OffchainLabs/nitro v2.0.12

Arbitrum Nitro v2.0.12

on GitHub

This release is available as a Docker image on Docker Hub at offchainlabs/nitro-node:v2.0.12-65d4e7a

What's Changed

This release splits out block validation into a separate process, makes various database and state cache sizes configurable, improves the relay reliability, and begins work on the still in-progress ArbOS version 11.

The default cache sizes are now also larger, which will result in about 3GB of extra memory usage in exchange for making the node faster. To use the old defaults, add --node.caching.database-cache=0 --node.caching.snapshot-cache=100 --node.caching.trie-clean-cache=150.

Configuration Changes

• All --node.validator.* options were renamed to --node.staker.*
• --node.block-validator.jit-validator was renamed to --validation.use-jit (both default to true)
• --node.block-validator.jit-validator-cranelift is now --validation.jit.cranelift
• --node.block-validator.arbitrator-validator was removed; it's now enabled if use-jit is false
• --node.validator.target-machine-count is now --validation.arbitrator.execution.cached-challenge-machines
• --node.block-validator.concurrent-runs-limit is now --validation.arbitrator.workers or --validation.jit.workers depending on if jit is enabled or not (default enabled)

User-facing Improvements

• Make more caches configurable: #1478
• Feed connection limit: #1481

Pending ArbOS Version 11 Improvements (work in progress)

• Fix precompiles using all gas on non-solidity errors: #1490
• Fix address set AllMembers list: #1525

Internal Highlights

• Populate feed broadcaster catchup buffer on startup: #1480
• Sequencer: retry txs over the block gas limit: #1466
• Wait for another sequencer to come alive before shutting down sequencer: #1482
• Add a type for nonceFailureCache: #1530
• Add metrics for the nonce failure cache: #1483
• Forward transaction to active sequencer from redis: #1369
• Inbox update without insertion lock: #1462
• Fix a panic on L1 inconsistency about which stakers exist: #1469
• Fix wsbroadcastserver connection name logging: #1496
• Every day, optionally handoff sequencer lockout to compact database: #1484
• Improve nonce failure cache handling on shutdown: #1493
• Fix precheckNonces which previously assumed txs would succeed: #1529
• Update default arb1 forwarding target to new sequencer URL: #1526
• Add reconnect cooldown period for relay limiter: #1520
• Relay new connection delay 2: #1528
• Fix disabling the nonce cache on the sequencer: #1523

Full Changelog: v2.0.11...v2.0.12