nimbus-eth2/beacon_chain/nimbus_light_client.nim

344 lines
13 KiB
Nim
Raw Normal View History

# beacon_chain
# Copyright (c) 2022-2023 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.
import
std/os,
chronicles, chronicles/chronos_tools, chronos, stew/io2,
eth/db/kvstore_sqlite3, eth/keys,
./eth1/eth1_monitor,
./gossip_processing/optimistic_processor,
./networking/topic_params,
./spec/beaconstate,
./spec/datatypes/[phase0, altair, bellatrix, capella, eip4844],
"."/[filepath, light_client, light_client_db, nimbus_binary_common, version]
from ./gossip_processing/block_processor import newExecutionPayload
from ./gossip_processing/eth2_processor import toValidationResult
# this needs to be global, so it can be set in the Ctrl+C signal handler
var globalRunning = true
programMain:
## 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"
globalRunning = false
try:
setControlCHook(controlCHandler)
except Exception as exc: # TODO Exception
warn "Cannot set ctrl-c handler", msg = exc.msg
var config = makeBannerAndConfig(
"Nimbus light client " & fullVersionStr, LightClientConf)
setupLogging(config.logLevel, config.logStdout, config.logFile)
notice "Launching light client",
version = fullVersionStr, cmdParams = commandLineParams(), config
let dbDir = config.databaseDir
if (let res = secureCreatePath(dbDir); res.isErr):
fatal "Failed to create create database directory",
path = dbDir, err = ioErrorMsg(res.error)
quit 1
let backend = SqStoreRef.init(dbDir, "nlc").expect("Database OK")
defer: backend.close()
let db = backend.initLightClientDB(LightClientDBNames(
legacyAltairHeaders: "altair_lc_headers",
headers: "lc_headers",
altairSyncCommittees: "altair_sync_committees")).expect("Database OK")
defer: db.close()
let metadata = loadEth2Network(config.eth2Network)
for node in metadata.bootstrapNodes:
config.bootstrapNodes.add node
template cfg(): auto = metadata.cfg
let
genesisState =
try:
template genesisData(): auto = metadata.genesisData
newClone(readSszForkedHashedBeaconState(
cfg, genesisData.toOpenArrayByte(genesisData.low, genesisData.high)))
except CatchableError as err:
raiseAssert "Invalid baked-in state: " & err.msg
beaconClock = BeaconClock.init(
getStateField(genesisState[], genesis_time))
getBeaconTime = beaconClock.getBeaconTimeFn()
genesis_validators_root =
getStateField(genesisState[], genesis_validators_root)
forkDigests = newClone ForkDigests.init(cfg, genesis_validators_root)
genesisBlockRoot = get_initial_beacon_block(genesisState[]).root
rng = keys.newRng()
netKeys = getRandomNetKeys(rng[])
network = createEth2Node(
rng, config, netKeys, cfg,
forkDigests, getBeaconTime, genesis_validators_root)
Support for driving multiple EL nodes from a single Nimbus BN (#4465) * Support for driving multiple EL nodes from a single Nimbus BN Full list of changes: * Eth1Monitor has been renamed to ELManager to match its current responsibilities better. * The ELManager is no longer optional in the code (it won't have a nil value under any circumstances). * The support for subscribing for headers was removed as it only worked with WebSockets and contributed significant complexity while bringing only a very minor advantage. * The `--web3-url` parameter has been deprecated in favor of a new `--el` parameter. The new parameter has a reasonable default value and supports specifying a different JWT for each connection. Each connection can also be configured with a different set of responsibilities (e.g. download deposits, validate blocks and/or produce blocks). On the command-line, these properties can be configured through URL properties stored in the #anchor part of the URL. In TOML files, they come with a very natural syntax (althrough the URL scheme is also supported). * The previously scattered EL-related state and logic is now moved to `eth1_monitor.nim` (this module will be renamed to `el_manager.nim` in a follow-up commit). State is assigned properly either to the `ELManager` or the to individual `ELConnection` objects where appropriate. The ELManager executes all Engine API requests against all attached EL nodes, in parallel. It compares their results and if there is a disagreement regarding the validity of a certain payload, this is detected and the beacon node is protected from publishing a block with a potential execution layer consensus bug in it. The BN provides metrics per EL node for the number of successful or failed requests for each type Engine API requests. If an EL node goes offline and connectivity is resoted later, we report the problem and the remedy in edge-triggered fashion. * More progress towards implementing Deneb block production in the VC and comparing the value of blocks produced by the EL and the builder API. * Adds a Makefile target for the zhejiang testnet
2023-03-05 01:40:21 +00:00
engineApiUrls = config.engineApiUrls
elManager =
if engineApiUrls.len > 0:
ELManager.new(
cfg,
metadata.depositContractBlock,
metadata.depositContractBlockHash,
db = nil,
Support for driving multiple EL nodes from a single Nimbus BN (#4465) * Support for driving multiple EL nodes from a single Nimbus BN Full list of changes: * Eth1Monitor has been renamed to ELManager to match its current responsibilities better. * The ELManager is no longer optional in the code (it won't have a nil value under any circumstances). * The support for subscribing for headers was removed as it only worked with WebSockets and contributed significant complexity while bringing only a very minor advantage. * The `--web3-url` parameter has been deprecated in favor of a new `--el` parameter. The new parameter has a reasonable default value and supports specifying a different JWT for each connection. Each connection can also be configured with a different set of responsibilities (e.g. download deposits, validate blocks and/or produce blocks). On the command-line, these properties can be configured through URL properties stored in the #anchor part of the URL. In TOML files, they come with a very natural syntax (althrough the URL scheme is also supported). * The previously scattered EL-related state and logic is now moved to `eth1_monitor.nim` (this module will be renamed to `el_manager.nim` in a follow-up commit). State is assigned properly either to the `ELManager` or the to individual `ELConnection` objects where appropriate. The ELManager executes all Engine API requests against all attached EL nodes, in parallel. It compares their results and if there is a disagreement regarding the validity of a certain payload, this is detected and the beacon node is protected from publishing a block with a potential execution layer consensus bug in it. The BN provides metrics per EL node for the number of successful or failed requests for each type Engine API requests. If an EL node goes offline and connectivity is resoted later, we report the problem and the remedy in edge-triggered fashion. * More progress towards implementing Deneb block production in the VC and comparing the value of blocks produced by the EL and the builder API. * Adds a Makefile target for the zhejiang testnet
2023-03-05 01:40:21 +00:00
engineApiUrls,
metadata.eth1Network)
else:
nil
optimisticHandler = proc(signedBlock: ForkedMsgTrustedSignedBeaconBlock):
Future[void] {.async.} =
notice "New LC optimistic block",
opt = signedBlock.toBlockId(),
wallSlot = getBeaconTime().slotOrZero
withBlck(signedBlock):
when stateFork >= ConsensusFork.Bellatrix:
if blck.message.is_execution_block:
template payload(): auto = blck.message.body.execution_payload
Support for driving multiple EL nodes from a single Nimbus BN (#4465) * Support for driving multiple EL nodes from a single Nimbus BN Full list of changes: * Eth1Monitor has been renamed to ELManager to match its current responsibilities better. * The ELManager is no longer optional in the code (it won't have a nil value under any circumstances). * The support for subscribing for headers was removed as it only worked with WebSockets and contributed significant complexity while bringing only a very minor advantage. * The `--web3-url` parameter has been deprecated in favor of a new `--el` parameter. The new parameter has a reasonable default value and supports specifying a different JWT for each connection. Each connection can also be configured with a different set of responsibilities (e.g. download deposits, validate blocks and/or produce blocks). On the command-line, these properties can be configured through URL properties stored in the #anchor part of the URL. In TOML files, they