Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

rustbot sends the same pings multiple times #1738

Open
RalfJung opened this issue Oct 30, 2023 · 5 comments
Open

rustbot sends the same pings multiple times #1738

RalfJung opened this issue Oct 30, 2023 · 5 comments
Labels
A-mentions Area: PR auto mentions bug Something isn't working

Comments

@RalfJung
Copy link
Member

This happens fairly regularly, here's a recent example:

I thought rustbot was supposed to remember who already got pinged, and not ping them again when the PR changes?

@ehuss ehuss added the A-mentions Area: PR auto mentions label Oct 30, 2023
@RalfJung
Copy link
Member Author

RalfJung commented Dec 1, 2023

Here's another double-ping: rust-lang/rust#118500

@ehuss ehuss added the bug Something isn't working label Jan 21, 2024
@RalfJung
Copy link
Member Author

RalfJung commented Dec 2, 2024

rust-lang/rust#133701 is another example of this.

@RalfJung
Copy link
Member Author

This is still happening: rust-lang/rust#137330.

Speaking of ping groups, is there anyone to ping who maintains triagebot? Is it even still maintained?

@ehuss
Copy link
Contributor

ehuss commented Feb 20, 2025

For rust-lang/rust#137330, it looks like the initial open hook was unable to finish in time, and I would guess it was unable to save the database entry. The time from the first to last log message is 7.35s. There is a 5second gap in the middle of handling assignments. The only thing I see happening during that time is the call to is_new_contributor, so I'm guessing that the response there was just slow. We may want to consider a different approach for detecting new contributors.

It could also have been a delay talking to the database. We have been trying to diagnose similar problems at https://rust-lang.zulipchat.com/#narrow/channel/224082-triagebot/topic/dropped.20assignment. It could be the database was overloaded (like from activity from perf).

We may also want to consider a different approach for handling webhooks in general. As triagebot adds more and more functionality, it makes it harder to respond to everything within 10s.

Another thought I had is that maybe there is some slow-running code in some handler somewhere, and tokio isn't balancing those tasks in parallel? The logs don't say much (there were 4 hook events, but none of them did anything significant). That seems unlikely, though.

is there anyone to ping who maintains triagebot?

You can ask on #triagebot if there is something that needs discussion outside of an issue or is urgent. Otherwise, we all get notifications from your comments on issues.

Is it even still maintained?

I don't know what your expectations are here. As you can see from the history, there is very regular development going on. New people were also just added to the team.

@RalfJung
Copy link
Member Author

Thanks for the reply and for taking a look. I felt like calling into the void in this issue, good to know that was not the case. :)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
A-mentions Area: PR auto mentions bug Something isn't working
Projects
None yet
Development

No branches or pull requests

2 participants