db,
2022-03-31 14:43:05 +00:00
getBeaconTime,
config.web3Urls,
eth1Network,
config.web3ForcePolling,
optJwtSecret,
ttdReached = not dag.loadExecutionBlockRoot(dag.finalizedHead.blck).isZero)
if config.rpcEnabled.isSome:
warn "Nimbus's JSON-RPC server has been removed. This includes the --rpc, --rpc-port, and --rpc-address configuration options. https://nimbus.guide/rest-api.html shows how to enable and configure the REST Beacon API server which replaces it."
2021-03-17 18:46:45 +00:00
let restServer = if config.restEnabled:
RestServerRef.init(config.restAddress, config.restPort,
config.keymanagerAllowedOrigin,
validateBeaconApiQueries,
config)
else:
nil
let
netKeys = getPersistentNetKeys(rng[], config)
nickname = if config.nodeName == "auto": shortForm(netKeys)
else: config.nodeName
network = createEth2Node(
rng, config, netKeys, cfg, dag.forkDigests, getBeaconTime,
getStateField(dag.headState, genesis_validators_root))
case config.slashingDbKind
of SlashingDbKind.v2:
discard
of SlashingDbKind.v1:
error "Slashing DB v1 is no longer supported for writing"
quit 1
of SlashingDbKind.both:
warn "Slashing DB v1 deprecated, writing only v2"
info "Loading slashing protection database (v2)",
path = config.validatorsDir()
proc getValidatorAndIdx(pubkey: ValidatorPubKey): Opt[ValidatorAndIndex] =
withState(dag.headState):
getValidator(forkyState().data.validators.asSeq(), pubkey)
let
slashingProtectionDB =
SlashingProtectionDB.init(
getStateField(dag.headState, genesis_validators_root),
2021-05-19 06:38:13 +00:00
config.validatorsDir(), SlashingDbName)
validatorPool = newClone(ValidatorPool.init(
slashingProtectionDB, config.doppelgangerDetection))
keymanagerInitResult = initKeymanagerServer(config, restServer)
keymanagerHost = if keymanagerInitResult.server != nil:
newClone KeymanagerHost.init(
validatorPool,
rng,
keymanagerInitResult.token,
config.validatorsDir,
config.secretsDir,
config.defaultFeeRecipient,
getValidatorAndIdx,
getBeaconTime)
else: nil
stateTtlCache =
if config.restCacheSize > 0:
StateTtlCache.init(
cacheSize = config.restCacheSize,
cacheTtl = chronos.seconds(config.restCacheTtl))
else:
nil
let payloadBuilderRestClient =
if config.payloadBuilderEnable:
2022-10-13 11:38:33 +00:00
RestClientRef.new(config.payloadBuilderUrl).valueOr:
warn "Payload builder REST client setup failed",
payloadBuilderUrl = config.payloadBuilderUrl
nil
else:
nil
let node = BeaconNode(
nickname: nickname,
graffitiBytes: if config.graffiti.isSome: config.graffiti.get
else: defaultGraffitiBytes(),
network: network,
netKeys: netKeys,
db: db,
config: config,
attachedValidators: validatorPool,
eth1Monitor: eth1Monitor,
payloadBuilderRestClient: payloadBuilderRestClient,
2021-03-17 18:46:45 +00:00
restServer: restServer,
keymanagerHost: keymanagerHost,
keymanagerServer: keymanagerInitResult.server,
eventBus: eventBus,
gossipState: {},
blocksGossipState: {},
beaconClock: beaconClock,
validatorMonitor: validatorMonitor,
stateTtlCache: stateTtlCache,
nextExchangeTransitionConfTime:
# https://github.com/ethereum/execution-apis/blob/v1.0.0-beta.1/src/engine/specification.md#specification-3
# Consensus Layer client software **SHOULD** poll this endpoint every
# 60 seconds.
# Delay first call by that time to allow for EL syncing to begin; it can
# otherwise generate an EL warning by claiming a zero merge block.
Moment.now + chronos.seconds(60),
dynamicFeeRecipientsStore: newClone(DynamicFeeRecipientsStore.init()))
node.initLightClient(
rng, cfg, dag.forkDigests, getBeaconTime, dag.genesis_validators_root)
node.initFullNode(
rng, dag, taskpool, getBeaconTime)
2020-12-16 13:03:04 +00:00
node.updateLightClientFromDag()
node
func verifyFinalization(node: BeaconNode, slot: Slot) =
# Epoch must be >= 4 to check finalization
const SETTLING_TIME_OFFSET = 1'u64
let epoch = slot.epoch()
# Don't static-assert this -- if this isn't called, don't require it
doAssert SLOTS_PER_EPOCH > SETTLING_TIME_OFFSET
# Intentionally, loudly assert. Point is to fail visibly and unignorably
# during testing.
if epoch >= 4 and slot mod SLOTS_PER_EPOCH > SETTLING_TIME_OFFSET:
let finalizedEpoch =
node.dag.finalizedHead.slot.epoch()
# Finalization rule 234, that has the most lag slots among the cases, sets
# state.finalized_checkpoint = old_previous_justified_checkpoint.epoch + 3
# and then state.slot gets incremented, to increase the maximum offset, if
# finalization occurs every slot, to 4 slots vs scheduledSlot.
doAssert finalizedEpoch + 4 >= epoch
func subnetLog(v: BitArray): string =
$toSeq(v.oneIndices())
func forkDigests(node: BeaconNode): auto =
let forkDigestsArray: array[BeaconStateFork, auto] = [
node.dag.forkDigests.phase0,
node.dag.forkDigests.altair,
node.dag.forkDigests.bellatrix,
2022-12-07 16:47:23 +00:00
node.dag.forkDigests.capella,
node.dag.forkDigests.eip4844]
forkDigestsArray
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-rc.0/specs/phase0/validator.md#phase-0-attestation-subnet-stability
proc updateAttestationSubnetHandlers(node: BeaconNode, slot: Slot) =
if node.gossipState.card == 0:
# When disconnected, updateGossipState is responsible for all things
# subnets - in particular, it will remove subscriptions on the edge where
# we enter the disconnected state.
return
let
aggregateSubnets =
node.consensusManager[].actionTracker.aggregateSubnets(slot)
stabilitySubnets =
node.consensusManager[].actionTracker.stabilitySubnets(slot)
subnets = aggregateSubnets + stabilitySubnets
node.network.updateStabilitySubnetMetadata(stabilitySubnets)
# Now we know what we should be subscribed to - make it so
let
prevSubnets = node.consensusManager[].actionTracker.subscribedSubnets
unsubscribeSubnets = prevSubnets - subnets
subscribeSubnets = subnets - prevSubnets
# Remember what we subscribed to, so we can unsubscribe later
node.consensusManager[].actionTracker.subscribedSubnets = subnets
let forkDigests = node.forkDigests()
for gossipFork in node.gossipState:
let forkDigest = forkDigests[gossipFork]
node.network.unsubscribeAttestationSubnets(unsubscribeSubnets, forkDigest)
node.network.subscribeAttestationSubnets(subscribeSubnets, forkDigest)
debug "Attestation subnets",
slot, epoch = slot.epoch, gossipState = node.gossipState,
stabilitySubnets = subnetLog(stabilitySubnets),
aggregateSubnets = subnetLog(aggregateSubnets),
prevSubnets = subnetLog(prevSubnets),
subscribeSubnets = subnetLog(subscribeSubnets),
unsubscribeSubnets = subnetLog(unsubscribeSubnets),
gossipState = node.gossipState
proc updateBlocksGossipStatus*(
node: BeaconNode, slot: Slot, dagIsBehind: bool) =
template cfg(): auto = node.dag.cfg
let
isBehind =
if node.shouldSyncOptimistically(slot):
# If optimistic sync is active, always subscribe to blocks gossip
false
else:
# Use DAG status to determine whether to subscribe for blocks gossip
dagIsBehind
targetGossipState = getTargetGossipState(
2022-12-04 07:42:03 +00:00
slot.epoch, cfg.ALTAIR_FORK_EPOCH, cfg.BELLATRIX_FORK_EPOCH,
cfg.CAPELLA_FORK_EPOCH, cfg.EIP4844_FORK_EPOCH, isBehind)
template currentGossipState(): auto = node.blocksGossipState
if currentGossipState == targetGossipState:
return
if currentGossipState.card == 0 and targetGossipState.card > 0:
debug "Enabling blocks topic subscriptions",
wallSlot = slot, targetGossipState
elif currentGossipState.card > 0 and targetGossipState.card == 0:
debug "Disabling blocks topic subscriptions",
wallSlot = slot
else:
# Individual forks added / removed
discard
let
newGossipForks = targetGossipState - currentGossipState
oldGossipForks = currentGossipState - targetGossipState
2023-01-12 13:33:14 +00:00
discard $eip4844ImplementationMissing & ": for EIP4844, https://github.com/ethereum/consensus-specs/blob/v1.3.0-rc.0/specs/eip4844/p2p-interface.md#beacon_block notes use beacon_block_and_blobs_sidecar rather than beacon_block"
for gossipFork in oldGossipForks:
let forkDigest = node.dag.forkDigests[].atStateFork(gossipFork)
node.network.unsubscribe(getBeaconBlocksTopic(forkDigest))
for gossipFork in newGossipForks:
let forkDigest = node.dag.forkDigests[].atStateFork(gossipFork)
node.network.subscribe(
getBeaconBlocksTopic(forkDigest), blocksTopicParams,
enableTopicMetrics = true)
node.blocksGossipState = targetGossipState
proc addPhase0MessageHandlers(
node: BeaconNode, forkDigest: ForkDigest, slot: Slot) =
node.network.subscribe(getAttesterSlashingsTopic(forkDigest), basicParams)
node.network.subscribe(getProposerSlashingsTopic(forkDigest), basicParams)
node.network.subscribe(getVoluntaryExitsTopic(forkDigest), basicParams)
node.network.subscribe(
getAggregateAndProofsTopic(forkDigest), aggregateTopicParams,
enableTopicMetrics = true)
# updateAttestationSubnetHandlers subscribes attestation subnets
proc removePhase0MessageHandlers(node: BeaconNode, forkDigest: ForkDigest) =
node.network.unsubscribe(getVoluntaryExitsTopic(forkDigest))
node.network.unsubscribe(getProposerSlashingsTopic(forkDigest))
node.network.unsubscribe(getAttesterSlashingsTopic(forkDigest))
node.network.unsubscribe(getAggregateAndProofsTopic(forkDigest))
for subnet_id in SubnetId:
node.network.unsubscribe(getAttestationTopic(forkDigest, subnet_id))
node.consensusManager[].actionTracker.subscribedSubnets = default(AttnetBits)
func hasSyncPubKey(node: BeaconNode, epoch: Epoch): auto =
# Only used to determine which gossip topics to which to subscribe
if node.config.subscribeAllSubnets:
(func(pubkey: ValidatorPubKey): bool {.closure.} = true)
else:
(func(pubkey: ValidatorPubKey): bool =
node.consensusManager[].actionTracker.hasSyncDuty(pubkey, epoch) or
pubkey in node.attachedValidators[].validators)
func getCurrentSyncCommiteeSubnets(node: BeaconNode, epoch: Epoch): SyncnetBits =
let syncCommittee = withState(node.dag.headState):
when stateFork >= BeaconStateFork.Altair:
forkyState.data.current_sync_committee
else:
return static(default(SyncnetBits))
getSyncSubnets(node.hasSyncPubKey(epoch), syncCommittee)
func getNextSyncCommitteeSubnets(node: BeaconNode, epoch: Epoch): SyncnetBits =
let syncCommittee = withState(node.dag.headState):
when stateFork >= BeaconStateFork.Altair:
forkyState.data.next_sync_committee
else:
return static(default(SyncnetBits))
getSyncSubnets(
node.hasSyncPubKey((epoch.sync_committee_period + 1).start_slot().epoch),
syncCommittee)
func getSyncCommitteeSubnets(node: BeaconNode, epoch: Epoch): SyncnetBits =
let
subnets = node.getCurrentSyncCommiteeSubnets(epoch)
epochsToSyncPeriod = nearSyncCommitteePeriod(epoch)
