Pull requests show negative number -1 after merge

[lifeofguenter]

Description

not sure if this can be reproduced but I have never experienced this in earlier versions.

Gitea Version

1.25.0

Can you reproduce the bug on the Gitea demo site?

No

Log Gist

No response

Screenshots

Git Version

No response

Operating System

No response

How are you running Gitea?

Download binary

Database

MySQL/MariaDB

[wxiaoguang]

It's not related to 1.25, but a longstanding bug ..... due to abused ORM & SQL

And there are other similar issues (duplicate), but I can't find them at the moment (difficult to search ...)

[wxiaoguang]

OK, found one: Pull requests number showing "-1" (negative PR count) #27472

The number of pull requests is different from the actual total in the list. #35168

Negative numbers in menu pull request #34927

The total number of PRs in the repositorie's PR label column is incorrect #34548

[lifeofguenter]

I am re-opening / keeping this one open. It really got worse. I can even make it count more in negative by closing + reopening the PR.

And this never happened prior to upgrading to 1.25.0

[lifeofguenter]

switched to postgres but it did not solve the issue, now PR counter does not work.

[mike12806]

I am re-opening / keeping this one open. It really got worse. I can even make it count more in negative by closing + reopening the PR.

And this never happened prior to upgrading to 1.25.0

Agree - I've had this happen sporadically previously, but now it's happening frequently. I can force a stats resync and it'll fix it (or restart the gitea container), but then it happens again immediately on the next PR

[lunny]

Could this issue be reproduced consistently?
I found an example on https://gitea.com/gitea/gitea-mcp where num_pulls is incorrect but num_closed_pulls is correct.

Could someone share the related SQL queries and other relevant logs when this issue occurs?

[lunny]

Yes. For the first, I sent #35819 used the same method as you.
Thank you very much. I think this is the real problem and I will update #35819 to include this fix.

[lunny]

I have updated #35819 to include the second fix.