Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Sometimes approves the wrong commit

See original GitHub issue

There have been two circumstances where a normal @bors r+ approved an old commit:

My hunch is that somehow bors missed the push notification, and the state.head_sha doesn’t get updated to the latest version.

I think Homu should never approve a commit that is not the latest commit. I imagine it should be check what the latest commit is when approving instead of assuming that the database is in sync. There’s definitely risks about race conditions here that may not be solvable, but some extra effort might make it more resilient.

What’s scary is that this may be happening and nobody notices. These two instances only got noticed because the old commit failed in CI.

Issue Analytics

State:
Created a year ago
Reactions:3
Comments:8 (8 by maintainers)

Top GitHub Comments

1reaction

notriddlecommented, Jul 2, 2022

if I understood it correctly where as bors-ng is stateless and homu stateful

Unfortunately, no. There are three borses, not two.

bors — https://github.com/graydon/bors — is stateless
homu — https://github.com/rust-lang/homu — is stateful. It exists because of GitHub’s rate limits.
bors-ng — https://github.com/bors-ng/bors-ng — is also stateful. It exists because self-hosting homu is too complicated. Since literally anybody can sign up their repository at any time, bors-ng also has to worry about rate limits.

0reactions

pinkforestcommented, Jul 17, 2022

Raised #178 for suspected general WebHook delivery instability which addressing commit mixups alone would not fix (this ticket)