Consensus-layer Call 131

[ralexstokes]

Consenus-layer Call 131

prev: call 130

Meeting Date/Time: Thursday 2024/4/4 at 14:00 UTC
Meeting Duration: 1.5 hours
stream

1. Deneb
   • Clarify voluntary_exits gossip topic after Capella consensus-specs#3638 (comment)
   • Network outage with Bloxroute relay
   • any other items?
2. Electra
   • Discuss currently included EIPs, including status of spec and any open questions
     • 6110
     • 7002
     • 7251 (@ethDreamer to give update)
     • 7549
   • Client team updates on current implementation status of each included EIP
   • Discuss any additional EIPs to include in Electra scope for early devnets
     • 7547 - Inclusion lists
     • anything else?
   • Agree on final scope for pectra-devnet-0
     • estimates on first devnet timeline?
3. Research, spec, etc
   • PeerDAS
     • spec status update?
     • implementation status update?
     • start implementer’s call?
4. Open discussion/Closing remarks

[dapplion]

Confirm that all clients use Deneb fork id for exits gossip topics on Deneb.

• Clarify voluntary_exits gossip topic after Capella consensus-specs#3638 (comment)

[jrudolf]

@ethDreamer to give an update on maxEB

[terencechain]

It may be useful to bring closure to the Bloxroute BDN issue. We left the matter unresolved last Thursday, pending a post-mortem report.

Today's update:

• I've counted, incident started on Mar 18 12:25:55, ended on Mar 27 17:51:34. Total 240 blocks missed
• Blockroute post mortem: https://gist.github.com/benhenryhunter/687299bcfe064674537dc9348d771e83
• Michael's response: https://twitter.com/sproulM_/status/1773853486373130708
• Since then, LH and Prysm have updated their implementations to allow the Beacon-API submission of blobs even if a block has been observed over P2P, regardless of whether the equivocation flag is set to true
• My general sentiment is that this issue could have easily been prevented, and the urgency has now shifted towards thoroughly considering how to prevent it in the future

[lucassaldanha]

Electra Testnets: I'm wondering if there are any plans for an Electra testnet for EIP-7002 and EIP-6110? I believe those two EIPs could be tested together and there is no need to feature flag them (basically both enabled as part of Electra fork).

[ralexstokes]

@lucassaldanha yeah, i think we should push for electra testnets with all 4 included EIPs, and whatever we can hopefully agree to on today's call!

[tbenr]

Unfortunately I can't join the call today. Just sharing here teku's view about 7547, as shared yesterday on discord.

The design seems a bit unclean so far. My concerns are that, if we go for something like this, then ePBS\ET could change the context that IL design will be affected. Moreover, recent mainnet issues with block production and relay made me feel we should prioritize the CL-EL block separation, which will set a more solid base to build IL on top of.
I'd keep it outside Electra and allocate time to ePBS related efforts.

[realbigsean]

Since then, LH and Prysm have updated their implementations to allow the Beacon-API submission of blobs even if a block has been observed over P2P, regardless of whether the equivocation flag is set to true

still a WIP on LH :) aiming for inclusion in our next release

[timbeiko]

Closed in favor of #1010