.play() on job returns 500 even though job is started successfully

[wardbox]

Description of the problem, including code/CLI snippet

I'm starting a job via the API and getting a 400 error even though I can see the job getting kicked off and running in the UI. Because of this I have no way of knowing if the error is reliable.

job = project.jobs.get(job.id, lazy=True)
job.play()

Expected Behavior

200 OK - Job played
Job runs and completes successfully

Actual Behavior

400 Bad request - Unplayable Job
Job runs and completes successfully

Specifications

• python-gitlab version: 4.7.0
• API version you are using (v3/v4): v4
• Gitlab server version (or gitlab.com): gitlab.com

[JohnVillalovos]

Hi,

Can you try:

gl.enable_debug()
<your prep work here>
job = project.jobs.get(job.id, lazy=False)
job.pprint()
job.play()

And hopefully will have more info on what is going on?

And you are sure that the job isn't already running before you do the play?

[JohnVillalovos]

Possibly useful: https://gitlab.com/gitlab-org/gitlab/-/issues/276938

Which mentions needing to use: when: manual in the job definition.

[wardbox]

I'll try enabling the debug and getting you the result. I confirmed 100% that the job was not running beforehand and my job does use the when: manual setting which is where I actually started on my journey to figure out what's going on!

[wardbox]

I'm not sure what changed, but it is not failing now.

[JohnVillalovos]

I'm guessing some issue with Gitlab. python-gitlab is fairly simple in sending the request so I doubt it was the cause of the issue.

I'll go ahead and close this.

[nejch]

@JohnVillalovos agreed, especially 500 indicates something on the GitLab side.