Skip to main content

Release Notes

v1.7.17 - Verbier (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged. The supported plugin version is 15.

VMs

  • Refactored P-chain block state management
    • Supporting easier parsing and usage of blocks
    • Improving separation of block execution with block definition
    • Unifying state definitions
  • Introduced support to send custom X-chain assets to the P-chain to be activated in a future upgrade
  • Introduced support to use custom assets on the P-chain to be activated in a future upgrade
  • Added VMs README to begin fully documenting plugin invariants
  • Added various comments around expected usages of VM tools

Coreth

  • Added optional JSON logging
  • Added interface for supporting stateful precompiles
  • Removed legacy code format from the database

Fixes

  • Fixed ungraceful gRPC connection closure during very long running requests
  • Fixed LevelDB panic during shutdown
  • Fixed verification of --stake-max-consumption-rate to include the upper-bound
  • Fixed various CI failures
  • Fixed flaky unit tests

Miscellaneous

  • Added bootstrapping ETA metrics
  • Converted all logs to support structured fields
  • Improved Snowman++ oracle block verification error messages
  • Removed deprecated or unused scripts

v1.7.16 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged. The supported plugin version is 15.

LevelDB

  • Fix rapid disk growth by manually specifying the maximum manifest file size

v1.7.15 (View on GitHub)

danger

This version should not be used in a production environment. v1.7.16 resolves a database leak introduced in this release.

This version is backwards compatible to v1.7.0. It is optional, but encouraged. The supported plugin version is 15.

PlatformVM

  • Replaced copy-on-write validator set data-structure to use tree diffs to optimize validator set additions
  • Replaced validation transactions with a standardized representation to remove transaction type handling
  • Migrated transaction execution to its own package
  • Removed child pointers from processing blocks
  • Added P-chain wallet helper for providing initial transactions

Coreth

  • Bumped go-ethereum dependency to v1.10.20
  • Updated API names used to enable services in eth-api config flag. Prior names are supported but deprecated, please update configurations accordingly
  • Optimized state sync by parallelizing trie syncing
  • Added eth_syncing API for compatibility. Note: This API is only accessible after bootstrapping and always returns "false", since the node will no longer be syncing at that point
  • Added metrics to the atomic transaction mempool
  • Added metrics for incoming/outgoing mempool gossip

Fixes

  • Updated Snowman and Avalanche consensus engines to report original container preferences before processing the provided container
  • Fixed inbound message byte throttler context cancellation cleanup
  • Removed case sensitivity of IP resolver services
  • Added failing health check when a whitelisted Subnet fails to initialize a chain

Miscellaneous

  • Added gRPC client metrics for dynamically created connections
  • Added uninitialized continuous time averager for when initial predictions are unreliable
  • Updated linter version
  • Documented various platform invariants
  • Cleaned up various dead parameters
  • Improved various tests

v1.7.14 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

APIs

danger

These API format changes are breaking changes. https://api.avax.network and https://api.avax-test.network have been updated with this format. If you are using AvalancheGo APIs in your code, please ensure you have updated to the latest versions. See this doc for details about the CB58 removal.

  • Removed CB58 as an encoding option from all APIs
  • Added HexC and HexNC as encoding options for all APIs that accept an encoding format
  • Removed the Success response from all APIs
  • Replaced containerID with id in the indexer API

PlatformVM

  • Fixed incorrect P-chain height in Snowman++ when staking is disabled
  • Moved platformvm transactions to be defined in a sub-package
  • Moved platformvm genesis management to be defined in a sub-package
  • Moved platformvm state to be defined in a sub-package
  • Standardized platformvm transactions to always be referenced via pointer
  • Moved the platformvm transaction builder to be defined in a sub-package
  • Fixed uptime rounding during node shutdown

Networking

  • Updated Connected and Disconnected messages to only be sent to chains if the peer is tracking the Subnet
  • Updated the minimum TLS version on the p2p network to v1.3
  • Supported context cancellation in the networking rate limiters
  • Added ChitsV2 message format for the p2p network to be used in a future upgrade

Miscellaneous

  • Fixed --public-ip-resolution-frequency invalid overwrite of the resolution service
  • Added additional metrics to distinguish between virtuous and rogue currently processing transactions
  • Suppressed the super cool avalanchego banner when stdout is not directed to a terminal
  • Updated linter version
  • Improved various comments and documentation
  • Standardized primary network handling across Subnet maps

v1.7.13 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

State Sync v0.1 (v1.7.11) and v0.2 (v1.7.12) are not compatible with State Sync v0.3 (v1.7.13). When the majority of stake upgrades, old clients will no longer be able to State Sync. In addition, v1.7.13 modifies the VM interface, so you'll need to update your Custom VM dependency.

State Sync

  • Added peer bandwidth tracking to optimize coreth state sync message routing
  • Fixed coreth leaf request handler bug to ensure the handler delivers a valid range proof
  • Removed redundant proof keys from coreth leafs response message format
  • Improved coreth state sync request retry logic
  • Improved coreth state sync handler metrics
  • Improved coreth state sync ETA
  • Added avalanche_{chainID}_handler_async_expired metric

Miscellaneous

  • Fixed platform.getCurrentValidators API to correctly mark a node as connected to itself on Subnets.
  • Fixed platform.getBlockchainStatus to correctly report Unknown for blockchains that are not managed by the P-Chain
  • Added process metrics by default in the rpcchainvm#Server
  • Added Database health checks
  • Removed the deprecated Database.Stat call from the rpcdb#Server
  • Added fail fast logic to duplicated Snowman additions to avoid undefined behavior
  • Added additional testing around Snowman diverged voting tests
  • Deprecated --dynamic-update-duration and --dynamic-public-ip CLI flags
  • Added --public-ip-resolution-frequency and --public-ip-resolution-service to replace --dynamic-update-duration and --dynamic-public-ip, respectively

v1.7.12 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

State Sync

  • Fixed proposervm state summary acceptance to only accept state summaries with heights higher than the locally last accepted block
  • Fixed proposervm state summary serving to only respond to requests after height indexing has finished
  • Improved C-chain state sync leaf request serving by optimistically reading leaves from snapshot
  • Refactored C-chain state sync block fetching

Networking

  • Reduced default peerlist and accepted frontier gossipping
  • Increased the default at-large outbound buffer size to 32 MiB

Metrics

  • Added leveldb metrics
  • Added process and golang metrics for the avalanchego binary
  • Added available disk space health check
    • Ensured that the disk space will not be fully utilized by shutting down the node if there is a critically low amount of free space remaining
  • Improved C-chain state sync metrics

Performance

  • Added C-chain acceptor queue within core/blockchain.go
  • Removed rpcdb locking when committing batches and using iterators
  • Capped C-chain TrieDB dirties cache size during block acceptance to reduce commit size at 4096 block interval

Cleanup

  • Refactored the avm to utilize the external txs package
  • Unified platformvm dropped tx handling
  • Clarified snowman child block acceptance calls
  • Fixed small consensus typos
  • Reduced minor duplicated code in consensus
  • Moved the platformvm key factory out of the VM into the test file
  • Removed unused return values from the timeout manager
  • Removed weird json rpc private interface
  • Standardized json imports
  • Added vm factory interface checks

v1.7.11 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

The first startup of the C-Chain will cause an increase in CPU and IO usage due to an index update. This index update runs in the background and does not impact restart time.

State Sync

  • Added state syncer engine to facilitate VM state syncing, rather than full historical syncing
  • Added GetStateSummaryFrontier, StateSummaryFrontier, GetAcceptedStateSummary, AcceptedStateSummary as P2P messages
  • Updated Ancestors message specification to expect an empty response if the container is unknown
  • Added --state-sync-ips and --state-sync-ids flags to allow manual overrides of which nodes to query for accepted state summaries
  • Updated networking library to permanently track all manually tracked peers, rather than just beacons
  • Added state sync support to the metervm
  • Added state sync support to the proposervm
  • Added state sync support to the rpcchainvm
  • Added beta state sync support to coreth

ProposerVM

  • Prevented rejected blocks from overwriting the proposervm height index
  • Optimized proposervm block rewind to utilize the height index if available
  • Ensured proposervm height index is marked as repaired in Initialize if it is fully repaired on startup
  • Removed --reset-proposervm-height-index. The height index will be reset upon first restart
  • Optimized proposervm height index resetting to periodically flush deletions

Bug Fixes

  • Fixed IPC message issuance and restructured consensus event callbacks to be checked at compile time
  • Fixed coreth metrics initialization
  • Fixed bootstrapping startup logic to correctly startup if initially connected to enough stake
  • Fixed coreth panic during metrics collection
  • Fixed panic on concurrent map read/write in P-chain wallet SDK
  • Fixed rpcchainvm panic by sanitizing http response codes
  • Fixed incorrect JSON tag on platformvm.BaseTx
  • Fixed AppRequest, AppResponse, and AppGossip stringers used in logging

API/Client

  • Supported client implementations pointing to non-standard URIs
  • Introduced ids.NodeID type to standardize logging and simplify API service and client implementations
  • Changed client implementations to use standard types rather than strings wherever possible
  • Added subnetID as an argument to platform.getTotalStake
  • Added connected to the Subnet validators in responses to platform.getCurrentValidators and platform.getPendingValidators
  • Add missing admin API client methods
  • Improved indexer API client implementation to avoid encoding edge cases

Networking

  • Added --snow-mixed-query-num-push-vdr and --snow-mixed-query-num-push-non-vdr to allow parameterization of sending push queries
    • By default, non-validators now send only pull queries, not push queries.
    • By default, validators now send both pull queries and push queries upon inserting a container into consensus. Previously, nodes sent only push queries.
  • Added metrics to track the amount of over gossipping of peerlist messages
  • Added custom message queueing support to outbound Peer messages
  • Reused Ping messages to avoid needless memory allocations

Logging

  • Replaced AvalancheGo's internal logger with uber-go/zap.
  • Replaced AvalancheGo's log rotation with lumberjack.
  • Renamed log-display-highlight to log-format and added json option.
  • Added log-rotater-max-size, log-rotater-max-files, log-rotater-max-age, log-rotater-compress-enabled options for log rotation.

