Add info-level diagnostic logs to aid with resolving stalled BFT chains #7271
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Signed-off-by: Matthew Whitehead <[email protected]>
Signed-off-by: Matthew Whitehead <[email protected]>
macfarla
approved these changes
Jun 27, 2024
Signed-off-by: Matt Whitehead <[email protected]>
daniellehrner
pushed a commit
to daniellehrner/besu
that referenced
this pull request
Jul 16, 2024
…ns (hyperledger#7271) * Add info-level diagnostic logs to aid with resolving stalled BFT chains Signed-off-by: Matthew Whitehead <[email protected]> * Add javadoc Signed-off-by: Matthew Whitehead <[email protected]> --------- Signed-off-by: Matthew Whitehead <[email protected]> Signed-off-by: Matt Whitehead <[email protected]>
daniellehrner
pushed a commit
to daniellehrner/besu
that referenced
this pull request
Jul 16, 2024
…ns (hyperledger#7271) * Add info-level diagnostic logs to aid with resolving stalled BFT chains Signed-off-by: Matthew Whitehead <[email protected]> * Add javadoc Signed-off-by: Matthew Whitehead <[email protected]> --------- Signed-off-by: Matthew Whitehead <[email protected]> Signed-off-by: Matt Whitehead <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]>
daniellehrner
added a commit
that referenced
this pull request
Jul 17, 2024
* EIP 7702 first draft Signed-off-by: Daniel Lehrner <[email protected]> * added CHANGELOG.md entry Signed-off-by: Daniel Lehrner <[email protected]> * bug fixes, added first tests Signed-off-by: Daniel Lehrner <[email protected]> * container verify GitHub workflow (#7239) Container verification step in release process automated with the container verify GitHub workflow. New workflow is triggered at the end of the release workflow which will check the release container images starts successfully. Verification test only checks container starts and reach the Ethereum main loop Signed-off-by: Chaminda Divitotawela <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Investigate chain halts when syncing (#7162) Fix some reasons for chain download halts when syncing Signed-off-by: [email protected] <[email protected]> Signed-off-by: Stefan Pingel <[email protected]> Co-authored-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Check for EOFCreate subcontainer rules (#7232) Check and test for the unused container rule, and only returncontract targets can have truncated data rule. Also test the other subcontainer rules in unit tests. Signed-off-by: Danno Ferrin <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Remove deprecation message for `--Xp2p-peer-lower-bound` (#7247) Signed-off-by: Gabriel-Trintinalia <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * less invasive code injection approach Signed-off-by: Daniel Lehrner <[email protected]> * added missing java doc & fixed test Signed-off-by: Daniel Lehrner <[email protected]> * added (currently non-working) acceptance test, some bug fixes in the transaction validation and tx pool logic Signed-off-by: Daniel Lehrner <[email protected]> * fix spotless Signed-off-by: Daniel Lehrner <[email protected]> * updated acceptance test, still not working, newPayload request seems to be necessary before final fork choice update Signed-off-by: Daniel Lehrner <[email protected]> * use correct world state to inject temporary code, inject code in existing accounts as well Signed-off-by: Daniel Lehrner <[email protected]> * renamed test service to prague, because the engine versions used are only available in the prague hard fork Signed-off-by: Daniel Lehrner <[email protected]> * fixed acceptance test, some bug fixes if authorized account does not yet exist Signed-off-by: Daniel Lehrner <[email protected]> * Add build version option to prefix git hash with custom version property (#7222) * Add build version option to prefix git hash with custom version property * Refactor to make appending the git hash a boolean property. Include a commented-out example of how to use the properties in the gradle file Signed-off-by: Matthew Whitehead <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Handle invalid snap getTrieNode requests with empty paths gracefully (#7221) Signed-off-by: Jason Frame <[email protected]> Co-authored-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * fix typos in CHANGELOG (#7226) Signed-off-by: Ties <[email protected]> Co-authored-by: Matt Nelson <[email protected]> Co-authored-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * feat: Add network option for LUKSO Mainnet (#7223) * Add option for LUKSO network Signed-off-by: Wolmin <[email protected]> * Add