CLI: Return non-zero exit code when deployment fails in nomad run
#11550
Conversation
There was a problem hiding this comment.
Hi @lukas-w! This is looking good! Once you've addressed the comments here, if you'd like to add a changelog item for the improvement that'd be great. If that's a confusing mess (understandable), let me know and I'll get that added.
Co-authored-by: Tim Gross <tgross@hashicorp.com>
|
Thanks for the review @tgross! I added a changelog entry via e78fd43 but included the |
|
LGTM. The changelog entry is perfect. Thanks @lukas-w! This will ship in the next planned patch release. |
PR #11550 changed the job stop exit behaviour when monitoring the deployment. When stopping a job, the deployment becomes cancelled and therefore the CLI now exits with status code 1 as it see this as an error. This change adds a new utility e2e function that accounts for this behaviour.
PR #11550 changed the job stop exit behaviour when monitoring the deployment. When stopping a job, the deployment becomes cancelled and therefore the CLI now exits with status code 1 as it see this as an error. This change adds a new utility e2e function that accounts for this behaviour.
|
I'm going to lock this pull request because it has been closed for 120 days ⏳. This helps our maintainers find and focus on the active contributions. |
This PR changes the
nomad runcommand (and other commands using themonitor.monitorfunction) to return a non-zero exit code when monitoring is enabled and deployment fails just likelevant deploydoes. This enables CI scripts to easily detect deployment failures. I chose return code1to match the existing documentation stating0on success,2on scheduling problems, and1for any other failures. A dedicated return code for deployment failures may makes sense though.I also noticed and fixed a supposed typo from #10661 via 24fbb5e: The monitoring code was using
DeploymentStatusDescriptionBlockedinstead ofDeploymentStatusBlocked.Fixes #11082, fixes #10994.