Miscellaneous

  • Standardized RPC specification of timestamp fields
  • Logged health checks whenever a failing health check is queried
  • Added callback support for the validator set manager
  • Increased coreth trie tip buffer size to 32
  • Added CPU usage metrics for AvalancheGo and all sub-processes
  • Added Disk IO usage metrics for AvalancheGo and all sub-processes

Cleanup

  • Refactored easily separable platformvm files into separate smaller packages
  • Simplified default version parsing
  • Fixed various typos
  • Converted some structs to interfaces to better support mocked testing
  • Refactored IP utils

Documentation

  • Increased recommended disk size to 1 TB
  • Updated issue template
  • Documented additional snowman.Block invariants

v1.7.10 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

Networking

  • Improved vertex and block gossiping for validators with low stake weight.
  • Added peers metric by Subnet.
  • Added percentage of stake connected metric by Subnet.

APIs

  • Added support for specifying additional headers and query params in the RPC client implementations.
  • Added static API clients for the platformvm and the avm.

PlatformVM

  • Introduced time based windowing of accepted P-chain block heights to ensure that local networks update the proposer list timely in the proposervm.
  • Improved selection of decision transactions from the mempool.

RPCChainVM

  • Increased buf version to v1.3.1.
  • Migrated all proto definitions to a dedicated /proto folder.
  • Removed the dependency on the non-standard grpc broker to better support other language implementations.
  • Added grpc metrics.
  • Added grpc server health checks.

Coreth

  • Fixed a bug where a deadlock on shutdown caused historical re-generation on restart.
  • Added an API endpoint to fetch the current VM Config.
  • Added AvalancheGo custom log formatting to the logs.
  • Removed support for the JS Tracer.

Logging

  • Added piping of Subnet logs to stdout.
  • Lazily initialized logs to avoid opening files that are never written to.
  • Added support for arbitrarily deleted log files while avalanchego is running.
  • Removed redundant logging configs.

Miscellaneous

  • Updated minimum go version to v1.17.9.
  • Added Subnet bootstrapping health checks.
  • Supported multiple tags per codec instantiation.
  • Added minor fail-fast optimization to string packing.
  • Removed dead code.
  • Fixed typos.
  • Simplified consensus engine Shutdown notification dispatching.
  • Removed Sleep call in the inbound connection throttler.

v1.7.9 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

Updates

  • Improved Subnet gossip to only send messages to nodes participating in that Subnet.
  • Fixed inlined VM initialization to correctly register static APIs.
  • Added logging for file descriptor limit errors.
  • Removed dead code from network packer.
  • Improved logging of invalid hash length errors.

v1.7.8 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

Networking

  • Fixed duplicate reference decrease when closing a peer.
  • Freed allocated message buffers immediately after sending.
  • Added --network-peer-read-buffer-size and --network-peer-write-buffer-size config options.
  • Moved peer IP signature verification to enable concurrent verifications.
  • Reduced the number of connection flushes when sending messages.
  • Canceled outbound connection requests on shutdown.
  • Reused dialer across multiple outbound connections.
  • Exported NewTestNetwork for easier external testing.

Coreth

  • Reduced log level of snapshot regeneration logs.
  • Enabled atomic tx replacement with higher gas fees.
  • Parallelized trie index re-generation.

Miscellaneous

  • Fixed incorrect BlockchainID usage in the X-chain ImportTx builder.
  • Fixed incorrect OutputOwners in the P-chain ImportTx builder.
  • Improved FD limit error logging and warnings.
  • Rounded bootstrapping ETAs to the nearest second.
  • Added gossip config support to the Subnet configs.
  • Optimized various queue removals for improved memory freeing.
  • Added a basic X-chain E2E usage test to the new testing framework.

v1.7.7 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

Networking

  • Refactored the networking library to track potential peers by nodeID rather than IP.
  • Separated peer connections from the mesh network implementation to simplify testing.
  • Fixed duplicate Connected messages bug.
  • Supported establishing outbound connections with peers reporting different inbound and outbound IPs.

Database

  • Disabled seek compaction in leveldb by default.

GRPC

  • Increased protocol version, this requires all plugin definitions to update their communication dependencies.
  • Merged services to be served using the same server when possible.
  • Implemented a fast path for simple HTTP requests.
  • Removed duplicated message definitions.
  • Improved error reporting around invalid plugins.

Coreth

  • Optimized FeeHistory API.
  • Added protection to prevent accidental corruption of archival node trie index.
  • Added capability to restore complete trie index on best effort basis.
  • Rounded up fastcache sizes to utilize all mmap'd memory in chunks of 64MB.

Configs

  • Removed --inbound-connection-throttling-max-recent
  • Renamed --network-peer-list-size to --network-peer-list-num-validator-ips
  • Removed --network-peer-list-gossip-size
  • Removed --network-peer-list-staker-gossip-fraction
  • Added --network-peer-list-validator-gossip-size
  • Added --network-peer-list-non-validator-gossip-size
  • Removed --network-get-version-timeout
  • Removed --benchlist-peer-summary-enabled
  • Removed --peer-alias-timeout

Miscellaneous

  • Fixed error reporting when making Avalanche chains that did not manually specify a primary alias.
  • Added beacon utils for easier programmatic handling of beacon nodes.
  • Resolved the default log directory on initialization to avoid additional error handling.
  • Added support to the chain state module to specify an arbitrary new accepted block.

v1.7.6 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

Consensus

  • Introduced a new vertex type to support future Avalanche based network upgrades.
  • Added pending message metrics to the chain message queues.
  • Refactored event dispatchers to simplify dependencies and remove dead code.

PlatformVM

  • Added json encoding option to the platform.getTx call.
  • Added platform.getBlock API.
  • Cleaned up block building logic to be more modular and testable.

Coreth

  • Increased FeeHistory maximum historical limit to improve MetaMask UI on the C-Chain.
  • Enabled chain state metrics.
  • Migrated go-ethereum v1.10.16 changes.

Miscellaneous

  • Added the ability to load new VM plugins dynamically.
  • Implemented X-chain + P-chain wallet that can be used to build and sign transactions. Without providing a full node private keys.
  • Integrated e2e testing to the repo to avoid maintaining multiple synced repos.
  • Fixed proposervm height indexing check to correctly mark the indexer as repaired.
  • Introduced message throttling overrides to be used in future improvements to reliably send messages.
  • Introduced a cap on the client specified request deadline.
  • Increased the default leveldb open files limit to 1024.
  • Documented the leveldb configurations.
  • Performed various cleanup passes.

v1.7.5 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

Consensus

  • Added asynchronous processing of App.* messages.
  • Added height indexing support to the proposervm and rpcchainvm. If a node is updated to >=v1.7.5 and then downgraded to <v1.7.5, the user must enable the --reset-proposervm-height-index=true flag to ensure the proposervm height index is correctly updated going forward.
  • Fixed bootstrapping job counter initialization that could cause negative ETAs to be reported.
  • Fixed incorrect processing check that could log incorrect information.
  • Removed incorrect warning logs.

Miscellaneous

  • Added tracked Subnets to be reported in calls to the info.peers API.
  • Updated gRPC implementations to use buf tooling and standardized naming and locations.
  • Added a consistent hashing implementation to be used in future improvements.
  • Fixed database iteration invariants to report ErrClosed rather than silently exiting.
  • Added additional sanity checks to prevent users from incorrectly configuring their node.
  • Updated log timestamps to include milliseconds.

Coreth

  • Added beta support for offline pruning.
  • Refactored peer networking layer.
  • Enabled cheap metrics by default.
  • Marked RPC call metrics as expensive.
  • Added Abigen support for native asset call precompile.
  • Fixed bug in BLOCKHASH opcode during traceBlock.
  • Fixed bug in handling updated chain config on startup.

v1.7.4 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

The first startup of the C-Chain will take a few minutes longer due to an index update.

Consensus

  • Removed deprecated Snowstorm consensus implementation that no longer aligned with the updated specification.
  • Updated bootstrapping logs to no longer reset counters after a node restart.
  • Added bootstrapping ETAs for fetching Snowman blocks and executing operations.
  • Renamed the MultiPut message to the Ancestors message to match other message naming conventions.
  • Introduced Whitelist conflicts into the Snowstorm specification that will be used in future X-chain improvements.
  • Refactored the separation between the Bootstrapping engine and the Consensus engine to support Fast-Sync.