tests for LUKSO Signed-off-by: Wolmin <[email protected]> * Apply spotless Signed-off-by: Wolmin <[email protected]> * Add changelog entry Signed-off-by: Wolmin <[email protected]> * Fix duplicate func Signed-off-by: Wolmin <[email protected]> * Fix changelog Signed-off-by: Wolmin <[email protected]> * Add bootnodes to genesis Signed-off-by: Wolmin <[email protected]> --------- Signed-off-by: Wolmin <[email protected]> Signed-off-by: Wolmin <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Update Docker base image to Ubuntu 24.04 (#7251) Signed-off-by: Fabio Di Fabio <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Reconfigure how Protocol Specs are created WRT EVM condiguration (#7245) Make the max code size and max initcode size a part of the EVM configuration. As part of the change we need to move the tasks CodeFactory once handled as a static class and move it into the EVM. This has a nice follow on effect that we don't need to pass in max EOF versions or max code sizes anymore. Signed-off-by: Danno Ferrin <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Fix the wrong 'Identifier' and 'Synchronizer' usage (#7252) * fix the synchronizer usage Signed-off-by: Leni <[email protected]> * fix Identifier usage Signed-off-by: Leni <[email protected]> --------- Signed-off-by: Leni <[email protected]> Co-authored-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Fix flaky SECP256R1 test (#7249) Signed-off-by: Daniel Lehrner <[email protected]> Co-authored-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * update to work with the new max retries value (#7253) Signed-off-by: Justin Florentine <[email protected]> Co-authored-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Temporary CancunEOF fork for EOF testing. (#7227) Add Genesis ("CancunEOFTime") and reference test ("CancunEOF") support for a temporary Cancun+EOF fork, in anticipation of potential devnets. Signed-off-by: Danno Ferrin <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * fixed bug introduced through merge of main, made acceptance test easier to understand Signed-off-by: Daniel Lehrner <[email protected]> * added missing java docs Signed-off-by: Daniel Lehrner <[email protected]> * removed unnecessary tag Signed-off-by: Daniel Lehrner <[email protected]> * make encodeSingleSetCode public again Signed-off-by: Daniel Lehrner <[email protected]> * Snapserver responses to return at least one response (#7190) Signed-off-by: Jason Frame <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * copy setCodeTransactionPayloads as well Signed-off-by: Daniel Lehrner <[email protected]> * fixed bug during tests with forrest db Signed-off-by: Daniel Lehrner <[email protected]> * Snapserver GetTrieNodes request to handle short hash for storage (#7264) Signed-off-by: Jason Frame <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * javadoc: Adding javadoc for ethstats module (#7269) * javadoc: Adding javadoc for ethstats module --------- Signed-off-by: Usman Saleem <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Fix javadoc for ethereum:core top level package (#7270) * javadoc - Apply javadoc to ethereum core top level package --------- Signed-off-by: Usman Saleem <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Disable Flaky tests - permissioning (#7256) * disable some flaky tests Signed-off-by: Sally MacFarlane <[email protected]> * correct name for test Signed-off-by: Sally MacFarlane <[email protected]> * formatting Signed-off-by: Sally MacFarlane <[email protected]> * disable some flaky tests Signed-off-by: Sally MacFarlane <[email protected]> --------- Signed-off-by: Sally MacFarlane <[email protected]> Co-authored-by: Usman Saleem <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Add bootnodes to the maintained peer list (#7257) * Add bootnodes to the maintained peer list Signed-off-by: Matthew Whitehead <[email protected]> * Update unit tests Signed-off-by: Matthew Whitehead <[email protected]> * Add entry in changelog Signed-off-by: Matthew Whitehead <[email protected]> * Tweak unit test Signed-off-by: Matthew Whitehead <[email protected]> * Refactor to keep common steps the same for both cases Signed-off-by: Matthew Whitehead <[email protected]> * Add debug log, call sanitizePeers() only once Signed-off-by: Matthew Whitehead <[email protected]> --------- Signed-off-by: Matthew Whitehead <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Fix javadoc for ethereum api module, graphql package (#7272) * javadoc - Adding missing javadocs ethereum:api graphql package Signed-off-by: Usman Saleem <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * T8n support for isStateTest and empty accounts (#7275) Update t8n executor to support new isStateTest env flag that will disable extra-transactional processing such as block rewards and beacon root. Also, make sure such extra-transactional commits don't create empty accounts. Signed-off-by: Danno Ferrin <[email protected]> Co-authored-by: Usman Saleem <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Promote storage x-trie-log subcommand to trie-log (#7278) Signed-off-by: Simon Dudley <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Evm tool readme update (#7274) * update paths to binary. update docker build to use java 21 * updated suggested jdk --------- Signed-off-by: Justin Florentine <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * javadoc - Add missing javadoc for evmtool module (#7277) Signed-off-by: Usman Saleem <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Rename ValidatorPublicKey to ValidatorPubKey (#7280) Adapt to EIP-7002 name change for validator public key in all places. Signed-off-by: Danno Ferrin <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Add info-level diagnostic logs to aid with resolving stalled BFT chains (#7271) * Add info-level diagnostic logs to aid with resolving stalled BFT chains Signed-off-by: Matthew Whitehead <[email protected]> * Add javadoc Signed-off-by: Matthew Whitehead <[email protected]> --------- Signed-off-by: Matthew Whitehead <[email protected]> Signed-off-by: Matt Whitehead <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Update EIP-2935 contract (#7281) Use the updated contract and address for EIP-2539. Signed-off-by: Danno Ferrin <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * add evmtool compability, fixing bugs related to sender recovery of 7702 txs and handling authorizations to empty accounts Signed-off-by: Daniel Lehrner <[email protected]> * Deeper tracing of self-destructed accounts (#7284) Consider previously self-destructed accounts when creating accounts. Signed-off-by: Danno Ferrin <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * next release version after 24.7.0 (#7285) Signed-off-by: garyschulte <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Add experimental `--Xsnapsync-bft-enabled` which enables snap sync for BFT chains (#7140) * Create a BFT-specific pivot block handler Signed-off-by: Matthew Whitehead <[email protected]> * Change visibility Signed-off-by: Matthew Whitehead <[email protected]> * Refactor sync-peer-count internal variable to match name, add experimental flag to enabled snap + BFT Signed-off-by: Matthew Whitehead <[email protected]> * Merge with main Signed-off-by: Matthew Whitehead <[email protected]> * Fix uppercase Signed-off-by: Matthew Whitehead <[email protected]> * Address synchronization issue with trie pruning. Create BFT-specific account range handler. Add pipeline name and logs Signed-off-by: Matthew Whitehead <[email protected]> * Remove debug log Signed-off-by: Matthew Whitehead <[email protected]> * fixing snapsync for empty state Signed-off-by: Karim Taam <[email protected]> * Don't queue up events we can't handle Signed-off-by: Matthew Whitehead <[email protected]> * Fix timing window where a validator with an empty data dir sometimes falls back to full sync if peer status isn't received quickly enough Signed-off-by: Matthew Whitehead <[email protected]> * Remove BFT-specific account request class. Not needed Signed-off-by: Matthew Whitehead <[email protected]> * Refactor some more 'fast' sync variables that are common to all pivot-based sync types Signed-off-by: Matthew Whitehead <[email protected]> * In FULL sync mode, disable bonsai-limit-trie-logs-enabled instead of failing to start Signed-off-by: Matthew Whitehead <[email protected]> * Add javadoc comments, clarify overriding bonsai-limit-trie-logs-enabled Signed-off-by: Matthew Whitehead <[email protected]> * Add BFT pivot block selector tests Signed-off-by: Matthew Whitehead <[email protected]> * Fix failure error message Signed-off-by: Matthew Whitehead <[email protected]> * Remove the unnamed Pipe constructor and update tests to set a pipe name Signed-off-by: Matthew Whitehead <[email protected]> * Revert some info logs back to debug given the feedback on noise in the logs syncing with holesky Signed-off-by: Matthew Whitehead <[email protected]> * Refactor fastSyncPivotDistance to syncPivotDistance Signed-off-by: Matthew Whitehead <[email protected]> * Incomplete refactoring Signed-off-by: Matthew Whitehead <[email protected]> * Update BFT event queueing tests