# The end-slot tracker might call this when it's theoretically applicable,
# but more than SYNC_COMMITTEE_SUBNET_COUNT epochs from when the next sync
# committee period begins, in which case `epochsToNextSyncPeriod` is none.
if epochsToSyncPeriod.isNone or
node.dag.cfg.stateForkAtEpoch(epoch + epochsToSyncPeriod.get) <
BeaconStateFork.Altair:
return subnets
subnets + node.getNextSyncCommitteeSubnets(epoch)
2022-11-24 14:38:07 +00:00
proc addAltairMessageHandlers(
node: BeaconNode, forkDigest: ForkDigest, slot: Slot) =
node.addPhase0MessageHandlers(forkDigest, slot)
# If this comes online near sync committee period, it'll immediately get
# replaced as usual by trackSyncCommitteeTopics, which runs at slot end.
2022-11-24 14:38:07 +00:00
let syncnets = node.getSyncCommitteeSubnets(slot.epoch)
for subcommitteeIdx in SyncSubcommitteeIndex:
if syncnets[subcommitteeIdx]:
node.network.subscribe(
getSyncCommitteeTopic(forkDigest, subcommitteeIdx), basicParams)
node.network.subscribe(
getSyncCommitteeContributionAndProofTopic(forkDigest), basicParams)
node.network.updateSyncnetsMetadata(syncnets)
2022-11-24 14:38:07 +00:00
proc addCapellaMessageHandlers(
node: BeaconNode, forkDigest: ForkDigest, slot: Slot) =
node.addAltairMessageHandlers(forkDigest, slot)
node.network.subscribe(getBlsToExecutionChangeTopic(forkDigest), basicParams)
2023-01-05 21:35:07 +00:00
proc addEIP4844MessageHandlers(
node: BeaconNode, forkDigest: ForkDigest, slot: Slot) =
node.addCapellaMessageHandlers(forkDigest, slot)
node.network.subscribe(getBeaconBlockAndBlobsSidecarTopic(forkDigest), basicParams)
proc removeAltairMessageHandlers(node: BeaconNode, forkDigest: ForkDigest) =
node.removePhase0MessageHandlers(forkDigest)
for subcommitteeIdx in SyncSubcommitteeIndex:
closureScope:
let idx = subcommitteeIdx
node.network.unsubscribe(getSyncCommitteeTopic(forkDigest, idx))
node.network.unsubscribe(
getSyncCommitteeContributionAndProofTopic(forkDigest))
2022-11-24 14:38:07 +00:00
proc removeCapellaMessageHandlers(node: BeaconNode, forkDigest: ForkDigest) =
node.removeAltairMessageHandlers(forkDigest)
node.network.unsubscribe(getBlsToExecutionChangeTopic(forkDigest))
2023-01-05 21:35:07 +00:00
proc removeEIP4844MessageHandlers(node: BeaconNode, forkDigest: ForkDigest) =
node.removeCapellaMessageHandlers(forkDigest)
node.network.unsubscribe(getBeaconBlockAndBlobsSidecarTopic(forkDigest))
proc updateSyncCommitteeTopics(node: BeaconNode, slot: Slot) =
template lastSyncUpdate: untyped =
node.consensusManager[].actionTracker.lastSyncUpdate
if lastSyncUpdate == Opt.some(slot.sync_committee_period()) and
nearSyncCommitteePeriod(slot.epoch).isNone():
# No need to update unless we're close to the next sync committee period or
# new validators were registered with the action tracker
# TODO we _could_ skip running this in some of the "near" slots, but..
return
lastSyncUpdate = Opt.some(slot.sync_committee_period())
let syncnets = node.getSyncCommitteeSubnets(slot.epoch)
debug "Updating sync committee subnets",
syncnets,
metadata_syncnets = node.network.metadata.syncnets,
gossipState = node.gossipState
# Assume that different gossip fork sync committee setups are in sync; this
# only remains relevant, currently, for one gossip transition epoch, so the
# consequences of this not being true aren't exceptionally dire, while this
# allows for bookkeeping simplication.
if syncnets == node.network.metadata.syncnets:
return
let
newSyncnets =
syncnets - node.network.metadata.syncnets
oldSyncnets =
node.network.metadata.syncnets - syncnets
forkDigests = node.forkDigests()
for subcommitteeIdx in SyncSubcommitteeIndex:
doAssert not (newSyncnets[subcommitteeIdx] and
oldSyncnets[subcommitteeIdx])
for gossipFork in node.gossipState:
template topic(): auto =
getSyncCommitteeTopic(forkDigests[gossipFork], subcommitteeIdx)
if oldSyncnets[subcommitteeIdx]:
node.network.unsubscribe(topic)
elif newSyncnets[subcommitteeIdx]:
node.network.subscribe(topic, basicParams)
node.network.updateSyncnetsMetadata(syncnets)
proc updateDoppelganger(node: BeaconNode, epoch: Epoch) =
if not node.processor[].doppelgangerDetectionEnabled:
return
# broadcastStartEpoch is set to FAR_FUTURE_EPOCH when we're not monitoring
# gossip - it is only viable to assert liveness in epochs where gossip is
# active
if epoch > node.processor[].doppelgangerDetection.broadcastStartEpoch:
for validator in node.attachedValidators[]:
validator.updateDoppelganger(epoch - 1)