come with a very natural syntax (althrough the URL scheme is also supported). * The previously scattered EL-related state and logic is now moved to `eth1_monitor.nim` (this module will be renamed to `el_manager.nim` in a follow-up commit). State is assigned properly either to the `ELManager` or the to individual `ELConnection` objects where appropriate. The ELManager executes all Engine API requests against all attached EL nodes, in parallel. It compares their results and if there is a disagreement regarding the validity of a certain payload, this is detected and the beacon node is protected from publishing a block with a potential execution layer consensus bug in it. The BN provides metrics per EL node for the number of successful or failed requests for each type Engine API requests. If an EL node goes offline and connectivity is resoted later, we report the problem and the remedy in edge-triggered fashion. * More progress towards implementing Deneb block production in the VC and comparing the value of blocks produced by the EL and the builder API. * Adds a Makefile target for the zhejiang testnet
2023-03-05 01:40:21 +00:00
if elManager != nil and not payload.block_hash.isZero:
discard await elManager.newExecutionPayload(payload)
discard await elManager.forkchoiceUpdated(
headBlockHash = payload.block_hash,
safeBlockHash = payload.block_hash, # stub value
finalizedBlockHash = ZERO_HASH)
else: discard
optimisticProcessor = initOptimisticProcessor(
getBeaconTime, optimisticHandler)
lightClient = createLightClient(
network, rng, config, cfg, forkDigests, getBeaconTime,
genesis_validators_root, LightClientFinalizationMode.Optimistic)
info "Listening to incoming network requests"
network.initBeaconSync(cfg, forkDigests, genesisBlockRoot, getBeaconTime)
network.addValidator(
getBeaconBlocksTopic(forkDigests.phase0),
proc (signedBlock: phase0.SignedBeaconBlock): ValidationResult =
toValidationResult(
optimisticProcessor.processSignedBeaconBlock(signedBlock)))
network.addValidator(
getBeaconBlocksTopic(forkDigests.altair),
proc (signedBlock: altair.SignedBeaconBlock): ValidationResult =
toValidationResult(
optimisticProcessor.processSignedBeaconBlock(signedBlock)))
network.addValidator(
getBeaconBlocksTopic(forkDigests.bellatrix),
proc (signedBlock: bellatrix.SignedBeaconBlock): ValidationResult =
toValidationResult(
optimisticProcessor.processSignedBeaconBlock(signedBlock)))
network.addValidator(
getBeaconBlocksTopic(forkDigests.capella),
proc (signedBlock: capella.SignedBeaconBlock): ValidationResult =
toValidationResult(
optimisticProcessor.processSignedBeaconBlock(signedBlock)))
network.addValidator(
getBeaconBlocksTopic(forkDigests.eip4844),
proc (signedBlock: eip4844.SignedBeaconBlock): ValidationResult =
toValidationResult(
optimisticProcessor.processSignedBeaconBlock(signedBlock)))
lightClient.installMessageValidators()
waitFor network.startListening()
waitFor network.start()
proc onFinalizedHeader(
lightClient: LightClient, finalizedHeader: ForkedLightClientHeader) =
withForkyHeader(finalizedHeader):
when lcDataFork > LightClientDataFork.None:
info "New LC finalized header",
finalized_header = shortLog(forkyHeader)
let
period = forkyHeader.beacon.slot.sync_committee_period
syncCommittee = lightClient.finalizedSyncCommittee.expect("Init OK")
db.putSyncCommittee(period, syncCommittee)
db.putLatestFinalizedHeader(finalizedHeader)
proc onOptimisticHeader(
lightClient: LightClient, optimisticHeader: ForkedLightClientHeader) =
withForkyHeader(optimisticHeader):
when lcDataFork > LightClientDataFork.None:
info "New LC optimistic header",
optimistic_header = shortLog(forkyHeader)
optimisticProcessor.setOptimisticHeader(forkyHeader.beacon)
lightClient.onFinalizedHeader = onFinalizedHeader
lightClient.onOptimisticHeader = onOptimisticHeader
lightClient.trustedBlockRoot = some config.trustedBlockRoot
let latestHeader = db.getLatestFinalizedHeader()
withForkyHeader(latestHeader):
when lcDataFork > LightClientDataFork.None:
let
period = forkyHeader.beacon.slot.sync_committee_period
syncCommittee = db.getSyncCommittee(period)
if syncCommittee.isErr:
error "LC store lacks sync committee", finalized_header = forkyHeader
else:
lightClient.resetToFinalizedHeader(latestHeader, syncCommittee.get)
