akula VS lodestar

Back in November, Artem Vorotnikov (Akula developer) made this tweet:

On a related note, I'm also keeping tabs on Akula, because a Rust-based client is a damn sexy proposition for running a performance-sensitive node 24/7.

It certainly seems it will be remembered as a major success story for open p2p protocols on the global Internet of our time.

A great multitude of developers and enthusiasts belonging or contributing to diverse teams spread across the world: developing, debating, and collaborating for years to arrive at the big event.

And it's all been done very much in the public view:

https://weekinethereumnews.com/

https://hackmd.io/@benjaminion/eth2_news

https://github.com/ethereum/pm

https://www.youtube.com/c/EthereumFoundation/videos

Consensus Clients:

https://github.com/sigp/lighthouse#readme

https://github.com/ChainSafe/lodestar#readme

https://github.com/status-im/nimbus-eth2#readme

https://github.com/prysmaticlabs/prysm#readme

https://github.com/ConsenSys/teku#readme

Execution Clients:

https://github.com/akula-bft/akula#readme

https://github.com/hyperledger/besu#readme

https://github.com/ledgerwatch/erigon#readme

https://github.com/ethereum/go-ethereum#readme

https://github.com/NethermindEth/nethermind#readme

https://github.com/status-im/nimbus-eth1#readme

Akula. Ethereum client written in Rust, based on Erigon client architecture.

I think I found it. This pull request helped me: https://github.com/ChainSafe/lodestar/pull/3782

Somer's guides are awesome and extremely thorough, but they help you learn to build a validator from scratch. In practice, I always download client binaries rather than building from scratch. You can find the lodestar releases page here.

https://github.com/ChainSafe/lodestar 63 contributors

My mainnet lodestar node crashed today. Thankfully, the validator process switched over to my backup lighthouse node without missing a single attestation.

Daily Goerli: I've been periodically running into some issues when updating nethermind which require force killing the Nethermind.Runner process. Recently had an issue where geth in lodestar-geth thought it was waiting for updates from lodestar, but lodestar was reporting connection refused. A restart of geth fixed it. I'm currently running into No state found for id 'head' in all lodestar instances, which was supposed to be fixed by this PR. I also recently encountered a long stack error from erigon in prysm-erigon when closing to update, but restarting both clients allowed sync to resume.

The following sections outline how to configure an execution-layer client and a consensus-layer client so that they have everything in place to execute the entire merge already in the genesis block. Here, we'll use Geth and Lodestar.

Consensus Layer client ✅ Teku v22.11.0: https://github.com/ConsenSys/teku/releases/tag/22.11.0 ✅ Lodestar v1.2.2: https://github.com/ChainSafe/lodestar/releases/tag/v1.2.2 🟡 Lighthouse: coming soon 🟡 Nimbus: coming soon ❌ Prysm: advised to switch to other clients