Consensus-layer Call 139

[ralexstokes]

Consensus-layer Call 139

prev: call 138

Meeting Date/Time: Thursday 2024/8/8 at 14:00 UTC
Meeting Duration: 1.5 hours
stream

1. Electra
   • Release v1.5.0-alpha.4 consensus-specs#3853 alpha.4 soon!
   • devnet-2
   • Using mempool for blob distribution
     • engine_getBlobsV1: Define engine_getBlobsV1 execution-apis#559
     • P2P clarification: P2P clarifications when introducing engine_getBlobsV1 consensus-specs#3864
   • mplex deprecation: Deprecate mplex consensus-specs#3866
   • EIP-7688 SSZ StableContainer inclusion in Pectra
2. PeerDAS
   • any development updates?
   • PeerDAS without sampling PR: Spec without peer sampling consensus-specs#3870
3. Research, spec, etc
   • consolidations and sync committees
   • Add quic ENR entry: p2p-interface.md: Add quic ENR entry. consensus-specs#3644
   • ProbeLab update
4. Open discussion/Closing remarks

[guillaumemichel]

I would like to present ProbeLab's weekly reports, providing measurements on Beacon nodes participating in discv5. I will explain the methodology, show the plots and discuss key results. We would appreciate your feedback and improvement suggestions.

[timbeiko]

On #1104, we agreed to make a final decision about ethereum/execution-apis#559 on the next ACDC, pending a PR to specify the expect "honest validator" usage of the new endpoint.

[AgeManning]

We just wanted to bring up mplex deprecation again. It's becoming a bit of a liability. We are thinking of setting a date in the near future, where Lighthouse will remove support for it. We want to make this date reasonable so that it doesn't impact other clients that may still depend on it.

There is a PR to comment on this here: ethereum/consensus-specs#3866

[ppopth]

Bringing back QUIC PR into discussion ethereum/consensus-specs#3644

[lucassaldanha]

@ralexstokes Teku team is mostly away this week. I am not sure if we will have anyone at the call (I won't be awake for it). I'm gonna drop some comments/notes here:

• devnet-2: Teku seems to be holding up well so far. We haven't heard of any recent issues with Teku. Please ping me on Discord if you find anything weird we are doing or if you have any concerns.
• engine_getBloblsV1: nothing controversial from our point-of-view (assuming that @tbenr 's concerns P2P clarifications when introducing engine_getBlobsV1 consensus-specs#3864 have been considered valid and we all agree on it).
• PeerDAS: I don't have a lot to update. Last I spoke with Dmitrii, we were preparing for a devnet-1 relaunch next week. We have some work to do on Megadata V3 as well.
• MPLEX deprecation: In Teku, we have made some progress with Yamux, it is in an experimental phase atm. We should be able to prioritize it and make it the new standard in the near (not so near) future. @AgeManning what timeframe are you guys working on for deprecation?

I think that's the gist of it. Maybe we get lucky and someone from Teku joins the call. But at least you guys have a rough idea of what's going on with Teku :)

I'll catch up with the recording tomorrow and follow-up on any questions and/or actions from the meeting. Cheers!

[etan-status]

• EIP-7688 SSZ StableContainer - last ACDC the sentiment I got was that the feature is generally seen useful, but there were prioritization concerns given the EL instabilities back then at devnet 2, and that some team members working on peerDAS may not leave enough time for 7688. Can we assess this more cleanly now (for devnet 3)?
  • Rocket Pool statement: https://x.com/KaneWallmann/status/1816729724145795258 (alternative without X login)
  • consensus-spec-tests updated to include SSZ data structure tests: Adopt EIP-7688: Forward compatible consensus data structures consensus-specs#3844 (comment)
  • For Pectra, the proposed SSZ scope is only EIP-7688 stable generalized indices; not SSZ transactions and other ideas (ELs are not ready at this time, besides Ethereumjs that has an SSZ transaction devnet available)
• Consolidations: What happens when the consolidation source is part of current / next sync committee? Should the consolidation be delayed until pending validator duties are fulfilled? Should the protocol prevent validators from getting new duties beyond any pending consolidations and exits?

[AgeManning]

@lucassaldanha - Thanks for the update. We haven't set a fixed timeframe for its deprecation. Ideally sooner rather than later, but don't want to jeopardize any connections with clients.
We currently support TCP/Yamux and QUIC. If Teku can support either of these, then we can remove mplex.

If you can give us a rough time-frame for shifting off mplex, then we can align with the deprecation.

[etan-status]

• Validator operation requests: Should we group the various validator request fields within a sub-Container to match the EL design with a requests_root more closely, and also to be able to group the various requests if there actually is a new one at some future time? see engine: unify request objects execution-apis#565 (comment)

[barnabasbusa]

Quick question, should we bring back testing call that we used to have during deneb/merge days?

[barnabasbusa]

Could we circle back to discussing ethereum/consensus-specs#3800 ?

[akashkshirsagar31]

Podcast (audio only) - https://open.spotify.com/episode/6mHHIbNWfbMssTF2NCSBUo?si=KDYLFz1_T_GWd9-fjAqSpg

[ralexstokes]

closing in lieu of #1129