Skip to content

Releases: NethermindEth/nethermind

v1.35.2

06 Nov 16:20
faa9b9e

Choose a tag to compare

Release notes

Important

This release replaces v1.35.1 and is mandatory for all node operators.

It includes the Ethereum Mainnet Fusaka hard fork configuration (Osaka, BPO1, and BPO2 upgrades).

All nodes must upgrade to this version before the Fusaka hard fork activation on December 3, 2025 at 9:49:11 PM to stay in sync.

Overview

This release activates the Ethereum Mainnet Fusaka hard fork, which introduces the Osaka, BPO1, and BPO2 upgrades.

What's Changed

Full Changelog

https://github.com/NethermindEth/nethermind/compare/1.35.1…1.35.2

v1.35.1

05 Nov 15:53
58f06cd

Choose a tag to compare

Caution

This release has block production and sync issues. Use v1.35.2 instead.

Release notes

Important

This release is mandatory for all node operators.

It includes the Ethereum Mainnet Fusaka hard fork (Osaka, BPO1, and BPO2 forks) and critical improvements that ensure network stability and compatibility.
All nodes must upgrade to this version before the Fusaka hard fork activation on December 3, 2025 at 9:49:11 PM to stay in sync.

Overview

This release continues the performance and reliability improvements introduced in v1.35.0 while activating the key Ethereum Mainnet Fusaka hard fork and strengthening node stability across networks.

This update focuses on network readiness and state consistency. It also resolves multiple edge cases affecting trie storage, block range updates, and protocol session handling.

Upgrading to this version ensures seamless operation through the Fusaka hard fork and reduces the risk of shutdowns or state inconsistencies during execution.

What's Changed


Full Changelog:
1.35.0...1.35.1

v1.35.0

20 Oct 09:53
e4a5377

Choose a tag to compare

Release notes

This version introduces important updates and optimizations across multiple components, improving network stability, performance, and test coverage.
It includes several fixes and enhancements to ensure smoother operation for all supported networks, including the Ethereum Mainnet.

Key improvements include:

  • Updated default Ethereum mainnet gas limit to 60M
  • Hardened code DB and trie persistence mechanisms
  • Optimizations for RPC workflows, pruning, and snap sync reliability
  • Enhancements to blob proofs, gas estimation, and network configuration updates
  • Reproducible Nethermind builds

These changes are strongly recommended for all operators to maintain compatibility and ensure optimal stability with the latest network parameters.

What's Changed

Read more

v1.34.1

06 Oct 13:40
c4238a3

Choose a tag to compare

Release notes

Important

This release is mandatory for all node operators.

It addresses the issue with nodes stuck at a certain block in Hoodi network, causing an issue:
Code 0x93c9d3b83e1711754119e36d3a33fa26ed814a0d17f47b67b7acfa8ced63bad5 is missing from the database.
and generally hardens the implementation around code DB data persistence.

If you have encountered the code missing issue, operators will need to update and resync their nodes. This fix should prevent it from happening in the future.

To resync your node, follow the guide at this link.

Changelog

Full Changelog: 1.34.0...1.34.1

v1.34.0

25 Sep 13:26
100632d

Choose a tag to compare

Release notes

Important

This release is mandatory for all testnet operators for the upcoming Fusaka hard fork.
It includes protocol changes and fixes required to remain in consensus on testnets.

  • Holesky: Wednesday, October 1, 2025 at 8:48:00 AM UTC
  • Sepolia: Tuesday, October 14, 2025 at 7:36:00 AM UTC
  • Hoodi: Tuesday, October 28, 2025 at 6:53:12 PM UTC

Fusaka Testnets Support

  • Implemented Osaka testnet forks (#9321).
  • Updated RPC behavior: eth_getBlobsV1 now returns UnsupportedFork post-Osaka (#9341).
  • Fixed eth_config: ensured ForkInfo is always initialized (#9233).