Coreth

  • Added an index mapping height to the list of accepted atomic operations at that height in a trie. Generating this index will cause the node to take a few minutes longer to startup the C-Chain for the first restart.
  • Updated Geth dependency to v1.10.15.
  • Updated networkID to match chainID.

VMs

  • Refactored platformvm rewards calculations to enable usage from an external library.
  • Fixed platformvm and avm UTXO fetching to not re-iterate the UTXO set if no UTXOs are fetched.
  • Refactored platformvm status definitions.
  • Added support for multiple address balance lookups in the platformvm.
  • Refactored platformvm and avm keystore users to reuse similar code.

RPCChainVM

  • Returned a 500 InternalServerError if an unexpected gRPC error occurs during the handling of an HTTP request to a plugin.
  • Updated gRPC server's max message size to enable responses larger than 4MiB from the plugin's handling of an HTTP request.

Configs

  • Added --stake-max-consumption-rate which defaults to 120,000.
  • Added --stake-min-consumption-rate which defaults to 100,000.
  • Added --stake-supply-cap which defaults to 720,000,000,000,000,000 nAVAX.
  • Renamed --bootstrap-multiput-max-containers-sent to --bootstrap-ancestors-max-containers-sent.
  • Renamed --bootstrap-multiput-max-containers-received to --bootstrap-ancestors-max-containers-received.
  • Enforced that --staking-enabled=false can not be specified on public networks (Fuji and Mainnet).

Metrics

  • All multi_put metrics were converted to ancestors metrics.

Miscellaneous

  • Improved corruptabledb error reporting by tracking the first reported error.
  • Updated CPU tracking to use the proper EWMA tracker rather than a linear approximation.
  • Separated health checks into readiness, healthiness, and liveness checks to support more fine-grained monitoring.
  • Refactored API client utilities to use a Context rather than an explicit timeout.

v1.7.3 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

The first startup of the C-Chain will take a few minutes longer due to a minor index update.

Consensus

  • Introduced a notion of vertex conflicts that will be used in future X-chain improvements.

Coreth

  • Added an index mapping height to the list of accepted atomic transactions at that height. Generating this index will cause the node to take approximately 2 minutes longer to startup the C-Chain for the first restart.
  • Fixed bug in base fee estimation API that impacted custom defined networks.
  • Decreased minimum transaction re-gossiping interval from 1s to 500ms.
  • Removed websocket handler from the static vm APIs.

Database

  • Reduced lock contention in prefixDBs.

Networking

  • Increase the gossip size from 6 to 10 validators.
  • Prioritized Connected and Disconnected messages in the message handler.

Miscellaneous

  • Notified VMs of peer versions on Connected.
  • Fixed acceptance broadcasting over IPC.
  • Fixed 32-bit architecture builds for AvalancheGo (not Coreth).

v1.7.2 (View on GitHub)

This version is backwards compatible to v1.7.0. It is optional, but encouraged.

Coreth

  • Fixed memory leak in the estimate gas API.
  • Reduced the default RPC gas limit to 50,000,000 gas.
  • Improved RPC logging.
  • Removed pre-AP5 legacy code.

PlatformVM

  • Optimized validator set change calculations.
  • Removed storage of non-decided blocks.
  • Simplified error handling.
  • Removed pre-AP5 legacy code.

Networking

  • Explicitly fail requests with responses that failed to be parsed.
  • Removed pre-AP5 legacy code.

Configs

  • Introduced the ability for a delayed graceful node shutdown.
  • Added the ability to take all configs as environment variables for containerized deployments.

Utils

  • Fixed panic bug in logging library when importing from external projects.

v1.7.1 (View on GitHub)

This update is backwards compatible with v1.7.0. Please see the expected update times in the v1.7.0 release.

Coreth

  • Reduced fee estimate volatility.

Consensus

  • Fixed vote bubbling for unverified block chits.

v1.7.0 (View on GitHub)

This upgrade adds support for issuing multiple atomic transactions into a single block and directly transferring assets between the P-chain and the C-chain.

The changes in the upgrade go into effect at 1 PM EST, December 2nd 2021 on Mainnet. One should upgrade their node before the changes go into effect, otherwise they may experience loss of uptime.

All nodes should upgrade before 1 PM EST/6 PM UTC, December 2nd 2021.

Networking

  • Added peer uptime reports as metrics.
  • Removed IP rate limiting over local networks.

PlatformVM

  • Enabled AtomicTxs to be issued into StandardBlocks and deprecated AtomicBlocks.
  • Added the ability to export/import AVAX to/from the C-chain.

Coreth

  • Enabled multiple AtomicTxs to be issued per block.
  • Added the ability to export/import AVAX to/from the P-chain.
  • Updated dynamic fee calculations.

ProposerVM

  • Removed storage of undecided blocks.

RPCChainVM

  • Added support for metrics to be reported by plugin VMs.

Configs

  • Removed --snow-epoch-first-transition and snow-epoch-duration as command line arguments.

PRE_RELEASE v1.7.0-fuji (View on GitHub)

Please note that this release is unable to run mainnet - and will display "mainnet is not supported" if attempted to run with a mainnet configuration.

This upgrade adds support for issuing multiple atomic transactions into a single block and directly transferring assets between the P-chain and the C-chain.

The changes in the upgrade go into effect at 10 AM EST, November 24th 2021 on the Fuji testnet. After Fuji is updated and verified, a mainnet compatible release will be published.

All Fuji nodes should upgrade before 10 AM EST, November 24th 2021.

Networking

  • Added peer uptime reports as metrics.
  • Removed IP rate limiting over local networks.

PlatformVM

  • Enabled AtomicTxs to be issued into StandardBlocks and deprecated AtomicBlocks.
  • Added the ability to export/import AVAX to/from the C-chain.

Coreth

  • Enabled multiple AtomicTxs to be issued per block.
  • Added the ability to export/import AVAX to/from the P-chain.
  • Updated fee constants.

RPCChainVM

  • Added support for metrics to be reported by plugin VMs.

Configs

  • Removed --snow-epoch-first-transition and snow-epoch-duration as command line arguments.

v1.6.5 (View on GitHub)

This version is backwards compatible to v1.6.0. It is optional, but encouraged.

Bootstrapping

  • Drop inbound messages to a chain if that chain is in the execution phase of bootstrapping.
  • Print beacon nodeIDs upon failure to connect to them.

Metrics

  • Added avalanche_{ChainID}_bootstrap_finished, which is 1 if the chain is done bootstrapping, 0 otherwise.

APIs

  • Added info.uptime API call that attempts to report the network's view of the local node.
  • Added observedUptime to each peer's result in info.peers.

Network

  • Added reported uptime to pong messages to be able to better track a local node's uptime as viewed by the network.
  • Refactored request timeout registry to avoid a potential race condition.

v1.6.4 (View on GitHub)

This version is backwards compatible to v1.6.0. It is optional, but encouraged.

Config

  • Added flag throttler-inbound-bandwidth-refill-rate, which specifies the max average inbound bandwidth usage of a peer.
  • Added flag throttler-inbound-bandwidth-max-burst-size, which specifies the max inbound bandwidth usage of a peer.

Networking

  • Updated peerlist gossiping to use the same mechanism as other gossip calls.
  • Added inbound message throttling based on recent bandwidth usage.

Metrics

  • Updated avalanche_{ChainID}_handler_gossip_{count,sum} to avalanche_{ChainID}_handler_gossip_request_{count,sum}.
  • Updated avalanche_{ChainID}_lat_get_accepted_{count,sum} to avalanche_{ChainID}_lat_accepted_{count,sum}.
  • Updated avalanche_{ChainID}_lat_get_accepted_frontier_{count,sum} to avalanche_{ChainID}_lat_accepted_frontier_{count,sum}.
  • Updated avalanche_{ChainID}_lat_get_ancestors_{count,sum} to avalanche_{ChainID}_lat_multi_put_{count,sum}.
  • Combined avalanche_{ChainID}_lat_pull_query_{count,sum} and avalanche_{ChainID}_lat_push_query_{count,sum} to avalanche_{ChainID}_lat_chits_{count,sum}.
  • Added avalanche_{ChainID}_app_response_{count,sum}.
  • Added avalanche_network_bandwidth_throttler_inbound_acquire_latency_{count,sum}
  • Added avalanche_network_bandwidth_throttler_inbound_awaiting_acquire
  • Added avalanche_P_vm_votes_won
  • Added avalanche_P_vm_votes_lost

Indexer

  • Added method GetContainerByID to client implementation.
  • Client methods now return []byte rather than string representations of a container.

C-Chain

  • Updated Geth dependency to 1.10.11.
  • Added a new admin API for updating the log level and measuring performance.
  • Added a new --allow-unprotected-txs flag to allow issuance of transactions without EIP-155 replay protection.

Subnet & Custom VMs

  • Ensured that all possible chains are run in --staking-enabled=false networks.

v1.6.3 (View on GitHub)

This version is backwards compatible to v1.6.0. It is optional, but encouraged.

