On May 11th, the beaconchain began experiencing issues reaching finality for 3 epochs. This does not affect end users, does not halt transactions, and is an issue that is primarily a concern for stakers and developers. The issue reoccurred on May 12th for 8 epochs. The chain is finalizing again. It’s possible the issue could reoccur before client updates are implemented.
The issues have been resolved by Prysm & Teku’s client updates. Please update your client software.
POST-MORTEM (17 May):
(Direct link: https://offchain.medium.com/post-mortem-report-ethereum-mainnet-finality-05-11-2023-95e271dfd8b2)
https://twitter.com/OffchainLabs/status/1659011594272030720
A block being ‘final’ means that that block can’t be reverted at a later date. The time between a block proposal and finalization is a short window of opportunity for reorgs. Finalization means that a block is certain and won’t change.
Did Ethereum go down?
No. Transactions continued to be processed as normal.
Which clients were affected?
All clients affected. Per Danny Ryan, many CL clients on ethereum experienced high load due to an exceptional scenario. Lighthouse did not because they took a different design approach.
Is it MEV-boost or relays that caused the problem?
No, they don’t seem to be involved.
(Firefox users may have trouble viewing embedded tweets):
The issue began to prevent finalization again at block 200751 and continued through 200758. The current status is that fixes are being worked on and a post-mortem will be published once those fixes have been implemented in client software. Client updates have been released for Prysm and Teku - please update your software if you run these clients.