proc updateGossipStatus(node: BeaconNode, slot: Slot) {.async.} =
## Subscribe to subnets that we are providing stability for or aggregating
## and unsubscribe from the ones that are no longer relevant.
# Let the tracker know what duties are approaching - this will tell us how
# many stability subnets we need to be subscribed to and what subnets we'll
# soon be aggregating - in addition to the in-beacon-node duties, there may
# also be duties coming from the validator client, but we don't control when
# these arrive
await node.registerDuties(slot)
# We start subscribing to gossip before we're fully synced - this allows time
# to subscribe before the sync end game
const
TOPIC_SUBSCRIBE_THRESHOLD_SLOTS = 64
HYSTERESIS_BUFFER = 16
let
head = node.dag.head
headDistance =
if slot > head.slot: (slot - head.slot).uint64
else: 0'u64
isBehind =
headDistance > TOPIC_SUBSCRIBE_THRESHOLD_SLOTS + HYSTERESIS_BUFFER
targetGossipState =
getTargetGossipState(
slot.epoch,
node.dag.cfg.ALTAIR_FORK_EPOCH,
node.dag.cfg.BELLATRIX_FORK_EPOCH,
2022-12-04 07:42:03 +00:00
node.dag.cfg.CAPELLA_FORK_EPOCH,
node.dag.cfg.EIP4844_FORK_EPOCH,
isBehind)
doAssert targetGossipState.card <= 2
let
newGossipForks = targetGossipState - node.gossipState
oldGossipForks = node.gossipState - targetGossipState
doAssert newGossipForks.card <= 2
doAssert oldGossipForks.card <= 2
func maxGossipFork(gossipState: GossipState): int =
var res = -1
for gossipFork in gossipState:
res = max(res, gossipFork.int)
res
if maxGossipFork(targetGossipState) < maxGossipFork(node.gossipState) and
targetGossipState != {}:
warn "Unexpected clock regression during transition",
targetGossipState,
gossipState = node.gossipState
if node.gossipState.card == 0 and targetGossipState.card > 0:
# We are synced, so we will connect
debug "Enabling topic subscriptions",
wallSlot = slot,
headSlot = head.slot,
headDistance, targetGossipState
node.processor[].setupDoppelgangerDetection(slot)
# Specially when waiting for genesis, we'll already be synced on startup -
# it might also happen on a sufficiently fast restart
# We "know" the actions for the current and the next epoch
withState(node.dag.headState):
if node.consensusManager[].actionTracker.needsUpdate(
forkyState, slot.epoch):
let epochRef = node.dag.getEpochRef(head, slot.epoch, false).expect(
"Getting head EpochRef should never fail")
node.consensusManager[].actionTracker.updateActions(epochRef)
if node.consensusManager[].actionTracker.needsUpdate(
forkyState, slot.epoch + 1):
let epochRef = node.dag.getEpochRef(head, slot.epoch + 1, false).expect(
"Getting head EpochRef should never fail")
node.consensusManager[].actionTracker.updateActions(epochRef)
if node.gossipState.card > 0 and targetGossipState.card == 0:
debug "Disabling topic subscriptions",
wallSlot = slot,
headSlot = head.slot,
headDistance
node.processor[].clearDoppelgangerProtection()
let forkDigests = node.forkDigests()
const removeMessageHandlers: array[BeaconStateFork, auto] = [
removePhase0MessageHandlers,
removeAltairMessageHandlers,
removeAltairMessageHandlers, # bellatrix (altair handlers, different forkDigest)
2022-12-07 16:47:23 +00:00
removeCapellaMessageHandlers,
2023-01-05 21:35:07 +00:00
removeEIP4844MessageHandlers
]
for gossipFork in oldGossipForks:
removeMessageHandlers[gossipFork](node, forkDigests[gossipFork])
const addMessageHandlers: array[BeaconStateFork, auto] = [
addPhase0MessageHandlers,
addAltairMessageHandlers,
2023-01-05 21:35:07 +00:00
addAltairMessageHandlers, # bellatrix (altair handlers, different forkDigest)
2022-12-07 16:47:23 +00:00
addCapellaMessageHandlers,
2023-01-05 21:35:07 +00:00
addEIP4844MessageHandlers
]
for gossipFork in newGossipForks:
addMessageHandlers[gossipFork](node, forkDigests[gossipFork], slot)
node.gossipState = targetGossipState
node.updateDoppelganger(slot.epoch)
node.updateAttestationSubnetHandlers(slot)
node.updateBlocksGossipStatus(slot, isBehind)
node.updateLightClientGossipStatus(slot, isBehind)
proc onSlotEnd(node: BeaconNode, slot: Slot) {.async.} =
# Things we do when slot processing has ended and we're about to wait for the
# next slot
if node.dag.needStateCachesAndForkChoicePruning():
if node.attachedValidators[].validators.len > 0:
node.attachedValidators[]
.slashingProtection
# pruning is only done if the DB is set to pruning mode.
.pruneAfterFinalization(
node.dag.finalizedHead.slot.epoch()
)
# Delay part of pruning until latency critical duties are done.
# The other part of pruning, `pruneBlocksDAG`, is done eagerly.
# ----
# This is the last pruning to do as it clears the "needPruning" condition.
node.consensusManager[].pruneStateCachesAndForkChoice()
if node.config.historyMode == HistoryMode.Prune:
node.dag.pruneHistory()
when declared(GC_fullCollect):
# The slots in the beacon node work as frames in a game: we want to make
# sure that we're ready for the next one and don't get stuck in lengthy
# garbage collection tasks when time is of essence in the middle of a slot -
# while this does not guarantee that we'll never collect during a slot, it
# makes sure that all the scratch space we used during slot tasks (logging,
# temporary buffers etc) gets recycled for the next slot that is likely to
# need similar amounts of memory.
try:
GC_fullCollect()
except Defect as exc:
raise exc # Reraise to maintain call stack
except Exception as exc:
# TODO upstream
raiseAssert "Unexpected exception during GC collection"
# Checkpoint the database to clear the WAL file and make sure changes in
# the database are synced with the filesystem.
node.db.checkpoint()
node.syncCommitteeMsgPool[].pruneData(slot)
if slot.is_epoch:
node.dynamicFeeRecipientsStore[].pruneOldMappings(slot.epoch)
# Update upcoming actions - we do this every slot in case a reorg happens
let head = node.dag.head
if node.isSynced(head) == SyncStatus.synced:
withState(node.dag.headState):
if node.consensusManager[].actionTracker.needsUpdate(
forkyState, slot.epoch + 1):
let epochRef = node.dag.getEpochRef(head, slot.epoch + 1, false).expect(
"Getting head EpochRef should never fail")
node.consensusManager[].actionTracker.updateActions(epochRef)
let
nextAttestationSlot =
node.consensusManager[].actionTracker.getNextAttestationSlot(slot)
nextProposalSlot =
node.consensusManager[].actionTracker.getNextProposalSlot(slot)
nextActionWaitTime = saturate(fromNow(
node.beaconClock, min(nextAttestationSlot, nextProposalSlot)))
# -1 is a more useful output than 18446744073709551615 as an indicator of
# no future attestation/proposal known.
template formatInt64(x: Slot): int64 =
if x == high(uint64).Slot:
-1'i64
else:
toGaugeValue(x)
template formatSyncCommitteeStatus(): string =
let slotsToNextSyncCommitteePeriod =
SLOTS_PER_SYNC_COMMITTEE_PERIOD - since_sync_committee_period_start(slot)
# int64 conversion is safe
doAssert slotsToNextSyncCommitteePeriod <= SLOTS_PER_SYNC_COMMITTEE_PERIOD
if not node.getCurrentSyncCommiteeSubnets(slot.epoch).isZeros:
"current"
elif not node.getNextSyncCommitteeSubnets(slot.epoch).isZeros:
"in " & toTimeLeftString(
SECONDS_PER_SLOT.int64.seconds * slotsToNextSyncCommitteePeriod.int64)
else:
"none"
info "Slot end",
slot = shortLog(slot),
nextActionWait =
if nextAttestationSlot == FAR_FUTURE_SLOT:
"n/a"
else:
Logging and startup improvements (#3038) * Logging and startup improvements Color support for released binaries! * startup scripts no longer log to file by default - this only affects source builds - released binaries don't support file logging * add --log-stdout option to control logging to stdout (colors, json) * detect tty:s vs redirected logs and log accordingly * add option to disable log colors at runtime * simplify several "common" logs, showing the most important information earlier and more clearly * remove line numbers / file information / tid - these take up space and are of little use to end users * still enabled in debug builds and tools * remove `testnet_servers_image` compile-time option * server images, released binaries and compile-from-source now offer the same behaviour and features * fixes https://github.com/status-im/nimbus-eth2/issues/2326 * fixes https://github.com/status-im/nimbus-eth2/issues/1794 * remove instanteneous block speed from sync message, keeping only average before: ``` INF 2021-10-28 16:45:59.000+02:00 Slot start topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:884 lastSlot=2384027 wallSlot=2384028 delay=461us84ns peers=0 head=75a10ee5:3348 headEpoch=104 finalized=cd6804ba:3264 finalizedEpoch=102 sync="wwwwwwwwww:0:0.0000:0.0000:00h00m (3348)" INF 2021-10-28 16:45:59.046+02:00 Slot end topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:821 slot=2384028 nextSlot=2384029 head=75a10ee5:3348 headEpoch=104 finalizedHead=cd6804ba:3264 finalizedEpoch=102 nextAttestationSlot=-1 nextProposalSlot=-1 nextActionWait=n/a ``` after: ``` INF 2021-10-28 22:43:23.033+02:00 Slot start topics="beacnde" slot=2385815 epoch=74556 sync="DDPDDPUDDD:10:5.2258:01h19m (2361088)" peers=37 head=eacd2dae:2361096 finalized=73782:a4751487 delay=33ms687us715ns INF 2021-10-28 22:43:23.291+02:00 Slot end topics="beacnde" slot=2385815 nextActionWait=n/a nextAttestationSlot=-1 nextProposalSlot=-1 head=eacd2dae:2361096 ``` * fix comment * documentation updates * mention `--log-file` may be deprecated in the future * update various docs
2021-11-02 17:06:36 +00:00
shortLog(nextActionWaitTime),
nextAttestationSlot = formatInt64(nextAttestationSlot),
nextProposalSlot = formatInt64(nextProposalSlot),
syncCommitteeDuties = formatSyncCommitteeStatus(),
head = shortLog(head)
if nextAttestationSlot != FAR_FUTURE_SLOT:
next_action_wait.set(nextActionWaitTime.toFloatSeconds)
let epoch = slot.epoch
2021-09-29 11:06:16 +00:00
if epoch + 1 >= node.network.forkId.next_fork_epoch:
# Update 1 epoch early to block non-fork-ready peers
node.network.updateForkId(epoch, node.dag.genesis_validators_root)
# When we're not behind schedule, we'll speculatively update the clearance
# state in anticipation of receiving the next block - we do it after logging
# slot end since the nextActionWaitTime can be short
let
advanceCutoff = node.beaconClock.fromNow(
slot.start_beacon_time() + chronos.seconds(int(SECONDS_PER_SLOT - 1)))
if advanceCutoff.inFuture:
# We wait until there's only a second left before the next slot begins, then
# we advance the clearance state to the next slot - this gives us a high
# probability of being prepared for the block that will arrive and the
# epoch processing that follows
await sleepAsync(advanceCutoff.offset)
node.dag.advanceClearanceState()
# Prepare action tracker for the next slot
node.consensusManager[].actionTracker.updateSlot(slot + 1)
# The last thing we do is to perform the subscriptions and unsubscriptions for
# the next slot, just before that slot starts - because of the advance cuttoff
# above, this will be done just before the next slot starts
node.updateSyncCommitteeTopics(slot + 1)
await node.updateGossipStatus(slot + 1)
func syncStatus(node: BeaconNode, wallSlot: Slot): string =
let optimistic_head = node.dag.is_optimistic(node.dag.head.root)
if node.syncManager.inProgress:
let
optimisticSuffix =
if optimistic_head:
"/opt"
else:
""
lightClientSuffix =
if node.consensusManager[].shouldSyncOptimistically(wallSlot):
" - lc: " & $shortLog(node.consensusManager[].optimisticHead)
else:
""
node.syncManager.syncStatus & optimisticSuffix & lightClientSuffix
elif node.backfiller.inProgress:
"backfill: " & node.backfiller.syncStatus
elif optimistic_head:
"synced/opt"
else:
"synced"