Signed-off-by: Matthew Whitehead <[email protected]> * Event queue test fixes Signed-off-by: Matthew Whitehead <[email protected]> * Remove automatic setting of bonsai-limit-trie-logs-enabled to false if sync-mode = FULL (moving to another PR) Signed-off-by: Matthew Whitehead <[email protected]> --------- Signed-off-by: Matthew Whitehead <[email protected]> Signed-off-by: Karim Taam <[email protected]> Signed-off-by: Matt Whitehead <[email protected]> Co-authored-by: Karim Taam <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Turn off CicleCI for Besu (#7291) All the CI jobs run in GitHub actions and Circle CI it no longer needed in Besu project Signed-off-by: Chaminda Divitotawela <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Check for snap server (#6609) * EthPeer add isServingSnap to be able to make sure that we have enough snap servers connected when we are snap syncing Signed-off-by: [email protected] <[email protected]> Signed-off-by: Sally MacFarlane <[email protected]> Co-authored-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Implement System Calls (#7263) Signed-off-by: Gabriel-Trintinalia <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * wrap WorldUpdater inside a WorldUpdaterService to inject the authorized code whenever needed Signed-off-by: Daniel Lehrner <[email protected]> * Update limit trie logs validation message for sync-mode=FULL (#7279) Signed-off-by: Simon Dudley <[email protected]> Co-authored-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Execute requests before block persist (#7295) Signed-off-by: Gabriel-Trintinalia <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * fixed MainnetTransactionProcessor retrieval of correctn `to` account with injected code, fixed code injection Signed-off-by: Daniel Lehrner <[email protected]> * only first authorization is accepted, all the following ones are ignored Signed-off-by: Daniel Lehrner <[email protected]> * don't cache account with empty code Signed-off-by: Daniel Lehrner <[email protected]> * revert wrapping of world updater, as its `updater()` method creates a type of nesting that is not compatible with wrapping it. Instead a service is injected in the world updater to inject the code into the authorized accounts Signed-off-by: Daniel Lehrner <[email protected]> * Fixed outdated tech redirect link. (#7297) * fix wiki link Signed-off-by: Snazzy <[email protected]> * fix format Signed-off-by: Snazzy <[email protected]> * change knownHash Signed-off-by: Snazzy <[email protected]> --------- Signed-off-by: Snazzy <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Increment private nonce even if transaction failed. (#6593) Increment private nonce even if transaction failed Signed-off-by: George Tebrean <[email protected]> Signed-off-by: [email protected] <[email protected]> Co-authored-by: [email protected] <[email protected]> Co-authored-by: Stefan Pingel <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * feat: Enhance --profile to load external profiles (#7292) * feat: --profile can load external profiles * fix external profile name method * fix ProfilesCompletionCandidate * test: Add unit tests * changelog: Update changelog * test: Fix TomlConfigurationDefaultProviderTest * test: Fix BesuCommandTest --------- Signed-off-by: Usman Saleem <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Fix status badge for documentation (#7304) Documentation has been moved to GitHub pages and no longer use readthedocs. Updated the README status badge for docs with correct link Signed-off-by: Chaminda Divitotawela <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * [MINOR] Fixed some typos (#7299) * typos Signed-off-by: Sally MacFarlane <[email protected]> --------- Signed-off-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * refactored to share one AuthorizedAccountService between the different instances of the world updater, renamed some classes Signed-off-by: Daniel Lehrner <[email protected]> * spotless Signed-off-by: Daniel Lehrner <[email protected]> * load code for authorization at the beginning of the transaction Signed-off-by: Daniel Lehrner <[email protected]> * ignore authorization if chain id doesn't match Signed-off-by: Daniel Lehrner <[email protected]> * cache authority address, evmtool: do not fail if sender address is wrong Signed-off-by: Daniel Lehrner <[email protected]> * Add evmtool block-test subcommand (#7293) * Add evmtool block-test subcommand Add an evmtool subcommand that will run non-hive blockchain tests. Signed-off-by: Danno Ferrin <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Make