Brooks

> It seems like `clean_accounts` is written with the assumption that a snapshot _already_ exists at `last_full_snapshot_slot`, rather than is being currently created. Yah! One thing that is not necessarily...

> do we have any metrics that track uncleaned zero lamport accounts? Maybe? Not that I'm aware of. That's probably a good thing to check before merging this.

I'm going to run this against mnb for a while to ensure it's safe.

> I'm going to run this against mnb for a while to ensure it's safe. I ran this branch for ~4 days without issue (until the mnb restart happened). I...

Tagging the folks that already commented over in the other issue: @jstarry @taozhu-chicago @joncinque Please tag in others that would be valuable for the discussion. TIA!

> 2. when the state hits that size, the lowest accounts by lamport balance compress - gets put into a Merkle tree Sounds like this would require scanning all the...

I added the `noCI` label so buildkite doesn't need to churn on this PR before it realizes nothing needs to be done.

@diman-io From [your comment on the other issue](https://github.com/solana-labs/solana/issues/26259#issuecomment-1169116655), it looks like you are running v1.10.26. Is that correct?

@diman-io Do you have a node on testnet? If yes, did https://github.com/solana-labs/solana/pull/27461 fix your issue? Edit: Also, you can pull in this change for v1.10 on mnb: https://github.com/solana-labs/solana/pull/27650

@jeffwashington I have to resolve why some `local-cluster` tests are failing, but I think this PR is good enough to review for direction/design. I'll mark the PR as "ready for...