Discussion thread for EIP-8080
I have a few questions about this EIP:
Firstly, the entire EIP proposal hinges on the assumption that the consolidation queue will often be empty while the exit queue is congested: “When there is an imbalance between exit and consolidation queues favoring the latter, as has been the case since the introduction of the consolidation feature, this becomes an appealing way to speed up withdrawals.” This observation has been true so far whilst Pectra adoption is still low (97.74% of all Mainnet validators are still 0x01 validators, 01.04.2026, as per https://pectrified.com/mainnet).
However, consolidations are still a relatively new feature, and a large portion of validators has not yet consolidated. It seems likely that consolidation demand will increase in the short and mid-term as parties adopt Pectra. In that world, the “idle consolidation queue” assumption may not hold.
Secondly, allowing exits to use the consolidation queue creates a policy tension. The Ethereum Foundation has consistently signalled that consolidation is a desired network outcome (reducing validator count, improving efficiency). If that’s the case, consolidations should arguably be prioritised and kept frictionless, not put into direct competition with exits. By allowing exits to consume consolidation churn, this EIP introduces a race between two fundamentally different intents:
Even if total churn remains constant, reallocating consolidation capacity to exits risks:
-
Slowing down consolidation precisely when adoption ramps up
-
Creating unpredictable dynamics depending on relative queue lengths
In other words, we move from a system with implicit prioritisation of consolidations to one where exits can crowd out consolidations?
Is this tradeoff intentional? And if so, why is improving exit liquidity considered equally (or more) important than consolidations, given the network’s long-term scaling goals?
Finally, I understand that the technical loophole which Pectra introduced must be fixed and I would be keen to know why EIP-8071 was rejected?
Please @fradamt correct me if I’m misunderstanding any of this, and I’d also be happy to discuss this more synchronously too.