Config Options

  • Updated the default value of --inbound-connection-throttling-max-conns-per-sec to 256.
  • Updated the default value of --meter-vms-enabled to true.
  • Updated the default value of --staking-disabled-weight to 100.

Metrics

  • Changed the behavior of avalanche_network_buffer_throttler_inbound_awaiting_acquire to only increment if the message is actually blocking.
  • Changed the behavior of avalanche_network_byte_throttler_inbound_awaiting_acquire to only increment if the message is actually blocking.
  • Added Block/Tx metrics on meterVMs.
    • Added avalanche_{ChainID}_vm_metervm_build_block_err_{count,sum}.
    • Added avalanche_{ChainID}_vm_metervm_parse_block_err_{count,sum}.
    • Added avalanche_{ChainID}_vm_metervm_get_block_err_{count,sum}.
    • Added avalanche_{ChainID}_vm_metervm_verify_{count,sum}.
    • Added avalanche_{ChainID}_vm_metervm_verify_err_{count,sum}.
    • Added avalanche_{ChainID}_vm_metervm_accept_{count,sum}.
    • Added avalanche_{ChainID}_vm_metervm_reject_{count,sum}.
    • Added avalanche_{DAGID}_vm_metervm_parse_tx_err_{count,sum}.
    • Added avalanche_{DAGID}_vm_metervm_get_tx_err_{count,sum}.
    • Added avalanche_{DAGID}_vm_metervm_verify_tx_{count,sum}.
    • Added avalanche_{DAGID}_vm_metervm_verify_tx_err_{count,sum}.
    • Added avalanche_{DAGID}_vm_metervm_accept_{count,sum}.
    • Added avalanche_{DAGID}_vm_metervm_reject_{count,sum}.

Coreth

  • Applied callTracer fault handling fix.
  • Initialized multicoin functions in the runtime environment.

ProposerVM

  • Updated block Delay in --staking-enabled=false networks to be 0.

v1.6.2 (View on GitHub)

This version is backwards compatible to v1.6.0. It is optional, but encouraged.

Config Options

  • Removed --coreth-config. See here.
  • Added --throttler-inbound-node-max-processing-msgs. See here.
  • Added --db-config-file. See here.

API

  • API method avm.exportAVAX has been removed. Use avm.export instead.
  • API method avm.importAVAX has been removed. Use avm.import instead.
  • API method info.peers now treats the PublicIP field as option and only populates it when a valid IP is provided.
  • API client platform.getValidatorsAt has been added.
  • API client admin.lockProfile has been fixed to correctly call lockProfile.
  • API client health.health has been fixed to properly handle responses from an unhealthy server.
  • Improved responses from the Health Check API to be more descriptive.

Benchlist

  • Changed the minimum time a validator must be unresponsive and the maximum amount of time a validator will be benched for. These used to be 5 minutes and 30 minutes, respectively, and are now 2.5 minutes and 15 minutes.

Database

  • Allowed users to specify database config with flag --db-config-file.

Subnets

  • Added the ability for a client to configure a Subnet as private to restrict membership to only approved validators.

Networking

  • Changed the default size of the inbound at-large message allocation from 32 MiB to 6 MiB.
  • Changed the default size of the outbound at-large message allocation from 32 MiB to 6 MiB.
  • Changed the default maximum number of bytes a node can take from the inbound at-large message allocation from 4 MiB to 2 MiB.
  • Changed the default maximum number of bytes a node can take from the outbound at-large message allocation from 4 MiB to 2 MiB.
  • Added additional inbound message rate-limiting. A node will not read more messages from a peer until it is processing less than --throttler-inbound-node-max-processing-msgs from that peer.
  • Changed default number of non-validators an AppGossip message is gossiped to from 2 to 0.
  • Changed default number of validators an AppGossip message is gossiped to from 4 to 6.
  • Introduced the ability for a VM to gossip to specific validators rather than just uniformly randomly.
  • Fixed an issue that caused some nodes to never attempt to reconnect to a previously disconnected node.

ProposerVM

  • Introduced a pessimistic P-chain height lag to improve stability during high P-chain block issuance.
  • Correctly applied the requested block delay.

Metrics

  • Removed API request histogram metrics from the X-chain and the P-chain.
  • Added P-chain mempool metrics.
  • Added validator_sets metrics to the platformvm.

Other

  • Refactored node startup and shutdown to avoid ungraceful shutdowns in the event that the node is started and then immediately stopped.
  • Fixed P-chain mempool to correctly track the number of allocated bytes.
  • Upgraded the C-chain to run geth 1.10.9.
  • Supported abigen for the C-chain.
  • Added support for pre-image support on the C-chain.
  • Added support for the fee history endpoint on the C-chain.
  • Refactored ID aliasing to better support GRPC tests.
  • Removed the end-to-end test branch matchup logic.
  • Removed the deprecated main entry point for the database migration's process manager.

v1.6.1 (View on GitHub)

This version is backwards compatible to v1.6.0. It is optional, but encouraged.

Upgrades

  • Added ability to specify Subnet configs
  • Added various new networking config values
  • Removed legacy messages from the network library
  • Fixed P-chain mempool bug that impacted AddValidator transactions on local networks
  • Changed transaction gossip rules to gossip to a fixed number of validators as well as all peers
  • Removed deprecated getLiveness method from the Health API
  • Added config option to disallow connections between non-validators

Note

The following are deprecated and should no longer be used. They may be removed in any future version:

  • API method avm.exportAVAX should be removed in favor of avm.export
  • API method avm.importAVAX should be removed in favor of avm.import
  • Config option coreth-config should be removed in favor of a chain config file.

v1.6.0 (View on GitHub)

This change is not backwards compatible with previous releases.

This upgrade adds a contention limiter to the C-chain and P-chain, introduces a block based fee on the C-chain, and tunes some dynamic fee parameters on the C-chain.

The changes in the upgrade go into effect at 5 PM EDT / 9 PM UTC, September 22th 2021 on Mainnet. You should upgrade your node before the changes go into effect, otherwise you may experience loss of uptime on your node.

More info can be found here.

Go

The minimum Go version required to build AvalancheGo is now Go 1.16.8

Bug Fixes

Fix race condition during timeout manager startup.

Upgrades

  • Introduced Snowman++ on the P-chain and C-chain.
  • Introduced mempool gossiping to the P-chain and C-chain using the VM<->VM communication layer.
  • Added a block based fee to C-chain blocks.
  • Set the minimum gas price to 25 nAVAX and the maximum gas price to 1000 nAVAX in the C-chain dynamic fee mechanism.
  • Rate limit incoming connections

New Metrics

  • avalanche_C_blks_built / avalanche_P_blks_built: Number of blocks that have been built locally on the C-Chain and P-Chain, respectively.
  • avalanche_C_blks_builds_failed / avalanche_P_blks_builds_failed: Number of calls to BuildBlock that failed on the C-Chain and P-Chain, respectively.

Config Options

  • Added flag inbound-connection-throttling-max-conns-per-sec.(See config documentation.)
  • Deprecated flag inbound-connection-throttling-max-recent. This flag is now ignored.

PRE_RELEASE v1.6.0-fuji (View on GitHub)

Please note that this release is unable to run mainnet - and will display "mainnet is not supported" if attempted to run with a mainnet configuration.

This upgrade adds a contention limiter to the C-chain and P-chain, introduces a block based fee on the C-chain, and tunes some dynamic fee parameters on the C-chain.

The changes in the upgrade go into effect at 5 PM EDT, September 16th 2021 on the Fuji testnet. After Fuji is updated and verified, a Mainnet compatible release will be published.

All Fuji nodes should upgrade before 5 PM EDT, September 16th 2021.

Upgrades

  • Introduced Snowman++ on the P-chain and C-chain.
  • Introduced mempool gossiping to the P-chain and C-chain using the VM<->VM communication layer.
  • Added a block based fee to C-chain blocks.
  • Set the minimum gas price to 25 nAVAX and the maximum gas price to 1000 nAVAX in the C-chain dynamic fee mechanism.
  • Added metrics for the number of blocks built and the number of failed build block attempts.

v1.5.3 (View on GitHub)

This version is backwards compatible to v1.5.0.

Uptime

  • Changed minimum uptime requirement to receive a staking reward from 60% to 80%.

Networking

  • Added 3 new network messages: AppRequest, AppResponse and AppGossip. These messages allow instances of a blockchain to send arbitrary data to each other as defined by their VM. Previously, instances of a blockchain could only communicate with one another by sending consensus messages (Put, PushQuery, etc.). See snow/engine/common/engine.go.
  • Upon receipt of a Pong message, disconnect from the sender if their version is incompatible.
  • Prepended method named in common.Sender with Send for clarity (e.g. Put --> SendPut).

P-Chain

  • Added functionality to track changes in validator weight by block.
  • Added API method GetValidatorsAt which allows for retrieval of a Subnet's (or the Primary Network's) validator set at a given P-Chain height.

C-Chain

  • Incorporate changes from Geth v1.10.8
  • Remove references to Ancients

Consensus

  • Added method Timestamp() to the snowman.Block interface.

