Andrew Bayer
Andrew Bayer
/assign @lbernick
https://github.com/tektoncd/pipeline/pull/5134/commits/6275baa83e1406d2545094f87bb1c276a3ff0597 changes handling of `Timeouts.Tasks` and `Timeouts.Finally` to also be done via signaling - I _think_ (or at least hope!) that this is the comprehensive solution. =)
Ok, an annoying quirk of the "just use `TaskRunSpecStatusCancelled`" approach vs the distinct `TaskRunSpecStatusTimedOut` - depending on which of the `TaskRun` or `PipelineRun` gets reconciled first after the overall `PipelineRun`...
> If we swap the TaskRun timeout to be just based on the PipelineTask timeout, it seems like this would only happen if the PipelineTask timeout happened to expire at...
I've marked this as WIP until we nail down the ordering between this change and #5150.
cc @lbernick, @jerop, @afrittoli, @vdemeester - with #5224 merged, I've updated this PR accordingly, and I think it's ready to go.
Whoops - missed a couple things! Fixing.
/hold ping @lbernick as a reminder that you said you'd take a look at my attempt to rework after-pipeline-timeout `TaskRun`/`Run` creation handling. =)
Adding this to the v0.39 milestone, because I'd really like to get this done. cc @pritidesai @lbernick for updated reviews. Thanks!