Jenkins Integration for Pipeline Stage


#1

We have a use case that requires us to trigger a job on our Jenkins server from Spinnaker (think running tests), wait for the results, and then continue to another pipeline when the results are successful. I’ve been following the instructions here https://www.spinnaker.io/setup/ci/jenkins and have configured hal with our jenkins “master”. I see the master showing in the UI when creating a new Jenkins stage but whenever I select the master, the job list seems to just be loading forever. I understand that most Jenkins instances have lots of jobs (ours is not so different) and maybe this takes a while, but I’m really not sure what to expect. I’ve tried modifying the json to just define a property called job (was just taking a stab in the dark) but it seems to still just want to spin. After waiting over 10 minutes with no change (just the loading bar), I figure it’s time to try and get more information for what steps are recommended to take next.


#2

It does seem like bad UI handling, but I probably should’ve started with the logs and realized that the problem was it not being able to resolve our jenkins dns (don’t post things at 1am :slight_smile:) . After fixing this, the jobs showed up instantly.


#3

Good to hear, Brian. I agree the problematic configuration should definitely be handled better.

As it stands, we relegate configuration problems to the “operator”, who typically works with halyard. This is different from the “user” (e.g. app dev team), for whom we don’t expose configuration problems.

So for the operator… it turns out we don’t have a halyard config validator for Jenkins. I’ve created an issue for it.