the retrying snap tasks switching (#7307) * make snap tasks switching Signed-off-by: [email protected] <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * 6612: Remove deprecated sync modes and related helper methods (#7309) * 6612: Remove deprecated sync modes and related helper methods Signed-off-by: Matilda-Clerke <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * EOF Reference Test Fixes (#7306) Fix a number of issues found in reference tests and evmone tests. - Be tolerant of more nulls in json - Support ContainerKind in reference tests - re-order EXTCALL oeprands - correct return value for REVERT in EXT*CALL - re-order EOFCREATE code validation Signed-off-by: Danno Ferrin <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * test template refactor, bump besu-native to 0.9.2 (#7315) Signed-off-by: garyschulte <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Feature/use gnark-crypto for eip-2537 (#7316) * use gnark-crypto for bls precompiles Signed-off-by: garyschulte <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * 6612 update changelog with removed syncmodes (#7320) * 6612: Update changelog with removal of deprecated sync modes Signed-off-by: Matilda Clerke <[email protected]> * 6612: Update changelog with removal of deprecated sync modes Signed-off-by: Matilda Clerke <[email protected]> * 6612: Update changelog with removal of deprecated sync modes Signed-off-by: Matilda Clerke <[email protected]> --------- Signed-off-by: Matilda Clerke <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Update datacopy (#7319) Check for OOG earlier in DataCopy. Add unit tests to cover operation branches. Signed-off-by: Danno Ferrin <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * disable flaky test (#7308) * disable flaky test Signed-off-by: Sally MacFarlane <[email protected]> * disable flaky test Signed-off-by: Sally MacFarlane <[email protected]> --------- Signed-off-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Update unit test (#7317) * Update parameterized unit tests so the enumerate with --dry-run * Update the prague-withdrawal.json unit test to handle current code Signed-off-by: Danno Ferrin <[email protected]> Co-authored-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * persist accounts that have storage updates, but no nonce, balance nor code Signed-off-by: Daniel Lehrner <[email protected]> * Revert "persist accounts that have storage updates, but no nonce, balance nor code" This reverts commit 9c9121a. Signed-off-by: Daniel Lehrner <[email protected]> * removed PKI backed QBFT (#7310) * removed PKI backed QBFT Signed-off-by: Sally MacFarlane <[email protected]> * changelog Signed-off-by: Sally MacFarlane <[email protected]> --------- Signed-off-by: Sally MacFarlane <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * EIP-7251 add consolidation request type (#7266) * add request type for consolidations, encoder, decoder and tests * added raw tx for consolidation * add consolidation reqs to EngineGetPayloadResultV4 * set storage slot value to 0 initially and value for tx * updates plugin api Signed-off-by: Justin Florentine <[email protected]> Signed-off-by: Sally MacFarlane <[email protected]> --------- Signed-off-by: Sally MacFarlane <[email protected]> Signed-off-by: Justin Florentine <[email protected]> Co-authored-by: Justin Florentine <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * fix: `eth_call` deserialization to correctly ignore unknown fields in the transaction object (#7323) * fix: Use Builder for JsonCallParameter * changelog * add additional unit tests * fix: Update builder to withGas to match the json eth_call --------- Signed-off-by: Usman Saleem <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * Stop transaction selection on TX_EVALUATION_TOO_LONG (#7330) Signed-off-by: Fabio Di Fabio <[email protected]> Signed-off-by: Daniel Lehrner <[email protected]> * message frame buider will create AuthorizedCodeService by itsef if it isn't set Signed-off-by: Daniel Lehrner <[email protected]> * get correct nonce for authorization Signed-off-by: Daniel Lehrner <[email protected]> * nonce only returns a vaid nonce, new method nonceList returns all the nonces Signed-off-by: Daniel Lehrner <[email protected]> * plugs leaky abstraction Signed-off-by: Justin Florentine <[email protected]> * some renaming, acceptance tests checks for exact balance of tx sponsor at the end. Signed-off-by: Daniel Lehrner <[email protected]> * inject the 7702 code into DiffBasedWorldStateUpdateAccumulator.createAccount Signed-off-by: Daniel