proc onSlotStart(node: BeaconNode, wallTime: BeaconTime,
lastSlot: Slot): Future[bool] {.async.} =
## Called at the beginning of a slot - usually every slot, but sometimes might
## skip a few in case we're running late.
## wallTime: current system time - we will strive to perform all duties up
## to this point in time
2020-06-26 13:51:20 +00:00
## lastSlot: the last slot that we successfully processed, so we know where to
## start work from - there might be jumps if processing is delayed
let
# The slot we should be at, according to the clock
wallSlot = wallTime.slotOrZero
# If everything was working perfectly, the slot that we should be processing
expectedSlot = lastSlot + 1
finalizedEpoch = node.dag.finalizedHead.blck.slot.epoch()
delay = wallTime - expectedSlot.start_beacon_time()
info "Slot start",
Logging and startup improvements (#3038) * Logging and startup improvements Color support for released binaries! * startup scripts no longer log to file by default - this only affects source builds - released binaries don't support file logging * add --log-stdout option to control logging to stdout (colors, json) * detect tty:s vs redirected logs and log accordingly * add option to disable log colors at runtime * simplify several "common" logs, showing the most important information earlier and more clearly * remove line numbers / file information / tid - these take up space and are of little use to end users * still enabled in debug builds and tools * remove `testnet_servers_image` compile-time option * server images, released binaries and compile-from-source now offer the same behaviour and features * fixes https://github.com/status-im/nimbus-eth2/issues/2326 * fixes https://github.com/status-im/nimbus-eth2/issues/1794 * remove instanteneous block speed from sync message, keeping only average before: ``` INF 2021-10-28 16:45:59.000+02:00 Slot start topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:884 lastSlot=2384027 wallSlot=2384028 delay=461us84ns peers=0 head=75a10ee5:3348 headEpoch=104 finalized=cd6804ba:3264 finalizedEpoch=102 sync="wwwwwwwwww:0:0.0000:0.0000:00h00m (3348)" INF 2021-10-28 16:45:59.046+02:00 Slot end topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:821 slot=2384028 nextSlot=2384029 head=75a10ee5:3348 headEpoch=104 finalizedHead=cd6804ba:3264 finalizedEpoch=102 nextAttestationSlot=-1 nextProposalSlot=-1 nextActionWait=n/a ``` after: ``` INF 2021-10-28 22:43:23.033+02:00 Slot start topics="beacnde" slot=2385815 epoch=74556 sync="DDPDDPUDDD:10:5.2258:01h19m (2361088)" peers=37 head=eacd2dae:2361096 finalized=73782:a4751487 delay=33ms687us715ns INF 2021-10-28 22:43:23.291+02:00 Slot end topics="beacnde" slot=2385815 nextActionWait=n/a nextAttestationSlot=-1 nextProposalSlot=-1 head=eacd2dae:2361096 ``` * fix comment * documentation updates * mention `--log-file` may be deprecated in the future * update various docs
2021-11-02 17:06:36 +00:00
slot = shortLog(wallSlot),
epoch = shortLog(wallSlot.epoch),
sync = node.syncStatus(wallSlot),
Logging and startup improvements (#3038) * Logging and startup improvements Color support for released binaries! * startup scripts no longer log to file by default - this only affects source builds - released binaries don't support file logging * add --log-stdout option to control logging to stdout (colors, json) * detect tty:s vs redirected logs and log accordingly * add option to disable log colors at runtime * simplify several "common" logs, showing the most important information earlier and more clearly * remove line numbers / file information / tid - these take up space and are of little use to end users * still enabled in debug builds and tools * remove `testnet_servers_image` compile-time option * server images, released binaries and compile-from-source now offer the same behaviour and features * fixes https://github.com/status-im/nimbus-eth2/issues/2326 * fixes https://github.com/status-im/nimbus-eth2/issues/1794 * remove instanteneous block speed from sync message, keeping only average before: ``` INF 2021-10-28 16:45:59.000+02:00 Slot start topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:884 lastSlot=2384027 wallSlot=2384028 delay=461us84ns peers=0 head=75a10ee5:3348 headEpoch=104 finalized=cd6804ba:3264 finalizedEpoch=102 sync="wwwwwwwwww:0:0.0000:0.0000:00h00m (3348)" INF 2021-10-28 16:45:59.046+02:00 Slot end topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:821 slot=2384028 nextSlot=2384029 head=75a10ee5:3348 headEpoch=104 finalizedHead=cd6804ba:3264 finalizedEpoch=102 nextAttestationSlot=-1 nextProposalSlot=-1 nextActionWait=n/a ``` after: ``` INF 2021-10-28 22:43:23.033+02:00 Slot start topics="beacnde" slot=2385815 epoch=74556 sync="DDPDDPUDDD:10:5.2258:01h19m (2361088)" peers=37 head=eacd2dae:2361096 finalized=73782:a4751487 delay=33ms687us715ns INF 2021-10-28 22:43:23.291+02:00 Slot end topics="beacnde" slot=2385815 nextActionWait=n/a nextAttestationSlot=-1 nextProposalSlot=-1 head=eacd2dae:2361096 ``` * fix comment * documentation updates * mention `--log-file` may be deprecated in the future * update various docs
2021-11-02 17:06:36 +00:00
peers = len(node.network.peerPool),
head = shortLog(node.dag.head),
finalized = shortLog(getStateField(
node.dag.headState, finalized_checkpoint)),
Logging and startup improvements (#3038) * Logging and startup improvements Color support for released binaries! * startup scripts no longer log to file by default - this only affects source builds - released binaries don't support file logging * add --log-stdout option to control logging to stdout (colors, json) * detect tty:s vs redirected logs and log accordingly * add option to disable log colors at runtime * simplify several "common" logs, showing the most important information earlier and more clearly * remove line numbers / file information / tid - these take up space and are of little use to end users * still enabled in debug builds and tools * remove `testnet_servers_image` compile-time option * server images, released binaries and compile-from-source now offer the same behaviour and features * fixes https://github.com/status-im/nimbus-eth2/issues/2326 * fixes https://github.com/status-im/nimbus-eth2/issues/1794 * remove instanteneous block speed from sync message, keeping only average before: ``` INF 2021-10-28 16:45:59.000+02:00 Slot start topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:884 lastSlot=2384027 wallSlot=2384028 delay=461us84ns peers=0 head=75a10ee5:3348 headEpoch=104 finalized=cd6804ba:3264 finalizedEpoch=102 sync="wwwwwwwwww:0:0.0000:0.0000:00h00m (3348)" INF 2021-10-28 16:45:59.046+02:00 Slot end topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:821 slot=2384028 nextSlot=2384029 head=75a10ee5:3348 headEpoch=104 finalizedHead=cd6804ba:3264 finalizedEpoch=102 nextAttestationSlot=-1 nextProposalSlot=-1 nextActionWait=n/a ``` after: ``` INF 2021-10-28 22:43:23.033+02:00 Slot start topics="beacnde" slot=2385815 epoch=74556 sync="DDPDDPUDDD:10:5.2258:01h19m (2361088)" peers=37 head=eacd2dae:2361096 finalized=73782:a4751487 delay=33ms687us715ns INF 2021-10-28 22:43:23.291+02:00 Slot end topics="beacnde" slot=2385815 nextActionWait=n/a nextAttestationSlot=-1 nextProposalSlot=-1 head=eacd2dae:2361096 ``` * fix comment * documentation updates * mention `--log-file` may be deprecated in the future * update various docs
2021-11-02 17:06:36 +00:00
delay = shortLog(delay)
# Check before any re-scheduling of onSlotStart()
if checkIfShouldStopAtEpoch(wallSlot, node.config.stopAtEpoch):
quit(0)
when defined(windows):
if node.config.runAsService:
reportServiceStatus(SERVICE_RUNNING, NO_ERROR, 0)
beacon_slot.set wallSlot.toGaugeValue
beacon_current_epoch.set wallSlot.epoch.toGaugeValue
# both non-negative, so difference can't overflow or underflow int64
finalization_delay.set(
wallSlot.epoch.toGaugeValue - finalizedEpoch.toGaugeValue)
if node.config.strictVerification:
verifyFinalization(node, wallSlot)
node.consensusManager[].updateHead(wallSlot)
await node.handleValidatorDuties(lastSlot, wallSlot)
2019-08-16 11:16:56 +00:00
await onSlotEnd(node, wallSlot)
# https://github.com/ethereum/builder-specs/blob/v0.2.0/specs/validator.md#registration-dissemination
# This specification suggests validators re-submit to builder software every
# `EPOCHS_PER_VALIDATOR_REGISTRATION_SUBMISSION` epochs.
if wallSlot.is_epoch and
wallSlot.epoch mod EPOCHS_PER_VALIDATOR_REGISTRATION_SUBMISSION == 0:
asyncSpawn node.registerValidators(wallSlot.epoch)
return false
proc handleMissingBlocks(node: BeaconNode) =
let missingBlocks = node.quarantine[].checkMissing()
if missingBlocks.len > 0:
debug "Requesting detected missing blocks", blocks = shortLog(missingBlocks)
node.requestManager.fetchAncestorBlocks(missingBlocks)
proc onSecond(node: BeaconNode, time: Moment) =
## This procedure will be called once per second.
if not(node.syncManager.inProgress):
node.handleMissingBlocks()
# Nim GC metrics (for the main thread)
updateThreadMetrics()
if time >= node.nextExchangeTransitionConfTime and not node.eth1Monitor.isNil:
# The EL client SHOULD log a warning when not receiving an exchange message
# at least once every 120 seconds. If we only attempt to exchange every 60
# seconds, the warning would be triggered if a single message is missed.
# To accommodate for that, exchange slightly more frequently.
# https://github.com/ethereum/execution-apis/blob/v1.0.0-beta.1/src/engine/specification.md#engine_exchangetransitionconfigurationv1
node.nextExchangeTransitionConfTime = time + chronos.seconds(45)
if node.currentSlot.epoch >= node.dag.cfg.BELLATRIX_FORK_EPOCH:
traceAsyncErrors node.eth1Monitor.exchangeTransitionConfiguration()
if node.config.stopAtSyncedEpoch != 0 and
node.dag.head.slot.epoch >= node.config.stopAtSyncedEpoch:
notice "Shutting down after having reached the target synced epoch"
bnStatus = BeaconNodeStatus.Stopping
proc runOnSecondLoop(node: BeaconNode) {.async.} =
const
sleepTime = chronos.seconds(1)
nanosecondsIn1s = float(sleepTime.nanoseconds)
while true:
let start = chronos.now(chronos.Moment)
await chronos.sleepAsync(sleepTime)
let afterSleep = chronos.now(chronos.Moment)
let sleepTime = afterSleep - start
node.onSecond(start)
let finished = chronos.now(chronos.Moment)
let processingTime = finished - afterSleep
ticks_delay.set(sleepTime.nanoseconds.float / nanosecondsIn1s)
trace "onSecond task completed", sleepTime, processingTime
func connectedPeersCount(node: BeaconNode): int =
len(node.network.peerPool)
2021-03-17 18:46:45 +00:00
proc installRestHandlers(restServer: RestServerRef, node: BeaconNode) =
restServer.router.installBeaconApiHandlers(node)
restServer.router.installConfigApiHandlers(node)
restServer.router.installDebugApiHandlers(node)
restServer.router.installEventApiHandlers(node)
restServer.router.installNimbusApiHandlers(node)
2021-03-17 18:46:45 +00:00
restServer.router.installNodeApiHandlers(node)
restServer.router.installValidatorApiHandlers(node)
if node.dag.lcDataStore.serve:
restServer.router.installLightClientApiHandlers(node)
2021-03-17 18:46:45 +00:00
2022-12-04 07:42:03 +00:00
from ./spec/datatypes/capella import SignedBeaconBlock
proc installMessageValidators(node: BeaconNode) =
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-rc.0/specs/phase0/p2p-interface.md#attestations-and-aggregation
# These validators stay around the whole time, regardless of which specific
# subnets are subscribed to during any given epoch.
let forkDigests = node.dag.forkDigests
node.network.addValidator(
getBeaconBlocksTopic(forkDigests.phase0),
proc (signedBlock: phase0.SignedBeaconBlock): ValidationResult =
if node.shouldSyncOptimistically(node.currentSlot):
toValidationResult(
node.optimisticProcessor.processSignedBeaconBlock(signedBlock))
else:
toValidationResult(node.processor[].processSignedBeaconBlock(
MsgSource.gossip, signedBlock)))
template installPhase0Validators(digest: auto) =
for it in SubnetId:
closureScope:
let subnet_id = it
node.network.addAsyncValidator(
getAttestationTopic(digest, subnet_id),
# This proc needs to be within closureScope; don't lift out of loop.
proc(attestation: Attestation): Future[ValidationResult] {.async.} =
return toValidationResult(
await node.processor.processAttestation(
MsgSource.gossip, attestation, subnet_id)))
node.network.addAsyncValidator(
getAggregateAndProofsTopic(digest),
proc(signedAggregateAndProof: SignedAggregateAndProof):
Future[ValidationResult] {.async.} =
return toValidationResult(
await node.processor.processSignedAggregateAndProof(
MsgSource.gossip, signedAggregateAndProof, false)))
node.network.addValidator(
getAttesterSlashingsTopic(digest),
proc (attesterSlashing: AttesterSlashing): ValidationResult =
toValidationResult(
node.processor[].processAttesterSlashing(
MsgSource.gossip, attesterSlashing)))
node.network.addValidator(
getProposerSlashingsTopic(digest),
proc (proposerSlashing: ProposerSlashing): ValidationResult =
toValidationResult(
node.processor[].processProposerSlashing(
MsgSource.gossip, proposerSlashing)))
node.network.addValidator(
getVoluntaryExitsTopic(digest),
proc (signedVoluntaryExit: SignedVoluntaryExit): ValidationResult =
toValidationResult(
node.processor[].processSignedVoluntaryExit(
MsgSource.gossip, signedVoluntaryExit)))
installPhase0Validators(forkDigests.phase0)
2022-12-04 07:42:03 +00:00
# Validators introduced in phase0 are also used in Altair and Bellatrix, but
# with different fork digests
installPhase0Validators(forkDigests.altair)
installPhase0Validators(forkDigests.bellatrix)
2022-12-04 07:42:03 +00:00
installPhase0Validators(forkDigests.capella)
if node.dag.cfg.EIP4844_FORK_EPOCH != FAR_FUTURE_EPOCH:
installPhase0Validators(forkDigests.eip4844)
node.network.addValidator(
getBeaconBlocksTopic(forkDigests.altair),
proc (signedBlock: altair.SignedBeaconBlock): ValidationResult =
if node.shouldSyncOptimistically(node.currentSlot):
toValidationResult(
node.optimisticProcessor.processSignedBeaconBlock(signedBlock))
else:
toValidationResult(node.processor[].processSignedBeaconBlock(
MsgSource.gossip, signedBlock)))
node.network.addValidator(
getBeaconBlocksTopic(forkDigests.bellatrix),
proc (signedBlock: bellatrix.SignedBeaconBlock): ValidationResult =
if node.shouldSyncOptimistically(node.currentSlot):
toValidationResult(
node.optimisticProcessor.processSignedBeaconBlock(signedBlock))
else:
toValidationResult(node.processor[].processSignedBeaconBlock(
MsgSource.gossip, signedBlock)))
2022-12-04 07:42:03 +00:00
node.network.addValidator(
getBeaconBlocksTopic(forkDigests.capella),
proc (signedBlock: capella.SignedBeaconBlock): ValidationResult =
if node.shouldSyncOptimistically(node.currentSlot):
toValidationResult(
node.optimisticProcessor.processSignedBeaconBlock(signedBlock))
else:
toValidationResult(node.processor[].processSignedBeaconBlock(
MsgSource.gossip, signedBlock)))
if node.dag.cfg.EIP4844_FORK_EPOCH != FAR_FUTURE_EPOCH:
node.network.addValidator(
getBeaconBlockAndBlobsSidecarTopic(forkDigests.eip4844),
proc (signedBlock: eip4844.SignedBeaconBlockAndBlobsSidecar): ValidationResult =
# TODO: take into account node.shouldSyncOptimistically(node.currentSlot)
toValidationResult(node.processor[].processSignedBeaconBlockAndBlobsSidecar(
MsgSource.gossip, signedBlock)))
template installSyncCommitteeeValidators(digest: auto) =
for subcommitteeIdx in SyncSubcommitteeIndex:
closureScope:
let idx = subcommitteeIdx
node.network.addAsyncValidator(
getSyncCommitteeTopic(digest, idx),
# This proc needs to be within closureScope; don't lift out of loop.
proc(msg: SyncCommitteeMessage): Future[ValidationResult] {.async.} =
return toValidationResult(
await node.processor.processSyncCommitteeMessage(
MsgSource.gossip, msg, idx)))
node.network.addAsyncValidator(
getSyncCommitteeContributionAndProofTopic(digest),
proc(msg: SignedContributionAndProof): Future[ValidationResult] {.async.} =
return toValidationResult(
await node.processor.processSignedContributionAndProof(
MsgSource.gossip, msg)))
installSyncCommitteeeValidators(forkDigests.altair)
installSyncCommitteeeValidators(forkDigests.bellatrix)
2022-12-04 07:42:03 +00:00
installSyncCommitteeeValidators(forkDigests.capella)
if node.dag.cfg.EIP4844_FORK_EPOCH != FAR_FUTURE_EPOCH:
installSyncCommitteeeValidators(forkDigests.eip4844)
node.installLightClientMessageValidators()
proc stop(node: BeaconNode) =
bnStatus = BeaconNodeStatus.Stopping
notice "Graceful shutdown"
if not node.config.inProcessValidators:
try:
node.vcProcess.close()
except Exception as exc:
warn "Couldn't close vc process", msg = exc.msg
try:
waitFor node.network.stop()
except CatchableError as exc:
warn "Couldn't stop network", msg = exc.msg
node.attachedValidators[].slashingProtection.close()
2022-08-07 21:53:20 +00:00
node.attachedValidators[].close()
node.db.close()
notice "Databases closed"
proc startBackfillTask(node: BeaconNode) {.async.} =
while node.dag.needsBackfill:
if not node.syncManager.inProgress:
# Only start the backfiller if it's needed _and_ head sync has completed -
# if we lose sync after having synced head, we could stop the backfilller,
# but this should be a fringe case - might as well keep the logic simple for
# now
node.backfiller.start()
return
await sleepAsync(chronos.seconds(2))
proc run(node: BeaconNode) {.raises: [Defect, CatchableError].} =
bnStatus = BeaconNodeStatus.Running
if not isNil(node.restServer):
node.restServer.installRestHandlers(node)
node.restServer.start()
2021-03-17 18:46:45 +00:00
if not isNil(node.keymanagerServer):
doAssert not isNil(node.keymanagerHost)
node.keymanagerServer.router.installKeymanagerHandlers(node.keymanagerHost[])
if node.keymanagerServer != node.restServer:
node.keymanagerServer.start()
let
wallTime = node.beaconClock.now()
wallSlot = wallTime.slotOrZero()
node.startLightClient()
node.requestManager.start()
node.syncManager.start()
if node.dag.needsBackfill(): asyncSpawn node.startBackfillTask()
waitFor node.updateGossipStatus(wallSlot)
asyncSpawn runSlotLoop(node, wallTime, onSlotStart)
asyncSpawn runOnSecondLoop(node)
asyncSpawn runQueueProcessingLoop(node.blockProcessor)
## Ctrl+C handling
proc controlCHandler() {.noconv.} =
when defined(windows):
# workaround for https://github.com/nim-lang/Nim/issues/4057
try:
setupForeignThreadGc()
except Exception as exc: raiseAssert exc.msg # shouldn't happen
notice "Shutting down after having received SIGINT"
bnStatus = BeaconNodeStatus.Stopping
try:
setControlCHook(controlCHandler)
except Exception as exc: # TODO Exception
warn "Cannot set ctrl-c handler", msg = exc.msg
# equivalent SIGTERM handler
when defined(posix):
proc SIGTERMHandler(signal: cint) {.noconv.} =
notice "Shutting down after having received SIGTERM"
bnStatus = BeaconNodeStatus.Stopping
c_signal(ansi_c.SIGTERM, SIGTERMHandler)
# main event loop
while bnStatus == BeaconNodeStatus.Running:
poll() # if poll fails, the network is broken
2020-04-20 14:59:18 +00:00
# time to say goodbye
node.stop()
var gPidFile: string
proc createPidFile(filename: string) {.raises: [Defect, IOError].} =
2019-07-07 09:53:58 +00:00
writeFile filename, $os.getCurrentProcessId()
gPidFile = filename
addQuitProc proc {.noconv.} = discard io2.removeFile(gPidFile)
proc initializeNetworking(node: BeaconNode) {.async.} =
node.installMessageValidators()
info "Listening to incoming network requests"
await node.network.startListening()
let addressFile = node.config.dataDir / "beacon_node.enr"
writeFile(addressFile, node.network.announcedENR.toURI)
await node.network.start()
proc start*(node: BeaconNode) {.raises: [Defect, CatchableError].} =
let
head = node.dag.head
finalizedHead = node.dag.finalizedHead
genesisTime = node.beaconClock.fromNow(start_beacon_time(Slot 0))
notice "Starting beacon node",
version = fullVersionStr,
2022-08-12 14:53:15 +00:00
nimVersion = NimVersion,
enr = node.network.announcedENR.toURI,
peerId = $node.network.switch.peerInfo.peerId,
2019-08-16 11:16:56 +00:00
timeSinceFinalization =
node.beaconClock.now() - finalizedHead.slot.start_beacon_time(),
head = shortLog(head),
Logging and startup improvements (#3038) * Logging and startup improvements Color support for released binaries! * startup scripts no longer log to file by default - this only affects source builds - released binaries don't support file logging * add --log-stdout option to control logging to stdout (colors, json) * detect tty:s vs redirected logs and log accordingly * add option to disable log colors at runtime * simplify several "common" logs, showing the most important information earlier and more clearly * remove line numbers / file information / tid - these take up space and are of little use to end users * still enabled in debug builds and tools * remove `testnet_servers_image` compile-time option * server images, released binaries and compile-from-source now offer the same behaviour and features * fixes https://github.com/status-im/nimbus-eth2/issues/2326 * fixes https://github.com/status-im/nimbus-eth2/issues/1794 * remove instanteneous block speed from sync message, keeping only average before: ``` INF 2021-10-28 16:45:59.000+02:00 Slot start topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:884 lastSlot=2384027 wallSlot=2384028 delay=461us84ns peers=0 head=75a10ee5:3348 headEpoch=104 finalized=cd6804ba:3264 finalizedEpoch=102 sync="wwwwwwwwww:0:0.0000:0.0000:00h00m (3348)" INF 2021-10-28 16:45:59.046+02:00 Slot end topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:821 slot=2384028 nextSlot=2384029 head=75a10ee5:3348 headEpoch=104 finalizedHead=cd6804ba:3264 finalizedEpoch=102 nextAttestationSlot=-1 nextProposalSlot=-1 nextActionWait=n/a ``` after: ``` INF 2021-10-28 22:43:23.033+02:00 Slot start topics="beacnde" slot=2385815 epoch=74556 sync="DDPDDPUDDD:10:5.2258:01h19m (2361088)" peers=37 head=eacd2dae:2361096 finalized=73782:a4751487 delay=33ms687us715ns INF 2021-10-28 22:43:23.291+02:00 Slot end topics="beacnde" slot=2385815 nextActionWait=n/a nextAttestationSlot=-1 nextProposalSlot=-1 head=eacd2dae:2361096 ``` * fix comment * documentation updates * mention `--log-file` may be deprecated in the future * update various docs
2021-11-02 17:06:36 +00:00
justified = shortLog(getStateField(
node.dag.headState, current_justified_checkpoint)),
Logging and startup improvements (#3038) * Logging and startup improvements Color support for released binaries! * startup scripts no longer log to file by default - this only affects source builds - released binaries don't support file logging * add --log-stdout option to control logging to stdout (colors, json) * detect tty:s vs redirected logs and log accordingly * add option to disable log colors at runtime * simplify several "common" logs, showing the most important information earlier and more clearly * remove line numbers / file information / tid - these take up space and are of little use to end users * still enabled in debug builds and tools * remove `testnet_servers_image` compile-time option * server images, released binaries and compile-from-source now offer the same behaviour and features * fixes https://github.com/status-im/nimbus-eth2/issues/2326 * fixes https://github.com/status-im/nimbus-eth2/issues/1794 * remove instanteneous block speed from sync message, keeping only average before: ``` INF 2021-10-28 16:45:59.000+02:00 Slot start topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:884 lastSlot=2384027 wallSlot=2384028 delay=461us84ns peers=0 head=75a10ee5:3348 headEpoch=104 finalized=cd6804ba:3264 finalizedEpoch=102 sync="wwwwwwwwww:0:0.0000:0.0000:00h00m (3348)" INF 2021-10-28 16:45:59.046+02:00 Slot end topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:821 slot=2384028 nextSlot=2384029 head=75a10ee5:3348 headEpoch=104 finalizedHead=cd6804ba:3264 finalizedEpoch=102 nextAttestationSlot=-1 nextProposalSlot=-1 nextActionWait=n/a ``` after: ``` INF 2021-10-28 22:43:23.033+02:00 Slot start topics="beacnde" slot=2385815 epoch=74556 sync="DDPDDPUDDD:10:5.2258:01h19m (2361088)" peers=37 head=eacd2dae:2361096 finalized=73782:a4751487 delay=33ms687us715ns INF 2021-10-28 22:43:23.291+02:00 Slot end topics="beacnde" slot=2385815 nextActionWait=n/a nextAttestationSlot=-1 nextProposalSlot=-1 head=eacd2dae:2361096 ``` * fix comment * documentation updates * mention `--log-file` may be deprecated in the future * update various docs
2021-11-02 17:06:36 +00:00
finalized = shortLog(getStateField(
node.dag.headState, finalized_checkpoint)),
finalizedHead = shortLog(finalizedHead),
SLOTS_PER_EPOCH,
SECONDS_PER_SLOT,
SPEC_VERSION,
dataDir = node.config.dataDir.string,
validators = node.attachedValidators[].count
if genesisTime.inFuture:
notice "Waiting for genesis", genesisIn = genesisTime.offset
waitFor node.initializeNetworking()
if node.eth1Monitor != nil:
node.eth1Monitor.start()
node.run()
2019-10-03 01:51:44 +00:00
func formatGwei(amount: uint64): string =
# TODO This is implemented in a quite a silly way.
# Better routines for formatting decimal numbers
# should exists somewhere else.
let
eth = amount div 1000000000
remainder = amount mod 1000000000
result = $eth
if remainder != 0:
result.add '.'
let remainderStr = $remainder
for i in remainderStr.len ..< 9:
result.add '0'
result.add remainderStr
2019-10-03 01:51:44 +00:00
while result[^1] == '0':
result.setLen(result.len - 1)
when not defined(windows):
proc initStatusBar(node: BeaconNode) {.raises: [Defect, ValueError].} =
if not isatty(stdout): return
if not node.config.statusBarEnabled: return
try:
enableTrueColors()
except Exception as exc: # TODO Exception
error "Couldn't enable colors", err = exc.msg
proc dataResolver(expr: string): string {.raises: [Defect].} =
template justified: untyped = node.dag.head.atEpochStart(
getStateField(
node.dag.headState, current_justified_checkpoint).epoch)