Stability and Sync Improvements

  • Fixed snap sync edge cases and hanging on Sepolia (#9166, #9338).
  • Corrected blockhash handling and rare invalid block cases (#9340).
  • Improved cancellation handling and graceful shutdown for peers and sync dispatcher (#9173, #9174, #9175).
  • Fixed concurrent storage root issues and reduced reliance on storage root reads (#9209, #9255).
  • Ensured consistent receipts handling and transaction processed counters (#9231, #9270).
  • Fixed issues in Taiko fee handling and default chains config (#9314, #9309).
  • Fixed missed SELFDESTRUCT between transactions (#9343).

Node Improvements

  • Reduced log noise from misbehaving peers and improved shutdown logging (#9156, #9177).
  • Improved logging and config autocompletion (#9159).
  • CLI now displays default values in help output (#9152).
  • Improved deposit request validation (#9199).
  • Enhanced Prometheus metrics in Kute tool (#9077).
  • Better RPC alignment with Geth (#8951).
  • Various performance optimizations: bloom filters, Patricia trie, TrieNode navigation, gas estimation (#9191, #9220, #9221, #9195, #9064, #9210).

These changes strengthen node stability, improve operator experience, and ensure readiness for the Fusaka fork on testnets.

What's Changed

Read more

v1.33.1

04 Sep 10:04
cda59e5

Choose a tag to compare

Release notes

Important

This release replaces v1.33.0 and addresses important issues reported for that version, specifically around eth_config and Receipts handling.

  • Receipts fix: Corrects an issue where receipts were removed for blocks beyond TxLookupLimit even when History pruning was disabled. See #9231.
  • eth_config fix: Resolves a configuration issue reported by the EF Testing team. See #9233.

These changes improve stability and consistency for users relying on long-term history and accurate transaction receipts.

Note

If you were affected by the missing receipts from running v1.33.0 and require full historical receipts, the two workarounds are:

  • Run Nethermind with Sync.FixReceipts set to true for offline patching of missing receipts
  • Delete the receipts folder in the Nethermind database directory (by default, nethermind_db/<network>) for a full download of all receipts while online

For further details, please review the full changelog and linked PRs below.

Full Changelog: 1.33.0...1.33.1

v1.33.0

01 Sep 14:36

Choose a tag to compare

Release notes

This release brings over 340 improvements across 2012 files, ranging from performance enhancements, protocol upgrades, and node-runner improvements.

Nethermind UI (initial release)

image

This version introduces the first iteration of the Nethermind UI (#9090), providing an integrated interface to interact with and monitor your node. This marks the beginning of a more user-friendly experience for running Nethermind.

The UI is enabled when UI health check is enabled: --healthchecks-enabled

If you wish to expose only the UI and not RPC externally, then you can do this using capabilities on HTTP endpoints with the JsonRpc.AdditionalRpcUrls parameter.

For example, if your consensus client is running locally (running in the same instance) and your external IP is 10.4.0.4, then you can expose only the UI using just the http|health capabilities on the external IP binding and want it enabled at HTTP port 80, then binding the other capabilities needed by the consensus client to localhost:

--jsonrpc-additionalrpcurls "[http://10.4.0.4:80|http|health,http://localhost:8551|http;ws|net;eth;subscribe;engine;web3;client]"

This ensures you don't additionally enable your node as a public RPC endpoint while also enabling the UI.

Historical data live pruning

We’ve added automatic history pruning capabilities (#8040), giving operators the ability to keep disk usage under control by pruning historical chain data as the node runs. This feature is disabled by default but can be enabled via configuration. It is an experimental feature, not yet recommended for use on production nodes.

History pruning can be enabled by passing the flag --history-pruning UseAncientBarriers. This will automatically prune any blocks older than the ancient barriers cutoff.

Alternatively, pruning can be enabled in rolling window mode with the flag --history-pruning Rolling. This will prune any blocks older than 82125 epochs, which is around 1 year on the Ethereum Mainnet. The specific number of epochs can be configured with the flag --history-retentionepochs <value> -- the number of epochs must be a minimum of 82125.

History pruning can be disabled with the flag --history-pruning disabled -- currently this is the default behavior, but this may change in the future.

