All Core Devs - Consensus (ACDC) #145, October 31 2024

[abcoathup]

Agenda

Consensus-layer Call 145 · Issue #1185 · ethereum/pm · GitHub moderated by @ralexstokes

Summary

All Core Devs - Consensus (ACDC) #145, October 31 2024

ACDC #145 summary Action Items Reminder: ACDC on 14 Nov is canceled due to Devcon Review the merged PRs below for the next consensus-specs release EL teams: can we flexibly deploy Osaka so that we have the option to deploy a CL fork and EL fork separately during testing? Check out the mekong testnet: https://mekong.ethpandaops.io/ Enjoy Devcon Summary Started with an overview of pectra-devnet-4 Devnet is going well; a few minor open issues but under review by the relevant c…

Recording

https://www.youtube.com/live/KMLqv60xg9w?t=89s

Additional info

Ethereum All Core Developers Consensus Call #145 Writeup | Galaxy by @Christine_dkim

[ralexstokes]

ACDC #145 summary

Action Items

• Reminder: ACDC on 14 Nov is canceled due to Devcon
• Review the merged PRs below for the next consensus-specs release
• EL teams: can we flexibly deploy Osaka so that we have the option to deploy a CL fork and EL fork separately during testing?
• Check out the mekong testnet: https://mekong.ethpandaops.io/
• Enjoy Devcon

Summary

• Started with an overview of pectra-devnet-4
  • Devnet is going well; a few minor open issues but under review by the relevant client teams
• Then, turned to pectra-devnet-5
  • Addressed some open PRs that are in a place to merge into the specs
    • Networking attestation support: Separate type for unaggregated network attestations by arnetheduck · Pull Request #3900 · ethereum/consensus-specs · GitHub
    • Update to networking timeouts: p2p: Deprecate TTFB, RESP_TIMEOUT, introduce rate limiting recommenda… by arnetheduck · Pull Request #3767 · ethereum/consensus-specs · GitHub
  • Did a temperature check on the updates to the requests structure in the communication b/t the CL and EL; work is under way
  • Turned to an open question around the blob base fee computation as we change the blob target, which becomes relevant with EIP-7742 and the intent to raise the blob target in Pectra
    • There are a spectrum of options here with an “endgame” solution that would be somewhat involved relative to the current EIP-7742 specification
    • Agreed to go with a middle ground solution so that we don’t delay Pectra and still have a similar amount of responsiveness to the blob base fee
  • Then discussed a bug fix to the consolidation flow under EIP-7251; PTAL: eip7251: Add missed exit checks to consolidation processing by mkalinin · Pull Request #4000 · ethereum/consensus-specs · GitHub
  • Asked for updates around the EIP-2537 gas benchmarking; but no one working on that was on the call
• And then turned to PeerDAS and blob scaling
  • peerdas-devnet-3 had a 100k slot reorg; the devnet has been deprecated
  • No one working on that devnet was on the call, but investigation is ongoing
  • Next, raised the point to go ahead and formally place EIP-7594 into the fulu specs
    • No opposition on the call; although, there’s a question around how to isolate PeerDAS testing from Osaka testing on the EL
• Then turned to longer-term R&D
  • New proposal to introduce execution layer requests for withdrawal credential updates: Add EIP: Withdrawal Credential Update Request by lucassaldanha · Pull Request #9005 · ethereum/EIPs · GitHub
  • Probelab had a quick update around some upcoming network bandwidth analysis via crawling of the disv5 network; expect results soon.
• And concluded with an announcement of the launch of the mekong public testnet, based off of pectra-devnet-4 intending for public preview of the upcoming Pectra hard fork
  • https://mekong.ethpandaops.io/
  • More details around running nodes, accessing the testnet, etc. coming in a future blog post!