# Full blocks gossip is required to portably drive an EL client:
# - EL clients may not sync when only driven with `forkChoiceUpdated`,
# e.g., Geth: "Forkchoice requested unknown head"
# - `newPayload` requires the full `ExecutionPayload` (most of block content)
# - `ExecutionPayload` block hash is not available in
# `altair.LightClientHeader`, so won't be exchanged via light client gossip
#
# Future `ethereum/consensus-specs` versions may remove need for full blocks.
# Therefore, this current mechanism is to be seen as temporary; it is not
# optimized for reducing code duplication, e.g., with `nimbus_beacon_node`.
func isSynced(wallSlot: Slot): bool =
let optimisticHeader = lightClient.optimisticHeader
withForkyHeader(optimisticHeader):
when lcDataFork > LightClientDataFork.None:
# Check whether light client has synced sufficiently close to wall slot
const maxAge = 2 * SLOTS_PER_EPOCH
forkyHeader.beacon.slot >= max(wallSlot, maxAge.Slot) - maxAge
else:
false
func shouldSyncOptimistically(wallSlot: Slot): bool =
# Check whether an EL is connected
Support for driving multiple EL nodes from a single Nimbus BN (#4465) * Support for driving multiple EL nodes from a single Nimbus BN Full list of changes: * Eth1Monitor has been renamed to ELManager to match its current responsibilities better. * The ELManager is no longer optional in the code (it won't have a nil value under any circumstances). * The support for subscribing for headers was removed as it only worked with WebSockets and contributed significant complexity while bringing only a very minor advantage. * The `--web3-url` parameter has been deprecated in favor of a new `--el` parameter. The new parameter has a reasonable default value and supports specifying a different JWT for each connection. Each connection can also be configured with a different set of responsibilities (e.g. download deposits, validate blocks and/or produce blocks). On the command-line, these properties can be configured through URL properties stored in the #anchor part of the URL. In TOML files, they come with a very natural syntax (althrough the URL scheme is also supported). * The previously scattered EL-related state and logic is now moved to `eth1_monitor.nim` (this module will be renamed to `el_manager.nim` in a follow-up commit). State is assigned properly either to the `ELManager` or the to individual `ELConnection` objects where appropriate. The ELManager executes all Engine API requests against all attached EL nodes, in parallel. It compares their results and if there is a disagreement regarding the validity of a certain payload, this is detected and the beacon node is protected from publishing a block with a potential execution layer consensus bug in it. The BN provides metrics per EL node for the number of successful or failed requests for each type Engine API requests. If an EL node goes offline and connectivity is resoted later, we report the problem and the remedy in edge-triggered fashion. * More progress towards implementing Deneb block production in the VC and comparing the value of blocks produced by the EL and the builder API. * Adds a Makefile target for the zhejiang testnet
2023-03-05 01:40:21 +00:00
if elManager == nil:
return false
isSynced(wallSlot)
var blocksGossipState: GossipState = {}
proc updateBlocksGossipStatus(slot: Slot) =
let
isBehind = not shouldSyncOptimistically(slot)
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.DENEB_FORK_EPOCH, isBehind)
template currentGossipState(): auto = 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
for gossipFork in oldGossipForks:
let forkDigest = forkDigests[].atStateFork(gossipFork)
network.unsubscribe(getBeaconBlocksTopic(forkDigest))
for gossipFork in newGossipForks:
let forkDigest = forkDigests[].atStateFork(gossipFork)
network.subscribe(
getBeaconBlocksTopic(forkDigest), blocksTopicParams,
enableTopicMetrics = true)
blocksGossipState = targetGossipState
proc onSlot(wallTime: BeaconTime, lastSlot: Slot) =
let
wallSlot = wallTime.slotOrZero()
expectedSlot = lastSlot + 1
delay = wallTime - expectedSlot.start_beacon_time()
finalizedHeader = lightClient.finalizedHeader
optimisticHeader = lightClient.optimisticHeader
finalizedBid = withForkyHeader(finalizedHeader):
when lcDataFork > LightClientDataFork.None:
forkyHeader.beacon.toBlockId()
else:
BlockId(root: genesisBlockRoot, slot: GENESIS_SLOT)
optimisticBid = withForkyHeader(optimisticHeader):
when lcDataFork > LightClientDataFork.None:
forkyHeader.beacon.toBlockId()
else:
BlockId(root: genesisBlockRoot, slot: GENESIS_SLOT)
syncStatus =
if optimisticHeader.kind == LightClientDataFork.None:
"bootstrapping(" & $config.trustedBlockRoot & ")"
elif not isSynced(wallSlot):
"syncing"
else:
"synced"
info "Slot start",
slot = shortLog(wallSlot),
epoch = shortLog(wallSlot.epoch),
sync = syncStatus,
peers = len(network.peerPool),
head = shortLog(optimisticBid),
finalized = shortLog(finalizedBid),
delay = shortLog(delay)
proc runOnSlotLoop() {.async.} =
var
curSlot = getBeaconTime().slotOrZero()
nextSlot = curSlot + 1
timeToNextSlot = nextSlot.start_beacon_time() - getBeaconTime()
while true:
await sleepAsync(timeToNextSlot)
let
wallTime = getBeaconTime()
wallSlot = wallTime.slotOrZero()
onSlot(wallTime, curSlot)
curSlot = wallSlot
nextSlot = wallSlot + 1
timeToNextSlot = nextSlot.start_beacon_time() - getBeaconTime()
proc onSecond(time: Moment) =
let wallSlot = getBeaconTime().slotOrZero()
if checkIfShouldStopAtEpoch(wallSlot, config.stopAtEpoch):
quit(0)
updateBlocksGossipStatus(wallSlot + 1)
lightClient.updateGossipStatus(wallSlot + 1)
proc runOnSecondLoop() {.async.} =
let sleepTime = chronos.seconds(1)
while true:
let start = chronos.now(chronos.Moment)
await chronos.sleepAsync(sleepTime)
let afterSleep = chronos.now(chronos.Moment)
let sleepTime = afterSleep - start
onSecond(start)
let finished = chronos.now(chronos.Moment)
let processingTime = finished - afterSleep
trace "onSecond task completed", sleepTime, processingTime
onSecond(Moment.now())
lightClient.start()
asyncSpawn runOnSlotLoop()
asyncSpawn runOnSecondLoop()
while globalRunning:
poll()
notice "Exiting light client"