# TODO:
# We should introduce a general API for resolving dot expressions
# such as `db.latest_block.slot` or `metrics.connected_peers`.
# Such an API can be shared between the RPC back-end, CLI tools
# such as ncli, a potential GraphQL back-end and so on.
# The status bar feature would allow the user to specify an
# arbitrary expression that is resolvable through this API.
case expr.toLowerAscii
of "version":
versionAsStr
of "full_version":
fullVersionStr
of "connected_peers":
$(node.connectedPeersCount)
of "head_root":
shortLog(node.dag.head.root)
of "head_epoch":
$(node.dag.head.slot.epoch)
of "head_epoch_slot":
$(node.dag.head.slot.since_epoch_start)
of "head_slot":
$(node.dag.head.slot)
of "justifed_root":
shortLog(justified.blck.root)
of "justifed_epoch":
$(justified.slot.epoch)
of "justifed_epoch_slot":
$(justified.slot.since_epoch_start)
of "justifed_slot":
$(justified.slot)
of "finalized_root":
shortLog(node.dag.finalizedHead.blck.root)
of "finalized_epoch":
$(node.dag.finalizedHead.slot.epoch)
of "finalized_epoch_slot":
$(node.dag.finalizedHead.slot.since_epoch_start)
of "finalized_slot":
$(node.dag.finalizedHead.slot)
of "epoch":
$node.currentSlot.epoch
of "epoch_slot":
$(node.currentSlot.since_epoch_start)
of "slot":
$node.currentSlot
of "slots_per_epoch":
$SLOTS_PER_EPOCH
of "slot_trailing_digits":
var slotStr = $node.currentSlot
if slotStr.len > 3: slotStr = slotStr[^3..^1]
slotStr
of "attached_validators_balance":
formatGwei(node.attachedValidatorBalanceTotal)
of "sync_status":
node.syncStatus(node.currentSlot)
else:
# We ignore typos for now and just render the expression
# as it was written. TODO: come up with a good way to show
# an error message to the user.
"$" & expr
var statusBar = StatusBarView.init(
node.config.statusBarContents,
dataResolver)
when compiles(defaultChroniclesStream.outputs[0].writer):
let tmp = defaultChroniclesStream.outputs[0].writer
defaultChroniclesStream.outputs[0].writer =
proc (logLevel: LogLevel, msg: LogOutputStr) {.raises: [Defect].} =
try:
# p.hidePrompt
erase statusBar
# p.writeLine msg
tmp(logLevel, msg)
render statusBar
# p.showPrompt
except Exception as e: # render raises Exception
logLoggingFailure(cstring(msg), e)
proc statusBarUpdatesPollingLoop() {.async.} =
try:
while true:
update statusBar
erase statusBar
render statusBar
await sleepAsync(chronos.seconds(1))
except CatchableError as exc:
warn "Failed to update status bar, no further updates", err = exc.msg
asyncSpawn statusBarUpdatesPollingLoop()
2022-06-21 08:29:16 +00:00
proc doRunBeaconNode(config: var BeaconNodeConf, rng: ref HmacDrbgContext) {.raises: [Defect, CatchableError].} =
info "Launching beacon node",
version = fullVersionStr,
bls_backend = $BLS_BACKEND,
cmdParams = commandLineParams(),
config
template ignoreDeprecatedOption(option: untyped): untyped =
if config.option.isSome:
warn "Config option is deprecated",
option = config.option.get
ignoreDeprecatedOption requireEngineAPI
ignoreDeprecatedOption safeSlotsToImportOptimistically
ignoreDeprecatedOption terminalTotalDifficultyOverride
ignoreDeprecatedOption optimistic
createPidFile(config.dataDir.string / "beacon_node.pid")
config.createDumpDirs()
if config.metricsEnabled:
let metricsAddress = config.metricsAddress
notice "Starting metrics HTTP server",
url = "http://" & $metricsAddress & ":" & $config.metricsPort & "/metrics"
try:
startMetricsHttpServer($metricsAddress, config.metricsPort)
except CatchableError as exc:
raise exc
except Exception as exc:
raiseAssert exc.msg # TODO fix metrics
# Nim GC metrics (for the main thread) will be collected in onSecond(), but
# we disable piggy-backing on other metrics here.
setSystemMetricsAutomaticUpdate(false)
# There are no managed event loops in here, to do a graceful shutdown, but
# letting the default Ctrl+C handler exit is safe, since we only read from
# the db.
let metadata = config.loadEth2Network()
# Updating the config based on the metadata certainly is not beautiful but it
# works
for node in metadata.bootstrapNodes:
config.bootstrapNodes.add node
let node = BeaconNode.init(rng, config, metadata)
if bnStatus == BeaconNodeStatus.Stopping:
return
when not defined(windows):
# This status bar can lock a Windows terminal emulator, blocking the whole
# event loop (seen on Windows 10, with a default MSYS2 terminal).
initStatusBar(node)
if node.nickname != "":
dynamicLogScope(node = node.nickname): node.start()
else:
node.start()
2022-06-21 08:29:16 +00:00
proc doCreateTestnet*(config: BeaconNodeConf, rng: var HmacDrbgContext) {.raises: [Defect, CatchableError].} =
let launchPadDeposits = try:
Json.loadFile(config.testnetDepositsFile.string, seq[LaunchPadDeposit])
except SerializationError as err:
error "Invalid LaunchPad deposits file",
err = formatMsg(err, config.testnetDepositsFile.string)
quit 1
var deposits: seq[DepositData]
2021-04-14 13:15:22 +00:00
for i in 0 ..< launchPadDeposits.len:
deposits.add(launchPadDeposits[i] as DepositData)
let
startTime = uint64(times.toUnix(times.getTime()) + config.genesisOffset)
outGenesis = config.outputGenesis.string
eth1Hash = if config.web3Urls.len == 0: eth1BlockHash
2022-03-31 14:43:05 +00:00
else: (waitFor getEth1BlockHash(
config.web3Urls[0], blockId("latest"),
rng.loadJwtSecret(config, allowCreate = true))).asEth2Digest
cfg = getRuntimeConfig(config.eth2Network)
var
initialState = newClone(initialize_beacon_state_from_eth1(
cfg, eth1Hash, startTime, deposits, {skipBlsValidation}))
# https://github.com/ethereum/eth2.0-pm/tree/6e41fcf383ebeb5125938850d8e9b4e9888389b4/interop/mocked_start#create-genesis-state
initialState.genesis_time = startTime
doAssert initialState.validators.len > 0
let outGenesisExt = splitFile(outGenesis).ext
if cmpIgnoreCase(outGenesisExt, ".json") == 0:
Json.saveFile(outGenesis, initialState, pretty = true)
echo "Wrote ", outGenesis
let outSszGenesis = outGenesis.changeFileExt "ssz"
SSZ.saveFile(outSszGenesis, initialState[])
echo "Wrote ", outSszGenesis
let bootstrapFile = config.outputBootstrapFile.string
if bootstrapFile.len > 0:
type MetaData = altair.MetaData
let
networkKeys = getPersistentNetKeys(rng, config)
netMetadata = MetaData()
forkId = getENRForkID(
cfg,
initialState[].slot.epoch,
initialState[].genesis_validators_root)
bootstrapEnr = enr.Record.init(
1, # sequence number
networkKeys.seckey.asEthKey,
some(config.bootstrapAddress),
some(config.bootstrapPort),
some(config.bootstrapPort),
[
toFieldPair(enrForkIdField, SSZ.encode(forkId)),
toFieldPair(enrAttestationSubnetsField, SSZ.encode(netMetadata.attnets))
])
writeFile(bootstrapFile, bootstrapEnr.tryGet().toURI)
echo "Wrote ", bootstrapFile
2022-06-21 08:29:16 +00:00
proc doRecord(config: BeaconNodeConf, rng: var HmacDrbgContext) {.
raises: [Defect, CatchableError].} =
case config.recordCmd:
of RecordCmd.create:
let netKeys = getPersistentNetKeys(rng, config)
var fieldPairs: seq[FieldPair]
for field in config.fields:
let fieldPair = field.split(":")
if fieldPair.len > 1:
fieldPairs.add(toFieldPair(fieldPair[0], hexToSeqByte(fieldPair[1])))
else:
fatal "Invalid field pair"
quit QuitFailure
let record = enr.Record.init(
config.seqNumber,
netKeys.seckey.asEthKey,
some(config.ipExt),
some(config.tcpPortExt),
some(config.udpPortExt),
fieldPairs).expect("Record within size limits")
echo record.toURI()
of RecordCmd.print:
echo $config.recordPrint
2022-06-21 08:29:16 +00:00
proc doWeb3Cmd(config: BeaconNodeConf, rng: var HmacDrbgContext)
2022-03-31 14:43:05 +00:00
{.raises: [Defect, CatchableError].} =
case config.web3Cmd:
of Web3Cmd.test:
let metadata = config.loadEth2Network()
2022-03-31 14:43:05 +00:00
waitFor testWeb3Provider(config.web3TestUrl,
2022-03-31 14:43:05 +00:00
metadata.cfg.DEPOSIT_CONTRACT_ADDRESS,
rng.loadJwtSecret(config, allowCreate = true))
2021-05-19 06:38:13 +00:00
proc doSlashingExport(conf: BeaconNodeConf) {.raises: [IOError, Defect].}=
let
dir = conf.validatorsDir()
filetrunc = SlashingDbName
# TODO: Make it read-only https://github.com/status-im/nim-eth/issues/312
let db = SlashingProtectionDB.loadUnchecked(dir, filetrunc, readOnly = false)
let interchange = conf.exportedInterchangeFile.string
db.exportSlashingInterchange(interchange, conf.exportedValidators)
echo "Export finished: '", dir/filetrunc & ".sqlite3" , "' into '", interchange, "'"
proc doSlashingImport(conf: BeaconNodeConf) {.raises: [SerializationError, IOError, Defect].} =
let
dir = conf.validatorsDir()
filetrunc = SlashingDbName
# TODO: Make it read-only https://github.com/status-im/nim-eth/issues/312
let interchange = conf.importedInterchangeFile.string
var spdir: SPDIR
try:
spdir = Json.loadFile(interchange, SPDIR,
requireAllFields = true)
2021-05-19 06:38:13 +00:00
except SerializationError as err:
writeStackTrace()
stderr.write $Json & " load issue for file \"", interchange, "\"\n"
2021-05-19 06:38:13 +00:00
stderr.write err.formatMsg(interchange), "\n"
quit 1
# Open DB and handle migration from v1 to v2 if needed
let db = SlashingProtectionDB.init(
genesis_validators_root = Eth2Digest spdir.metadata.genesis_validators_root,
basePath = dir,
dbname = filetrunc,
modes = {kCompleteArchive}
)
# Now import the slashing interchange file
# Failures mode:
# - siError can only happen with invalid genesis_validators_root which would be caught above
# - siPartial can happen for invalid public keys, slashable blocks, slashable votes
let status = db.inclSPDIR(spdir)
doAssert status in {siSuccess, siPartial}
echo "Import finished: '", interchange, "' into '", dir/filetrunc & ".sqlite3", "'"
proc doSlashingInterchange(conf: BeaconNodeConf) {.raises: [Defect, CatchableError].} =
case conf.slashingdbCmd
of SlashProtCmd.`export`:
conf.doSlashingExport()
of SlashProtCmd.`import`:
conf.doSlashingImport()
proc handleStartUpCmd(config: var BeaconNodeConf) {.raises: [Defect, CatchableError].} =
# Single RNG instance for the application - will be seeded on construction
# and avoid using system resources (such as urandom) after that
let rng = keys.newRng()
case config.cmd
of BNStartUpCmd.createTestnet: doCreateTestnet(config, rng[])
of BNStartUpCmd.noCommand: doRunBeaconNode(config, rng)
of BNStartUpCmd.deposits: doDeposits(config, rng[])
of BNStartUpCmd.wallets: doWallets(config, rng[])
of BNStartUpCmd.record: doRecord(config, rng[])
2022-03-31 14:43:05 +00:00
of BNStartUpCmd.web3: doWeb3Cmd(config, rng[])
of BNStartUpCmd.slashingdb: doSlashingInterchange(config)
of BNStartUpCmd.trustedNodeSync:
let
network = loadEth2Network(config)
cfg = network.cfg
genesis =
if network.genesisData.len > 0:
newClone(readSszForkedHashedBeaconState(
cfg,
network.genesisData.toOpenArrayByte(0, network.genesisData.high())))
else: nil
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
if config.blockId.isSome():
error "--blockId option has been removed - use --state-id instead!"