Local Networks

  • Updated the start time of the validators in the local genesis. The end time for validators specified in the local config in versions before v1.5.3 is Sep. 10, 2021 00:00:00 UTC. Because of this, you must upgrade to AvalancheGo v1.5.3 in order to run a local network after this time.

Config Options

  • Added AvalancheGo config option consensus-app-gossip-size, which defines the number of peers an AppGossip message is gossiped to.
  • Added C-Chain config option log-level. Options are: "trace", "debug", "info", "warn", "error", "crit". Defaults to "debug" (as before.)

v1.5.2 (View on GitHub)

This update is backwards compatible with v1.5.0. Please see the expected update times in the v1.5.0 release.

Coreth

AVM

  • Introduced stateless codec generation for improved tooling.

Consensus

  • Added additional logging around bubbling votes.

v1.5.1-eth_call (View on GitHub)

This update is backwards compatible with v1.5.0. Please see the expected network upgrade times in the v1.5.0 release.

This update is a hotfix for v1.5.1 that allows using eth_call without the externally owned account check.

v1.5.1 (View on GitHub)

This update is backwards compatible with v1.5.0. Please see the expected network upgrade times in the v1.5.0 release.

Configuration

  • Removed option bootstrap-retry-max-attempts and added option bootstrap-retry-warn-frequency

Subnets

  • Added subnetIDs to the handshake message. This notifies peers about which Subnets a node is interesting in syncing.
  • Optimized Subnet container gossiping.

AVM

  • Fixed avm.GetTx's JSON endpoint to properly report amounts on UTXOs.

Bootstrapping

  • Fixed busy loop that could occur if a node's internet dropped during bootstrapping, causing the node to report a fatal error.

RPCChainVM

  • Improved caching of unverified blocks.

Coreth

  • Updated to Geth v1.10.7.

v1.5.0 (View on GitHub)

This change is not backwards compatible with previous releases.

This upgrade adds dynamic fees to the C-chain, along with various other improvements.

The changes in the upgrade go into effect at 10 AM EDT, August 24th 2021 on Mainnet. You should upgrade your node before the changes go into effect, otherwise you may experience loss of uptime on your node.

More info can be found here.

Network Upgrades

  • Added dynamic fee calculations to the C-chain.
  • Increased CreateSubnetTx and CreateChainTx fees.
  • Fixed heap corruption bug in delegator validation.
  • Enforced MaxStakeWeight for delegation transactions.

Client Upgrades

  • Added transaction indexing capabilities to the X-chain to enable historical lookups of transactions by address and asset.
  • Added ./avalanchego as the default command in the docker image.
  • Used static dependency versions in the docker image.
  • Removed database migration support and deamon runner.
  • Refactored node config parsing.
  • Optimized container gossiping sampling.
  • Added the ability to statically build the AvalancheGo and EVM binaries.
  • Simplified the Block interface to only expose the parent block's ID rather than fetching the full parent block.
  • Added additional metrics for pending jobs in the consensus engines.
  • Refactored P-chain statuses to handle blockchain validation statuses separately from transaction confirmation statuses.

Updated APIs

  • Added GetAddressTxs to the avm API.
  • Added SetLoggerLevel and GetLoggerLevel to the Admin API to allow fine grained tuning of log levels while the node is still running.
  • Added GetConfig to the Admin API to allow fetching the node config that the node is currently using.
  • Updated platformvm.Client to allow specifying nodeIDs in GetCurrentValidators and GetPendingValidators and generalized the response to GetStake.

Updated CLI Arguments

  • Removed fetch-only.
  • Added JSON config parsing to avm VM.
    • Added indexTransactions
    • Added indexAllowIncomplete

PRE_RELEASE v1.5.0-fuji (View on GitHub)

Please note that this release is unable to run mainnet - and will display "this node version doesn't support mainnet" if attempted to run with a mainnet configuration. If you run a mainnet node, no action is required until the official release is published next week.

This change is not backwards compatible with previous releases.

This upgrade adds dynamic fees to the C-chain, along with various other improvements.

The changes in the upgrade go into effect at 3 PM EDT, August 16th 2021 on the Fuji testnet. After Fuji is updated and verified, a mainnet compatible release will be published.

Network Upgrades

  • Added dynamic fee calculations to the C-chain.
  • Increased CreateSubnetTx and CreateChainTx fees.
  • Fixed heap corruption bug in delegator validation.
  • Enforced MaxStakeWeight for delegation transactions.

Client Upgrades

  • Added transaction indexing capabilities to the X-chain to enable historical lookups of transactions by address and asset.
  • Added ./avalanchego as the default command in the docker image.
  • Used static dependency versions in the docker image.
  • Removed database migration support and deamon runner.
  • Refactored node config parsing.
  • Optimized container gossiping sampling.
  • Added the ability to statically build the AvalancheGo and EVM binaries.
  • Simplified the Block interface to only expose the parent block's ID rather than fetching the full parent block.
  • Added additional metrics for pending jobs in the consensus engines.
  • Refactored P-chain statuses to handle blockchain validation statuses separately from transaction confirmation statuses.

Updated APIs

  • Added GetAddressTxs to the avm API.
  • Added SetLoggerLevel and GetLoggerLevel to the Admin API to allow fine grained tuning of log levels while the node is still running.
  • Added GetConfig to the Admin API to allow fetching the node config that the node is currently using.
  • Updated platformvm.Client to allow specifying nodeIDs in GetCurrentValidators and GetPendingValidators and generalized the response to GetStake.

Updated CLI Arguments

  • Removed fetch-only.
  • Added JSON config parsing to avm VM.
    • Added indexTransactions
    • Added indexAllowIncomplete

v1.4.12 (View on GitHub)

This update is backwards compatible. It is optional, but encouraged.

X-Chain

  • Added formatting argument "json" to API method GetTx, which returns the JSON representation of the queried transaction
  • Added interface type assertions

Info API

  • Added method GetNodeVersion to Info API client

Prometheus Metrics

  • Fixed and renamed metrics for bytes not sent due to compression
  • Added metrics for bytes not received due to compression
  • Added helper struct noAverager to metrics package

Database

  • Updated/added benchmarks

Shared Memory

  • Replace Put and Remove with Apply to allow for future atomic transaction optimization

v1.4.11 (View on GitHub)

C-Chain

This release enables snapshots by default.

Config Flags

Removed

  • conn-meter-reset-duration
  • conn-meter-max-conns

Added

  • network-compression-enabled

Prometheus Metrics

Many Prometheus metrics were renamed, and many histograms were replaced with 2 gauges. See here for updated Grafana Dashboards.

This release also adds helper methods to the utils/metric package.

RocksDB

RocksDB is no longer built by default when running the build script, and it is not included in publicly released binaries. To build AvalancheGo with RocksDB, run export ROCKSDBALLOWED=1 in your terminal and then scripts/build.sh. You must do this before you can use --db-type=rocksdb.

The RocksDB database now places/looks for its files in a subdirectory rocksdb. Note that if you previously ran with RocksDB, you'll need to move the existing files.

Message Compression

Nodes now compress some P2P messages. If a peer is version >= v1.4.11, Put, Push Query, Peer List and Multiput messages sent to the peer are compressed using gzip before being sent over the network. This reduces AvalancheGo's bandwidth usage.

Inbound Connection Throttling Refactored inbound connection rate-limiting and enable it by default.

General Improvements

  • Refactored and improved performance of iteration over a database served by gRPC to a plugin.
  • On Linux, clean up the C-Chain if AvalancheGo dies ungracefully
  • Refactored P2P message definitions and move them from the network package.
  • Added VM aliases to the HTTP API server
  • Replaced 1024 with units.KiB, etc.
  • Improved partition tolerance by processing chits in order of the creation of the corresponding queries.

Fuji IPs

Updated the bootstrap IPs for the Fuji Testnet.

v1.4.10 (View on GitHub)

Apricot Phase 2 - Patch 10

caution

This update is backwards compatible. It is optional, but encouraged.

The patch includes performance, throttling, and VM improvements:

  • Added support to use RocksDB rather than LevelDB on supported architectures.
  • Restructured inbound network throttling to be on a per-node basis, to restrict the bandwidth usage of peer nodes.
  • Restructured outbound network throttling to weight allocated bytes by stake.
  • Updated the default value of the pruning-enabled flag to true for the C-chain.
  • Enabled registering of custom VMs over RPC.
  • Updated blockchain status to report validation status.
  • Moved TimestampVM into its own repository to match the expected VM creation path.
  • Fixed protobuf code-gen script to place grpc files in the correct location.
  • Passed the block bytes through the rpcchainvm#Block.Verify to avoid any potential cache eviction verification failures.

v1.4.9 (View on GitHub)

Apricot Phase 2 - Patch 9

caution

This update is backwards compatible. It is optional, but encouraged.

The patch includes performance improvements, and monitoring improvements:

  • Added support to run the C-chain with pruning enabled. Pruning is currently disabled by default.
  • Reduced C-chain Websocket ping interval to reduce disconnects when behind load balancer.
  • Added timestamp to snowman Block interface.
  • Fixed bug in C-chain API max duration enforcement for calls made via websockets.
  • Added gzip header support for the http endpoint.
  • Added additional version descriptions to the info.getNodeVersion endpoint.
  • Restricted connecting to node versions >= 1.4.5.
  • Moved daemon logs under the primary log folder.
  • Added support for deterministic sampling.
  • Added auto deployment GitHub action for new tags.
  • Refactored config management to better support launching nodes programmatically.

v1.4.8 (View on GitHub)

Apricot Phase 2 - Patch 8

caution

This update is backwards compatible. It is optional, but encouraged.

