We are running ansible 2.2.1.0 and upgraded tower from 3.0.2 to 3.1.1
After the tower upgrade, when we run a job template, the job fails because the with “SCM Update failed for Ops with ID 46720”
The issue I realized is that the job is looking for a job ID that doesn’t exist.
Example, running the job template will spawn two jobs, 46718 is the ID of the play and 46719 the ID for the scm update, but the play job will fail with the above message looking for job ID 46720 (not 46719) which doesn’t exits.
My SCM (using mercurial) is indeed on a non-standard port (8000) but when the I run an SCM update on the project it completes fine.
Also when running a template the SCM update job that spawns completes successfully but the play job fails because it is checking the wrong job ID.
I’m wondering if this was caused when updating. Maybe a member of the team tried to run a play while I was updating tower and something got messed up in the Tower DB?
It affects every other job run in my setup. I run template X it fails because of prior job (even though I am not doing “update on launch”) then i run it again and it goes through fine.
Only testing it at the moment so not end of the world. I was thinking of just doing a stunnel wrapper maybe.
I dont think this forum supports Tower however, so you would likely have to put a support ticket in with Red Hat to get more info/help. https://access.redhat.com/
I am seeing same - having just installed Tower for the first time and trying to get up and running with the simplest of projects and playbooks.
Under ‘Projects’ I can start an SCM Update using the cloud/arrow icon and it will complete (but ONLY if I select ‘Delete on Update’ but that seems to be a side issue).
However, when I run the job related to my simple playbook, it fails with Explanation: Previous Task Failed >> details; SCM Update failed for {project} with ID 41.
But my jobs listed only go as far as 40 at present. 39 = SCM Update (the one I ran from Projects tab). 40 = the failed playbook run.