Lehrner <[email protected]> * spotless Signed-off-by: Justin Florentine <[email protected]> * spotless fix, removed todos Signed-off-by: Daniel Lehrner <[email protected]> * make AuthorityProcessor & chain id for it optional Signed-off-by: Daniel Lehrner <[email protected]> --------- Signed-off-by: Daniel Lehrner <[email protected]> Signed-off-by: Chaminda Divitotawela <[email protected]> Signed-off-by: [email protected] <[email protected]> Signed-off-by: Stefan Pingel <[email protected]> Signed-off-by: Danno Ferrin <[email protected]> Signed-off-by: Gabriel-Trintinalia <[email protected]> Signed-off-by: Matthew Whitehead <[email protected]> Signed-off-by: Jason Frame <[email protected]> Signed-off-by: Ties <[email protected]> Signed-off-by: Wolmin <[email protected]> Signed-off-by: Wolmin <[email protected]> Signed-off-by: Fabio Di Fabio <[email protected]> Signed-off-by: Leni <[email protected]> Signed-off-by: Justin Florentine <[email protected]> Signed-off-by: Usman Saleem <[email protected]> Signed-off-by: Sally MacFarlane <[email protected]> Signed-off-by: Simon Dudley <[email protected]> Signed-off-by: Matt Whitehead <[email protected]> Signed-off-by: garyschulte <[email protected]> Signed-off-by: Karim Taam <[email protected]> Signed-off-by: Snazzy <[email protected]> Signed-off-by: George Tebrean <[email protected]> Signed-off-by: Matilda-Clerke <[email protected]> Signed-off-by: Matilda Clerke <[email protected]> Co-authored-by: Chaminda Divitotawela <[email protected]> Co-authored-by: Stefan Pingel <[email protected]> Co-authored-by: Sally MacFarlane <[email protected]> Co-authored-by: Danno Ferrin <[email protected]> Co-authored-by: Gabriel-Trintinalia <[email protected]> Co-authored-by: Matt Whitehead <[email protected]> Co-authored-by: Jason Frame <[email protected]> Co-authored-by: Ties <[email protected]> Co-authored-by: Matt Nelson <[email protected]> Co-authored-by: Wolmin <[email protected]> Co-authored-by: Fabio Di Fabio <[email protected]> Co-authored-by: leniram159 <[email protected]> Co-authored-by: Justin Florentine <[email protected]> Co-authored-by: Usman Saleem <[email protected]> Co-authored-by: Simon Dudley <[email protected]> Co-authored-by: garyschulte <[email protected]> Co-authored-by: Karim Taam <[email protected]> Co-authored-by: gringsam <[email protected]> Co-authored-by: George Tebrean <[email protected]> Co-authored-by: [email protected] <[email protected]> Co-authored-by: Matilda-Clerke <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR description
This PR adds
INFOlevel diagnostic logs to BFT nodes if (and only if) the round for a given block reaches 2 or higher:It also logs an
INFOmessage when the local node expires a round, clearly indicating when the next round will expire, again only if the node reaches round 2 or higher:Detail
When a BFT chain stalls, the logs are very quiet. Below is an example of a 4-validator chain where 2 validators go offline for 2 minutes and then come back online (note the 4-minute absence of any logs in between blocks
31and32):When this happens it is often necessary to look at
DEBUGorTRACElogs, sometimes on several of the nodes, to determine which round each node has reached and understand if/when the chain will begin producing new blocks.Reviewers may be concerned about the use of
INFOlogs for these diagnostic messages. My contention is that it is probably reasonable to useINFOmessages in this case for 3 reasons:INFOlogs are extremely quiet in a stalled-chain case. Giving a summary of where the round-changes are without needing to enableDEBUGorTRACEis a very useful indicator of whether the user needs to investigate further (e.g. if round changes aren't being received from other nodes) or whether the chain is just waiting forquorumnodes to reach the same round. In the cases where blocks are being produced successfully, none of the extra log messages are produced.INFOlogs in this case isn't obfuscating any otherINFOlogs.Below is an example of the same 4-validator chain during the same 2-minute stall. Note how once the chain begins producing blocks the new diagnostic messages cease to be written:
I did consider concatenating the summary list into a single line, and I accept that reviewers might prefer the PR to use that approach instead of the multi-line logs I've gone with. I tend to prefer logs that are easy for the user to consume by eye and I think the example above shows how easy it is to scan the current round-change status and determine if they need to do any more investigation. But I welcome any comments on the format/verbosity.