The patch includes performance improvements, monitoring improvements, and Subnet fixes:

  • Changed the AVM's fee definition to enforce fees to be paid in the chain's native asset. This doesn't change the X-Chain's behavior, but it makes other AVM instances usable.
  • Added the ability to specify configs to specific chains. This deprecates the coreth-config CLI parameter.
  • Added rate limiting to the number of new outbound connections.
  • Introduced a VM wrapper that adds transparent metrics to a chain.
  • Added the ability to enable continuous node profiling.
  • Reduced byte allocations in the networking layer.
  • Added various CLI parameters for tuning gossip parameters.
  • Enabled nodes to run using an ephemeral key pair, rather than one that is read from disk.
  • Removed incorrect spurious warning.
  • Moved CI tests to run in Github Actions rather than running in Travis.
  • Removed special cases from the VM interface.

Added Command Line Arguments:

  • profile-dir
  • profile-continuous-enabled
  • profile-continuous-freq
  • profile-continuous-max-files
  • chain-config-dir
  • bootstrap-multiput-max-containers-received
  • bootstrap-multiput-max-containers-sent
  • boostrap-max-time-get-ancestors
  • consensus-on-accept-gossip-size
  • consensus-accepted-frontier-gossip-size
  • meter-vms-enabled
  • staking-ephemeral-cert-enabled
  • outbound-connection-timeout
  • outbound-connection-throttling-rps

v1.4.7 (View on GitHub)

Apricot Phase 2 - Patch 7

caution

This update is backwards compatible. It is optional, but encouraged. The patch includes performance improvements and bug fixes.

If the previously installed node version is <= v1.4.4 then this node may have stopped processing blocks. This update will repair the node and perform a database migration. If the previously installed node version is >=v1.4.5 then this node will use the existing database and does not need to perform a database migration.

  • Fixed the pre-migration node to correctly verify the P-chain block SHraz7TtMfTQ5DX1rREhNZW1bi7PpPzAq7xoJAwrWNQrLhQcD.
  • Fixed regression in platformvm.GetBlockchains to correctly return the primary Subnet blockchains.
  • Updated the grpc version to v1.37.
  • Optimized peerlist sampling.
  • Added database benchmarks.
  • Reduced various repeated memory allocations.

v1.4.6 (View on GitHub)

Apricot Phase 2 - Patch 6

caution

This update is backwards compatible. It is optional, but encouraged. This patch includes performance improvements and bug fixes.

If the previously installed node version is <= v1.4.4 then this node will perform a database migration. For details about the database migration please see the v1.4.5 release notes. If the previously installed node version is v1.4.5 then this node use the existing database and does not need to perform a database migration.

This patch:

  • Removes invalid transaction issuance into P-chain mempool that caused high sustained DB writes.
  • Ignored non-database files and folders in the database directory. This should specifically fix errors reported on macOS with .DS_Store files.
  • Fixed the build-dir flag to be able to be specified via CLI without causing the preupgrade node to error.
  • Removed the plugin-dir flag that is no longer supported with the node-manager daemon. Typically not specifying the flag leads to the correct behavior. However, for complex installations the build-dir flag may be required.
  • Enforced gossiping messages only to connections that have finished the peer handshake.
  • Reduced memory allocations during consensus traversals and bootstrapping.

v1.4.5 (View on GitHub)

Apricot Phase 2 - Patch 5 - DB Upgrade

This upgrade is more involved than the typical version update.

caution

This update is backwards compatible. It is optional, but encouraged. The patch includes significant performance improvements and numerous other updates.

VM Improvements:

  • Fully redesigned the platformvm's state management.
    • Removed the usage of versiondbs being passed through blocks to pass state references that can be modified and read without re-parsing objects.
    • Implemented a base state manager to properly cache and mange writes to the underlying database.
    • Implemented CoW validator sets to enable caching multiple validator sets in memory.
    • Indexed chains by Subnet to avoid touching unused state objects.
    • Indexed validators by nodeID to avoid unnecessary iterations while accepting addDelegator and addSubnetValidator transactions.
    • Reduced the number of key-value pairs dedicated to managing validator sets on disk and validator uptimes.
  • Added staking reward look-ups to the platformvm's API to support indexing of rewards.
  • Refactored validator uptime metering to simplify testing.
  • Added block and transaction type metrics to the platformvm.
  • Added API call metrics to the avm and the platformvm.
  • Updated the avm's state management to use prefixdbs, record caching metrics, and share additional code with the platformvm.
  • Simplified UTXO management and indexing in the avm and platformvm.
  • Restructured address parsing and management to be fully shared across compatible VM instances.
  • Restructured shared memory of the primary Subnet to be fully shared across VM instances.
  • Added a chain state implementation to support seamless caching over existing VM implementations and to simplify the implementation of new VMs.
  • Integrated the new chain state manager into the rpcchainvm, which also adds various metrics.
  • Added upgradeBytes and configBytes to the standard VM interface to better support future network upgrades.
  • Added getAtomicTx and getAtomicTxStatus endpoints to the evm API.
  • Simplified evm block production to be synchronously performed with the consensus engine.
  • Added an atomic transaction mempool to re-introduce orphaned atomic transactions.
  • Fixed bug in the evm client to properly set the sourceChain in getAtomicUTXOs.
  • Integrated the new chain state manager into the evm to better optimize block management.

Bootstrapping Improvements:

  • Removed re-traversals during bootstrapping. This significantly improves the performance of the node during restarts of the bootstrapping process.
  • Fixed an ungraceful node shutdown when attempting to exit the node while executing bootstrapped containers.
  • Fixed duplicated IPC container broadcasts during bootstrapping.
  • Standardized the bootstrapping jobs queue to write to state using prefixdbs rather than implementing custom prefixing.
  • Added additional bootstrapping caching and cache metrics.

Database Migration Additions:

  • Added a daemon process manager to seamlessly migrate to the updated database format.
  • Refactored version handling to track database semantic versions.
  • Implemented a database manager to track and operate over different database versions.
  • Implemented a keystore migration that automatically copies users from the v1.0.0 database to the v1.4.5 database.
  • Implemented a validator uptime migration from the v1.0.0 database to the v1.4.5 database.

Node Improvements:

  • Updating config parsing to always expand environment variables.
  • Refactored the node config to allow specifying TLS certificates in memory without touching disk.
  • Added better support for meaningful exit codes.
  • Displayed listening address of the http and staking servers to aid in supporting non-specific port mappings.
  • Implemented a versionable database to be able to toggle between a pass through database and a versioned database.
  • Optimized ID Set pre-allocations and reduced the memory usage of the structs.
  • Enforced stricter linting rules.

Modified command line arguments:

For the following arguments "default" was previously treated as a keyword. Now, "default" will attempt to be treated as the intended value of the flag. To retain the default behavior, the flag should not be specified.

  • config-file
  • coreth-config
  • plugin-dir
  • staking-tls-key-file
  • staking-tls-cert-file
  • bootstrap-ips
  • bootstrap-ids
  • ipcs-path
  • db-dir

For the following arguments "" was previously treated as a keyword. Now, "" will attempt to be treated as the intended value of the flag. To retain the default behavior, the flag should not be specified.

  • ipcs-chain-ids
  • log-dir
  • log-display-level

It is no longer required that the bootstrap-ips and bootstrap-ids are paired. This means it is now valid to specify a different number of bootstrap-ips than bootstrap-ids. The bootstrap-ips are used to initially connect to the network and the bootstrap-ids are used as the beacons in bootstrapping.

Added command line arguments:

  • fetch-only
  • build-dir

Removed command line arguments:

  • xput-server-port
  • xput-server-enabled

v1.4.4 (View on GitHub)

Apricot Phase 2 - Patch 4

caution

This update is backwards compatible. It is optional, but encouraged.

The patch includes bug fixes and performance improvements that aim to optimize the upcoming db-upgrade release.

  • Skipped tailing delay in bootstrapping so that all chains finish as soon as the last chain is marked as bootstrapped in a Subnet.
  • Improved message handling during bootstrapping to handle messages while waiting for other chains to sync.
  • Reduced sampler allocations by re-using existing samplers.
  • Updated docker scripts to only push images from the master branch.
  • Fixed log formatting.
  • Improved error messages.

v1.4.3 (View on GitHub)

Apricot Phase 2 - Patch 3

caution

This update is backwards compatible. It is optional, but encouraged.

The patch includes bug fixes, updated uptime monitoring, and performance improvements.

