-
Notifications
You must be signed in to change notification settings - Fork 319
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Flood Docker slaves IDLE #54
Comments
I also observe the same comportment and haven't been able to track-it down. In fact, I'd like to be able to enable this as a feature, so Jenkins could "preload" the slaves (within the max container or max instance limit), reducing the cloud plugin overhead when starting a lot of jobs at the same time. (Jenkins 1.567, Docker 1.0.0 on CentOS 6) |
I need to do more investigating, which will probably involve building the plugin myself. But I believe what is happening is that jobs with no label restriction are causing a docker container to be provisioned. The logic in the DockerCloud is to use the first container when a null label is requested, thus they are all the same container getting provisioned. However, for some reason, Jenkins thinks the provisioned slave is not satisfactory and will not send the job to it. Even worse, the Docker retention policy will not terminate an idle slave if it has not done work. |
* Fix provisioning a slave on a job for 'any' slave Resolves cause of jenkinsci#54, jenkinsci#43
Here is a question for you. We are discussing on #59, whether it is worth having the "Utilize node as much as possible"/"Tied jobs only" option on containers. Alternatively should it be hard coded one way or another? |
Resolves effect of jenkinsci#54 Also fix DockerTemplateTest This was broken before by jenkinsci#46, but also needed to be updated for my changes
The plugin advertised supporting untied jobs but then all slaves were EXCLUSIVE Fixes the root cause of jenkinsci#54
Thanks alot for these fix ! can't wait to test them :) !!! |
Fixed in 0.7 release. |
Hi,
I got some flood in the compilation list:
(It is just a part of the screen...there is more)
Here is the jenkins logs:
And after some times this appear on the log:
If I look on the slaves informations, they are all created by SYSTEM and attached to none project.
The container is only used by one project that launched every 12 hour.
Every slaves are and stay connected:
I use docker 1.0.0 on debian jessie.
My Jenkins is ver. 1.564 on windows with docker plugin 0.6.2.
I usually restart docker service to make them disappear.
I can not tell when it happen...
The text was updated successfully, but these errors were encountered: