near/nearcore 1.37.0

on GitHub

CODE_COLOR: CODE_YELLOW_MAINNET
RELEASE_VERSION: 1.37.0
PROTOCOL_UPGRADE: TRUE
DATABASE_UPGRADE: FALSE
SECURITY_UPGRADE: FALSE

Protocol Upgrade Voting

This release contains a protocol upgrade. Voting for upgrading to protocol version 64 will start on Monday 2024-03-11 18:00:00 UTC.

Managing Resharding

This protocol upgrade includes resharding of shard 3. Which means that after protocol upgrade we will have 5 shards. New shard split is defined by these border accounts vec!["aurora", "aurora-0", "kkuuue2akv_1630967379.near", "tge-lockup.sweat"]

Important points:

• Resharding will happen in the epoch RIGHT AFTER protocol voting epoch (and RIGHT BEFORE actual protocol upgrade).
• If your node fails to reshard before protocol upgrade, it will not be able to continue being a part of the chain. If that happens, you should restart from a Pagoda provided backup.
• Resharding creates a database snapshot that is deleted right after resharding is finished.During resharding, though, you may notice progressive increase of data folder size (around 150GB).
• To monitor resharding you can use metrics near_resharding_status, near_resharding_batch_size, and near_resharding_batch_prepare_time_bucket. You can read more here https://github.com/near/nearcore/blob/master/docs/architecture/how/resharding.md#monitoring.
• If you observe problems with block production or resharding performance, you can adjust resharding throttling configuration. Read more here https://github.com/near/nearcore/blob/master/docs/architecture/how/resharding.md#monitoring.
• RAM usage during resharding is around 8-10GB.

Full resharding doc can be found here https://github.com/near/nearcore/blob/master/docs/architecture/how/resharding.md.

Protocol Changes

• Resharding v2 - new implementation for resharding and a new shard layout for production networks. #10303, NEP-0508
• Restrict the creation of non-implicit top-level accounts that are longer than 32 bytes. Only the registrar account can create them. #9589
• Adjust the number of block producers and chunk producers on testnet to facilitate testing of chunk-only producers #9563

Non-protocol Changes

• Add prometheus metrics for the internal state of the doomslug. #9458
• Fix EXPERIMENTAL_protocol_config to apply overrides from EpochConfig. #9692
• Add config option tx_routing_height_horizon to configure how many chunk producers are notified about the tx. #10251

Known issues

• [This item includes the mitigations] State snapshot compaction may cause stack overflow. To avoid this issue, make sure that fields store.state_snapshot_config.compaction_enabled and store.state_snapshot_compaction_enabled are set to false in config.
• [This one does not] A bug in network tracing spans management may lead to stack overflow. This issue will be fixed in 1.37.1 release that we plan to make after the protocol upgrade.