  • Fixed benched message handling that could cause a node to be unable to progress during bootstrapping. This was typically experienced when the node would fail to transition to normal execution as it was finishing bootstrapping.
  • Fixed a non-deterministic bug in the C-Chain codebase that could cause nodes that receive a lot of transaction broadcast requests to temporarily stop producing blocks until they process a block produced by another node.
  • Restricted the number of version messages to be sent to a peer to one.
  • Removed legacy handshake messages that were deprecated in Apricot Phase 2.
  • Marked nodes that have been benched as being offline for uptime calculations.
  • Updated the validator set to be more performant during validator set changes.
  • Updated the networking to only attempt to re-connect to a peer on disconnect if they are currently a validator.

v1.4.2 (View on GitHub)

Apricot Phase 2 - Patch 2

caution

This update is backwards compatible with v1.4.0 and v1.4.1. The changes in the upgrade go into effect at 10 AM EDT, May 5th 2021 on the Fuji testnet and 7 AM EDT, May 10th 2021 on mainnet.

The patch further reduces the size of gossiped peerlist messages and introduces several new flags:

  • network-peer-list-size allows for tuning the number of peers gossiped in each peerlist message.
  • network-peer-list-gossip-size allows for tuning the number of peers to gossip peerlist messages to.
  • network-peer-list-gossip-frequency allows for tuning how frequently peerlists are gossiped.

v1.4.1 (View on GitHub)

Apricot Phase 2 - Patch 1

caution

This update is backwards compatible with v1.4.0. Please see the expected update times in the v1.4.0 release.

The patch reduces the size of gossiped peerlist messages and introduces a new flag --bootstrap-beacon-connection-timeout that allows for the beacon connection timeout to be configured on startup.

v1.4.0 (View on GitHub)

Apricot Phase 2

danger

Please note that this change is not backwards compatible with previous releases.

The related blog post can be found here.

caution

This upgrade applies the Ethereum Berlin upgrade to the C-chain, adds a new AVM endpoint, and includes various stability improvements. We urge everyone in the community to update as soon as possible in order to ensure that their nodes remain healthy.

The changes in the upgrade go into effect at 10 AM EDT, May 5th 2021 on the Fuji testnet and 7 AM EDT, May 10th 2021 on mainnet.

The primary components to this upgrade include:

  • Updated Coreth to depend on v1.10.2 of go-ethereum.
  • Applied the Ethereum Berlin upgrade. Specifically EIP-2565, EIP-2718, EIP-2929, and EIP-2930.
  • Added new stateful pre-compiled smart contracts to the C-chain to support ANT transfers and ARC-20 wrappers around ANTs.
  • Added an AVM /events endpoint that supports websocket notification of transactions being accepted matching an addresses filter.
  • Added two new networking message types SignedVersion and SignedPeerlist to improve validator -> IP mappings.
  • Fixed a long standing bug where shutting down the node while a chain was bootstrapping could cause the chain to be shut down ungracefully.
  • Updated the plugin gRPC packages to paginate large requests to improve stability.
  • Added the ability to run avalanchego's main binary as a plugin.
  • Fixed a potential race condition in the leveldb corruption protection.
  • Updated the automated build scripts to better support multiple architectures.

Added command line arguments:

  • plugin-mode-enabled specifies the binary to run in plugin mode.

Removed command line arguments:

  • p2p-tls-enabled
  • disconnected-check-frequency
  • disconnected-restart-timeout
  • restart-on-disconnected

v1.3.2 (View on GitHub)

Apricot Phase 1 - Patch 2

danger

This update is backwards compatible. It is optional, but encouraged. The patch includes security improvements, bug fixes, and monitoring improvements.

Security Improvements

  • Enforced a strict canonical format for C-chain blocks made prior to Apricot Phase 1. This ensures that modifications to the extra-data block field can not result in modifications to the chain state during bootstrapping.
  • Changed the Keystore to ensure only encrypted values are sent over the IPC between avalanchego and plugin processes.

Bug Fixes:

  • Fixed delegation cap calculations to include updating the current delegation maximum before removing a delegator. This ensures that the delegation cap is always enforced.
  • Fixed AVM's static API to be registered correctly on startup.
  • Updated node uptime calculations to take network upgrades into account.

Monitoring Improvements

  • Added an optional node indexer that can provide a locally consistent ordering of operations accepted on a chain.
  • Updated ansible inventory to include numerous improvements (Huge thanks to @moreati).

v1.3.1 (View on GitHub)

Apricot Phase 1 - Patch 1

danger

This update is backwards compatible. It is optional, but encouraged. The patch includes stability, monitoring improvements, and minor bug fixes.

The primary components to this upgrade include:

  • Fixed C-chain segfault when performing compression on arm64 CPUs.
  • Added group permissions to local files to enable complex node monitoring.
  • Stripped white space from Auth passwords passed through the api-auth-password-file flag.
  • Removed timeSinceNoOutstandingRequests as it was replaced by longestRunningRequest.
  • Added additional metrics in network throttling.
  • Various code cleanup.

Added command line arguments:

  • network-health-max-outstanding-request-duration

Removed command line arguments:

  • network-health-max-time-since-no-requests

v1.3.0 (View on GitHub)

Apricot Phase 1

danger

Please note that this change is not backwards compatible with previous releases.

This upgrade reduces C-chain gas fees, removes C-chain gas refunds, and includes various security improvements. We urge everyone in the community to update as soon as possible in order to ensure that their nodes remain healthy.

The changes in the upgrade go into effect at 10 AM EST, March 25th 2021 on the Fuji testnet and 10 AM EST, March 31st 2021 on mainnet.

The primary components to this upgrade include:

  • Reduced C-chain gas cost from 470 nAVAX to 225 nAVAX.
  • Removed C-chain gas refunds. This change adopts EIP-3298.
  • Refactored C-chain verification to be cleaner when performing network upgrades.
  • Fixed the Auth API to properly enforce revoked tokens.
  • Strengthened the Auth API to ensure the expected signature format is used.
  • Removed the Auth API's password from the CLI arguments.
  • Added more strict file permissions checks.
  • Added some minor additional error handling.
  • Sanitized log writes before being written to disk.
  • Added configurable origins to the HTTP endpoint.
  • Removed attempted HTTPs to HTTP fail over on startup. Now the node will close on startup if upgrading the HTTP endpoint to HTTPs fails.

Added command line arguments:

  • api-auth-password-file specifies the file to read the Auth API's password from.

Removed command line arguments:

  • api-auth-password

v1.2.4 (View on GitHub)

Apricot Phase 0 - Upgrade 1 - Patch 4

danger

This update is backwards compatible. It is optional, but encouraged. The patch includes stability and monitoring improvements.

  • Updated readme to correct storage requirements.
  • Added additional error handling to Avalanche Tx verification during bootstrapping.
  • Updated numerous metrics, including adding numerous new metrics relating to node health and database usage, removing unused and invalid metrics, and fixing some metric names.
  • Added additional logging to CI.
  • Added the C-chain to the list of critical chains.

v1.2.3 (View on GitHub)

Apricot Phase 0 - Upgrade 1 - Patch 3

danger

This update is backwards compatible. It is optional, but encouraged. The patch includes stability and monitoring improvements.

  • Adjusted [network, router, consensus] health check parameters to remove flaky health checks.
  • Simplified C-chain block handling.

v1.2.2 (View on GitHub)

Apricot Phase 0 - Upgrade 1 - Patch 2

danger

This update is backwards compatible. It is optional, but encouraged. The patch includes stability, performance, and monitoring improvements.

  • Added IP aliases in the network library to avoid repeated SYN calls.
  • Fixed bootstrap message handling when bootstrapping from yourself.
  • Simplified AdvanceTimeTx issuance.
  • Added new consensus health checks.
  • Adding node health logging.
  • Added health responses to health GET requests.
  • Consolidated incoming message logs.
  • Added error logging to the LevelDB wrapper.
  • Added error codes to the rpcdb to avoid string parsing.
  • Improved C-chain handling of canonical chain to reduce the number of reorgs.
  • Improved C-chain handling of mock calls performed on the pending block.

v1.2.1 (View on GitHub)

Apricot Phase 0 - Upgrade 1 - Patch 1

danger

This update is backwards compatible. It is optional, but encouraged. The patch includes stability, performance, and monitoring improvements.

Please note that this update removes `network-timeout-increase` and ‘network-timeout-reduction` as command line arguments.

Change Summary:

  • Added `UTXO`s to the `platformvm.getStake` response.
  • Added benchlist reporting to the `info.peers` response.
  • Added additional health checks into the networking layer.
  • Added `percent of stake connected` as a reported metric.
  • Added bootstrapping restart logic to ensure the node has caught up to the current tip, even during times of high throughput.
  • Added Subnet-wide bootstrapping to ensure that a chain won't fall behind due to another chain bootstrapping.
  • Prevented verification of rejected blocks to avoid unnecessary computation.
  • Removed gossiping of non-preferred blocks to the network.
  • Switched the network timeout calculator to use an EWMA of the observed network latency.
  • Removed `Get` requests from the network latency calculations.
  • Cleaned up the benchlisting algorithm.
  • Cleaned up handling of dropped messages on send.
  • Cleaned up outstanding request and timeout logic.
  • Generalized performance tracking to allow for prefixing of profile names.
  • Added additional caching to the Avalanche bootstrapping traversal.
  • Fixed ansible linting.
  • The added command line arguments mainly consist of configurations of health checks. Additionally, the modified network latency calculations changed the name of some command line args.

