This is what I’ve noticed so far. I also agree that tagging shouldn’t skip any CI. Normally, tags are used to determine which CI jobs can start running, not to affect CI jobs that have already run.
Best regards, Zhen ---- 回复的原邮件 ---- | 发件人 | Mihai Budiu<[email protected]> | | 日期 | 2026年08月26日 06:55 | | 收件人 | dev<[email protected]> | | 抄送至 | | | 主题 | Re: Skipped CI jobs | Really? Then the CI is a bit too smart. The runs are not that long, we should probably always run them if needed. Mihai ________________________________ From: jensen <[email protected]> Sent: Tuesday, August 25, 2026 3:54 PM To: dev <[email protected]> Subject: Re: Skipped CI jobs I agree with Alessandro. Regardless of whether the current CI run has finished, adding the "LGTM will merge soon" label skips all CI checks except for a few mandatory checkers. That’s why I usually add the label first and then squash-merge, to avoid missing the latest CI status. I’ve brought this up before, but I’m not sure how to fix it. Best regards, Zhen ---- 回复的原邮件 ---- | 发件人 | Mihai Budiu<[email protected]> | | 日期 | 2026年08月26日 01:08 | | 收件人 | [email protected] | | 抄送至 | | | 主题 | Re: Skipped CI jobs | After I approved, it still skipped many CI jobs. Mihai ________________________________ From: Julian Hyde <[email protected]> Sent: Tuesday, August 25, 2026 9:58 AM To: [email protected] <[email protected]> Subject: Re: Skipped CI jobs Are you referring to the GitHub policy of not running CI for PRs from first-time contributors? That policy seems understandable to me, especially during the crypto-mining days, because it prevents people from using GitHub for free CPU. GitHub has also been flaky over the last few months. Last week I was hit by an issue, which lasted for several hours, where pushes to main didn’t trigger a CI job. (It’s not that the job failed. There was no job to restart. After GitHub resolved the issue, I had to force-push several commits to trigger CI.) > On Aug 25, 2026, at 9:48 AM, Mihai Budiu <[email protected]> wrote: > > I wonder why CI skips so many jobs. Some of them seem unjustified: > https://github.com/apache/calcite/pull/5219 > >