quit 1
waitFor doTrustedNodeSync(
cfg,
config.databaseDir,
config.eraDir,
config.trustedNodeUrl,
State-only checkpoint state startup (#4251) Currently, we require genesis and a checkpoint block and state to start from an arbitrary slot - this PR relaxes this requirement so that we can start with a state alone. The current trusted-node-sync algorithm works by first downloading blocks until we find an epoch aligned non-empty slot, then downloads the state via slot. However, current [proposals](https://github.com/ethereum/beacon-APIs/pull/226) for checkpointing prefer finalized state as the main reference - this allows more simple access control and caching on the server side - in particular, this should help checkpoint-syncing from sources that have a fast `finalized` state download (like infura and teku) but are slow when accessing state via slot. Earlier versions of Nimbus will not be able to read databases created without a checkpoint block and genesis. In most cases, backfilling makes the database compatible except where genesis is also missing (custom networks). * backfill checkpoint block from libp2p instead of checkpoint source, when doing trusted node sync * allow starting the client without genesis / checkpoint block * perform epoch start slot lookahead when loading tail state, so as to deal with the case where the epoch start slot does not have a block * replace `--blockId` with `--state-id` in TNS command line * when replaying, also look at the parent of the last-known-block (even if we don't have the parent block data, we can still replay from a "parent" state) - in particular, this clears the way for implementing state pruning * deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
config.stateId,
config.backfillBlocks,
config.reindex,
config.downloadDepositSnapshot,
genesis)
{.pop.} # TODO moduletests exceptions
when defined(windows):
proc reportServiceStatus*(dwCurrentState, dwWin32ExitCode, dwWaitHint: DWORD) {.gcsafe.} =
gSvcStatus.dwCurrentState = dwCurrentState
gSvcStatus.dwWin32ExitCode = dwWin32ExitCode
gSvcStatus.dwWaitHint = dwWaitHint
if dwCurrentState == SERVICE_START_PENDING:
gSvcStatus.dwControlsAccepted = 0
else:
gSvcStatus.dwControlsAccepted = SERVICE_ACCEPT_STOP
# TODO
# We can use non-zero values for the `dwCheckPoint` parameter to report
# progress during lengthy operations such as start-up and shut down.
gSvcStatus.dwCheckPoint = 0
# Report the status of the service to the SCM.
let status = SetServiceStatus(gSvcStatusHandle, addr gSvcStatus)
debug "Service status updated", status
proc serviceControlHandler(dwCtrl: DWORD): WINBOOL {.stdcall.} =
case dwCtrl
of SERVICE_CONTROL_STOP:
# We re reporting that we plan stop the service in 10 seconds
reportServiceStatus(SERVICE_STOP_PENDING, NO_ERROR, 10_000)
bnStatus = BeaconNodeStatus.Stopping
of SERVICE_CONTROL_PAUSE, SERVICE_CONTROL_CONTINUE:
warn "The Nimbus service cannot be paused and resimed"
of SERVICE_CONTROL_INTERROGATE:
# The default behavior is correct.
# The service control manager will report our last status.
discard
else:
debug "Service received an unexpected user-defined control message",
msg = dwCtrl
proc serviceMainFunction(dwArgc: DWORD, lpszArgv: LPSTR) {.stdcall.} =
# The service is launched in a fresh thread created by Windows, so
# we must initialize the Nim GC here
setupForeignThreadGc()
gSvcStatusHandle = RegisterServiceCtrlHandler(
SERVICE_NAME,
serviceControlHandler)
gSvcStatus.dwServiceType = SERVICE_WIN32_OWN_PROCESS
gSvcStatus.dwServiceSpecificExitCode = 0
reportServiceStatus(SERVICE_RUNNING, NO_ERROR, 0)
info "Service thread started"
var config = makeBannerAndConfig(clientId, BeaconNodeConf)
handleStartUpCmd(config)
info "Service thread stopped"
reportServiceStatus(SERVICE_STOPPED, NO_ERROR, 0) # we have to report back when we stopped!
programMain:
var config = makeBannerAndConfig(clientId, BeaconNodeConf)
2020-11-28 18:50:09 +00:00
if not(checkAndCreateDataDir(string(config.dataDir))):
# We are unable to access/create data folder or data folder's
# permissions are insecure.
quit QuitFailure
Logging and startup improvements (#3038) * Logging and startup improvements Color support for released binaries! * startup scripts no longer log to file by default - this only affects source builds - released binaries don't support file logging * add --log-stdout option to control logging to stdout (colors, json) * detect tty:s vs redirected logs and log accordingly * add option to disable log colors at runtime * simplify several "common" logs, showing the most important information earlier and more clearly * remove line numbers / file information / tid - these take up space and are of little use to end users * still enabled in debug builds and tools * remove `testnet_servers_image` compile-time option * server images, released binaries and compile-from-source now offer the same behaviour and features * fixes https://github.com/status-im/nimbus-eth2/issues/2326 * fixes https://github.com/status-im/nimbus-eth2/issues/1794 * remove instanteneous block speed from sync message, keeping only average before: ``` INF 2021-10-28 16:45:59.000+02:00 Slot start topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:884 lastSlot=2384027 wallSlot=2384028 delay=461us84ns peers=0 head=75a10ee5:3348 headEpoch=104 finalized=cd6804ba:3264 finalizedEpoch=102 sync="wwwwwwwwww:0:0.0000:0.0000:00h00m (3348)" INF 2021-10-28 16:45:59.046+02:00 Slot end topics="beacnde" tid=386429 file=nimbus_beacon_node.nim:821 slot=2384028 nextSlot=2384029 head=75a10ee5:3348 headEpoch=104 finalizedHead=cd6804ba:3264 finalizedEpoch=102 nextAttestationSlot=-1 nextProposalSlot=-1 nextActionWait=n/a ``` after: ``` INF 2021-10-28 22:43:23.033+02:00 Slot start topics="beacnde" slot=2385815 epoch=74556 sync="DDPDDPUDDD:10:5.2258:01h19m (2361088)" peers=37 head=eacd2dae:2361096 finalized=73782:a4751487 delay=33ms687us715ns INF 2021-10-28 22:43:23.291+02:00 Slot end topics="beacnde" slot=2385815 nextActionWait=n/a nextAttestationSlot=-1 nextProposalSlot=-1 head=eacd2dae:2361096 ``` * fix comment * documentation updates * mention `--log-file` may be deprecated in the future * update various docs
2021-11-02 17:06:36 +00:00
setupLogging(config.logLevel, config.logStdout, config.logFile)
## This Ctrl+C handler exits the program in non-graceful way.
## It's responsible for handling Ctrl+C in sub-commands such
## as `wallets *` and `deposits *`. In a regular beacon node
## run, it will be overwritten later with a different handler
## performing a graceful exit.
proc exitImmediatelyOnCtrlC() {.noconv.} =
when defined(windows):
# workaround for https://github.com/nim-lang/Nim/issues/4057
setupForeignThreadGc()
# in case a password prompt disabled echoing
resetStdin()
echo "" # If we interrupt during an interactive prompt, this
# will move the cursor to the next line
notice "Shutting down after having received SIGINT"
quit 0
setControlCHook(exitImmediatelyOnCtrlC)
# equivalent SIGTERM handler
when defined(posix):
proc exitImmediatelyOnSIGTERM(signal: cint) {.noconv.} =
notice "Shutting down after having received SIGTERM"
quit 0
c_signal(ansi_c.SIGTERM, exitImmediatelyOnSIGTERM)
when defined(windows):
if config.runAsService:
var dispatchTable = [
SERVICE_TABLE_ENTRY(lpServiceName: SERVICE_NAME, lpServiceProc: serviceMainFunction),
SERVICE_TABLE_ENTRY(lpServiceName: nil, lpServiceProc: nil) # last entry must be nil
]
let status = StartServiceCtrlDispatcher(LPSERVICE_TABLE_ENTRY(addr dispatchTable[0]))
if status == 0:
fatal "Failed to start Windows service", errorCode = getLastError()
quit 1
else:
handleStartUpCmd(config)
else:
handleStartUpCmd(config)