Added command line arguments:

  • `network-timeout-halflife`
  • `network-timeout-coefficient`
  • `network-health-min-conn-peers`
  • `network-health-max-time-since-msg-received`
  • `network-health-max-time-since-msg-sent`
  • `network-health-max-portion-send-queue-full`
  • `network-health-max-send-fail-rate`
  • `network-health-max-time-since-no-requests`
  • `router-health-max-drop-rate`
  • `router-health-max-outstanding-requests`
  • `health-check-frequency`
  • `health-check-averager-halflife`
  • `bootstrap-retry-enabled`
  • `bootstrap-retry-max-attempts`

Removed command line arguments:

  • `network-timeout-increase`
  • `network-timeout-reduction`

v1.2.0 (View on GitHub)

Apricot Phase 0 - Upgrade 1

danger

Please note that this patch is not backwards compatible with previous releases. This upgrade fixes performance issues related to interchange transfers between X, C, and P chains. We urge everyone in the community to upgrade as soon as possible in order to ensure that their nodes are not affected. Also, note that nodes may take several extra minutes to connect after the upgrade and the process should be allowed to complete uninterrupted.

The primary components to this upgrade include:

  • Fixed atomic import validation on C-Chain
  • Added rule exception logic to allow atomic bonus blocks
  • Added fail-fast logic into Shared Memory if duplicated deletes are issued
  • Fixed issue where polls could stall in snowman because of a failure to clear requests
  • Fixed BAD BLOCK issue in coreth due to unknown ancestors
  • Fixed a race condition in the repair canonical chain script in coreth
  • Limited number of processing blocks in Snowman and processing txs in Avalanche
  • Updated networking timeout default values and benchlist settings
  • Verified there was no safety violation after the initial network instability

v1.1.5 (View on GitHub)

Apricot Phase 0 - Patch 5

danger

This update is backwards compatible. It is optional but encouraged. The patch includes stability improvements.

  • Fixed a potential deadlock when registering new chains that could cause the P-chain and http(s) endpoint to block.
  • Repairs TxID -> Block Height indexing in the C-chain.
  • Added graceful handling of empty contract deployments in the debug_traceTransaction API in the C-chain.
  • Improved error handling in the C-chain.

v1.1.4 (View on GitHub)

Apricot Phase 0 - Patch 4

danger

This update is backwards compatible. It is optional but encouraged. The patch includes CLI upgrades, API bug fixes, stability improvements, and performance improvements.

  • Fixed an issue where C-chain block indexes could map to unaccepted blocks at a given height.
  • Fixed VM crash when the RPCChainVM experienced high API loads.
  • Fixed optimistic vote bubbling in the Avalanche Engine to correctly pass votes through processing vertices.
  • Added field IncludePartial to the AVM's GetBalance and GetAllBalances API methods. This changes the default behavior to only return the balances of spendable and uniquely owned assets.
  • Added the ability to specify custom genesis configs for custom network IDs.
  • Added additional IPC API functionality.
  • Added additional caching to the RPCChainVM.
  • Improved plugin directory lookup to always work with the binary releases.

v1.1.3 (View on GitHub)

Apricot Phase 0 - Patch 3

danger

This update is optional but encouraged. The patch includes minor bug fixes relating to APIs.

  • Fixed hanging call when attempting to filter C-chain logs.
  • Fixed C-chain client to call the proper multi-coin API.
  • Added getAtomicUTXOs to avm and platformvm API clients.

v1.1.2 (View on GitHub)

Apricot Phase 0 - Patch 2

danger

This update is optional but encouraged. The patch includes bug fixes and performance improvements.

  • Fixed bootstrapping processing cache to reduce duplicated traversals when bootstrapping Avalanche.
  • Optimized P-chain verification during bootstrapping.
  • Fixed maximum bench list calculation to use the proper input values.
  • Removed extra linter runs from CI.
  • Added Height to the snowman.Block interface.

v1.1.1 (View on GitHub)

Apricot Phase 0 - Patch 1

danger

This update is optional but encouraged. The patch includes bug fixes and performance improvements.

  • Fixed a node crash bug when users disabled the Health API.
  • Fixed a bug in uptime tracking that could over report a node's uptime.
  • Refactored vertex parsing to use a Codec.
  • Separated stateful and stateless vertex management.
  • Added per-field slice length checking to the Codec.
  • Introduced a new codec type that groups TypeIDs together.
  • Introduced message limit flags to the CLI.
  • Introduced a semanticdb package to be used during a future database migration.
  • Added Epoch tracking to the chain context.
  • Improved some of the error messages returned during transaction validation.
  • Reduced GC pressure in the version DB.

v1.1.0 (View on GitHub)

Apricot Phase 0

danger

Please note that this upgrade is not backwards compatible with previous releases. Upgrades must be performed no later than Monday, December 7th at 11 p.m. UTC (6 p.m. EST). The upgrade, which was originally scheduled around mid December, is now being expedited to fix an important token unlocking bug. We urge everyone in the community to upgrade as soon as possible in order to ensure that their nodes are not affected.

There are two primary components to this upgrade:

  • General preparations for our upcoming Apricot network upgrade called the Apricot Phase Zero Upgrade
  • Fixing an issue that prevented stake-able locked outputs from being unlocked after their lock _**_time had passed

v1.0.6 (View on GitHub)

danger

Please note that this release contains breaking changes. It changes the default response format of platform.getTxStatus and platform.getCurrentValidators. The update is optional but encouraged. The patch includes performance improvements and some quality of life improvements.

  • Removed deprecated formats of platform.getTxStatus and platform.getCurrentValidators.
  • Added support for hex encodings of imported and exported users from the keystore API.
  • Set golang requirement to v1.15.5 to avoid a DoS vulnerability found in the golang standard lib.
  • Added API clients to act as helpers interacting with the node software.
  • Enabled falling back to bootstrapping if a node becomes disconnected from the rest of the network.
  • Fixed the GetUTXOs APIs when UTXOs referenced multiple addresses.
  • Refactored binary encoding to better generalize RPC options.
  • Fixed IP block filtering to correctly set the window length.
  • Generalized the codec package to be able to manage multiple codecs with different versions.
  • Added Epoch to the Vertex interface in preparation of a future release.
  • Deferred transaction hashing to reduce CPU/Memory utilization past fast checks.
  • For those using https://explorerapi.avax-dev.network/, the URL is going to be shut down in a future release. Please switch over to https://explorerapi.avax.network/.

For assistance with this update, you can join our Discord for help.

v1.0.5 (View on GitHub)

danger

Please note that the release after this one, v1.0.6, will contain the breaking changes. Namely, the response format of platform.getTxStatus and platform.getCurrentValidators will change.

The changes in this release, v1.0.5, are backwards compatible with previous releases. The update is optional but encouraged. The patch includes performance improvements and some quality of life improvements.

  • Added IssueTx and GetUTXOs to the C-chain API to enable issuing atomic swaps without revealing private keys to a node.
  • Fixed memory leak in the snowman request manager with oracle block processing.
  • Fix UTXO pagination bug that under-reported available funds.
  • Moved chain http logs to live in the human-readable chain logs folder.
  • Restructure how IDs are managed to avoid heap allocations.
  • Optimized the UniformSamplers to avoid creating multiple maps.
  • Reduced usage of ids.Set in favor of []ids.ID to better utilize continuous memory.
  • Introduced []byte reuse in PrefixDB.
  • Implemented type-specific sorting functions to avoid frequent interface conversion allocations.
  • Optimized AVM load user to avoid reading unnecessary information from disk.
  • Removed a memory allocation + copy in socket sending for the full length of the message.

For assistance with this update, you can join our Discord for help.

v1.0.4 (View on GitHub)

danger

This update is optional but encouraged. The patch includes quality of life improvements and various performance enhancements. Note that this update requires the CLI parameters to be specified with -- rather than allowing for either - or --. For example, -public-ip=127.0.0.1 is no longer allowed and must be specified as --public-ip=127.0.0.1. Otherwise, this update is backwards compatible.

• Added Subnet whitelisting to allow a node owner to choose which Subnets to validate.
• Added config file parsing for node settings.
• Added more options for specifying a node's IP address and added getNodeIP to the info *endpoint.
• Added a TxID to the result of get.Validators in the platformvm.
• Updated Coreth version.
• Cleaned up the snowball trie implementation and added additional tests to align with mutation tests.
• Implemented and optimized continuous time averages for tracking CPU and network latency.
• Significantly optimized memory allocations in various locations.
• Increased the signature verification cache size.
• Reduced DB reads during vertex management.
• Added an optional argument includeReason to platform.getTxStatus.
If not provided, or if false, the output from getTxStatus is the same as before.

For example:
{
"jsonrpc": "2.0",
"result": "Dropped",
"id": 1
}

If includeReason is true, the output from getTxStatus has a new format. It's an object that looks like this:

{
"jsonrpc": "2.0",
"result": {
"status":"[Status]",
"reason":"[Reason tx was dropped, if applicable]"
},
"id": 1
}

In this new format, reason will not be present unless the status is Dropped.
Anything that depends on platform.getTxStatus should switch to using the includeReason argument and use the new response format. After a few releases, we'll only support the new response format.

For assistance with this update, you can join our Discord for help.