Peering improvements for OP Stack

Several improvements were made to Optimism and OP Stack, including:

  • better handling of finalized/safe blocks #8658
  • a dedicated optimism RPC namespace #8667
  • refined peer management to improve stability and sync performance #8684, #8960

Performance improvements

A wide range of EVM and networking optimizations were introduced, including:

  • Reduced allocations in opcode execution #8687, #8740 and transaction sender recovery #8781, #8824.
  • Warmup and precomputation improvements for EVM instructions #8699, #8836.
  • Faster logging and RPC handling #8879, #8976.
  • RocksDB upgrades and optimizations #8640, #9059.

These changes collectively reduce CPU and memory usage while improving throughput and responsiveness.

Other improvements

Additionally, this release includes important fixes and feature updates:

  • EIP implementations:
    • EIP-7823: Gas accounting #8649
    • EIP-7825: Transaction gas limit cap #8663
    • EIP-7883: ModExp changes #8489, #8964
    • EIP-7907: Contract code size metering #8845
    • EIP-7910: eth_config #8956
    • EIP-7918: Blob base fee updates (#8656, #8935
    • EIP-7934: Execution block size limit #8815
    • EIP-7939: CLZ instruction #8759
    • EIP-7951: RIP-7212 renamed & enabled in Osaka #8753
  • Consensus layer & OP-CL: Split finalized/safe blocks #8658, new RPC methods for Taiko and OP #8874, and refined health checks #8755.
  • Discovery & peering: Faster removal of inactive peers #9037, improved gossip prioritization for peers on the same fork #9043, and fixes for discv4/discv5 edge cases #9046, #9113.
  • Testing & tooling: Revamped Hive/Assertoor test integrations #8711, #8862, new worldstate benchmarks #9051, and initial support for HTML and Grafana reporting #9074, #8960.

What's Changed

Read more

v1.32.4

31 Jul 09:41

Choose a tag to compare

Release notes

Fixed the peer discovery issue

After the Geth v1.16.0 release, there was a sudden drop in discovered Nethermind nodes in network crawlers. This version exposed one edge-case issue, which was much more likely to happen, and is now addressed.

Other improvements

Additionally, a few extra improvements were added to the general discovery phase of finding new peers and ensuring a high peer count can be achieved:

  • #9046: Ensures we reuse the original request identifier when replying, fixing mismatches in ENR messages.
  • #8712: Fixes corrupt snappy message when disconnecting fast due to too many peers.
  • #9037: Speeds up how quickly inactive or problematic peers are dropped.
  • #9043: Prefers peers on the same blockchain fork for gossiping to improve message propagation.
  • #9040: Fixes an issue that could miss some transaction receipts during syncing.
  • #9054: Optimizes peer-distance calculations for faster network operations.

Full Changelog: 1.32.3...1.32.4

v1.32.3

21 Jul 14:45

Choose a tag to compare

Release notes

Important

This release is a mandatory upgrade for all node operators of the Energy Web Chain.

The Zurich hard-fork of EWC is set to the block number 36871000, which is expected to occur in the UTC afternoon of August 5, 2025.

Major highlights

Addressed OOM issues during full and archive sync

More resilient syncing of Archive nodes with better memory management thanks to #8975

Fixed the issue with broken block processing thread when a block is suggested out of order

This is an edge case scenario, but with some consensus clients and under very rare circumstances, it tends to break the processing in #8683

Improved the general performance of a node and its stability

Improved peering and validation performance under heavy P2P demand (snapserving, mempool txs, etc) in #8968

Few extra changes which are improving overall node performance, like #8926, #8947, #8927

Additionally, whenever a node breaks via random database corruption, now Nethermind will shut down to ensure no further corruption of the database is happening, and with proper setup (via Docker, systemd, or in systems like DappNode), it may self-recover in a few minutes; in #8986

Full Changelog: 1.32.2...1.32.3

v1.31.13

15 Jul 11:01
1b54872

Choose a tag to compare

Release notes

Important

This release is a mandatory upgrade for all node operators of the Energy Web Chain.

The Zurich hard-fork of EWC is set to the block number 36871000, which is expected to occur in the UTC afternoon of August 5, 2025.

Full Changelog: 1.31.12...1.31.13