Details
-
Type:
Improvement
-
Status: Closed (View Workflow)
-
Priority:
Minor
-
Resolution: Fixed
-
Component/s: jclouds-plugin
-
Labels:None
-
Environment:Jenkins 2.229
jclouds plugin 2.17
-
Similar Issues:
Description
When I use the jclouds plugin to spawn instances at DigitalOcean and EC2, I noticed that the single-use slaves are never removed. They are marked offline as soon as the job finishes, but they remain offline in Jenkins and they are active with the cloud provider.
As soon as the job finishes, I see this in the logs:
2020-04-02 18:30:42.634+0000 [id=3070] WARNING j.p.j.c.JCloudsOneOffSlave$JCloudsOneOffSlaveDisposer#tearDown: Taking single-use slave [COMPUTER_NAME] offline.
I let it sit for two hours and nothing changed. I had to delete the agent manually from inside Jenkins' build executors listing.
Am I missing a step? I'll gladly turn on more debug logging if that would help diagnose the issue.
Attachments
Activity
Please try this build. If this fixes the problem for you, I will create a new release.
Thanks for the quick fix. I'll try to get a look at this as soon as I can.
As I got no further feedback, I now released v1.18 and close this issue.
Feel free to reopen this issue, if you are not satisfied with the solution.
Note: A slight delay (~ 1min) is still happening between marking a slave offline and actually deleting the instance in the cloud. This works as designed.
Well this turned out to be a bug. The retention time for the single-use slaves was never reset to zero.
Therefore, the normal RetentionStrategy handler never touched them. Will fix this asap.
-Fritz