Michael Trainor activity https://gitlab.com/tmike 2026-03-18T04:40:06Z tag:gitlab.com,2026-03-18:5215651629 Michael Trainor commented on issue #39330 at GitLab.org / gitlab-runner 2026-03-18T04:40:06Z tmike Michael Trainor

GitLab Ultimate customer impacted by this. GitLab Support ticket refers (internal).

The customer notices that when a job is being canceled during a helm deployment, it results in the pending-upgrade status, instead of failed. Helm introduced SIGTERM handling in Helm v3.7.0 to ensure the status is failed when given a SIGTERM: https://github.com/helm/helm/pull/9180

The problem seems to be that the Runner does not wait for the processes to handle the signal, it just kills the containers/pods.

tag:gitlab.com,2026-03-18:5215641331 Michael Trainor opened issue #39330: Runner sends SIGTERM to processes but doesn't give time to process signal at GitLab.org / gitlab-runner 2026-03-18T04:33:04Z tmike Michael Trainor tag:gitlab.com,2026-03-16:5206281794 Michael Trainor pushed to project branch tmike-prevent-email-child-pipelines at GitLab.org / GitLab 2026-03-16T03:35:39Z tmike Michael Trainor

Michael Trainor (e4bdbaad) at 16 Mar 03:35

Add test coverage for disabled feature flag behavior

tag:gitlab.com,2026-03-16:5206095185 Michael Trainor commented on merge request !221010 at GitLab.org / GitLab 2026-03-16T01:25:17Z tmike Michael Trainor

@daniel-prause @hfyngvason I'm in the process of implementing the FF as suggested. If that's not necessary, I can drop those subsequent commits. The MR is not in a mergeable state currently as the FF implementation is ongoing.

tag:gitlab.com,2026-03-13:5199559300 Michael Trainor pushed to project branch tmike-prevent-email-child-pipelines at GitLab.org / GitLab 2026-03-13T03:52:53Z tmike Michael Trainor

Michael Trainor (5d0a0b38) at 13 Mar 03:52

Handle nil parent when checking feature flag in fields

... and 2795 more commits