High Sequence Number Elimination and No Redundant Sequence Number by Node

[l-monninger]

Summary

• RFCs: Link to RFC, Link to RFC, or $\emptyset$.
• Categories: any of protocol-units, networks, scripts, util, cicd, or misc.

Changelog

Testing

Outstanding issues

[andygolay]

Could you please fill out the PR description with a little more context? I'm not very familiar with this specific issue.

I'm seeing a couple conflicting files and also unclear on why to merge into l-monninger/high-sequence-number-dos.

[l-monninger]

@andygolay This is PR relative to #597. Comments on this means of preventing this kid of DOS are there.

[andygolay]

@andygolay This is PR relative to #597. Comments on this means of preventing this kid of DOS are there.

Got it. So the gc prevents redundant sequence numbers? Seems good as long as the suzuka-full-node-malicious check is passing.

Checks / unit-tests is failing with

test transaction_pipe::tests::test_pipe_mempool_with_duplicate_transaction has been running for over 60 seconds

Is that expected? It passes in #597.

[l-monninger]

Checks / unit-tests is failing with

test transaction_pipe::tests::test_pipe_mempool_with_duplicate_transaction has been running for over 60 seconds
Is that expected? It passes in #597.

That's unexpected. I will investigate.

[l-monninger]

@andygolay unit tests should be resolved, but e2e test just flaked.

[mzabaluev]

It would be good to give a description of how these slots are going to work.
Also, since this will affect transaction submission, this should make it into a MIP documenting the behavior, but perhaps this can be done later.

[andygolay]

Checks / suzuka-full-node-malicious is failing