First, let me give some context:
- We use Jenkins for the CI of the Hibernate projects.
- We have a master/slaves setup on AWS with slaves being terminated when
they don't have anything to do so slaves come and go and we can't count on
- It's a rather new setup as we used a couple of persistent slaves before.
Lately, we have seen that when we push something to a given GitHub branch,
all the jobs targeting this repository are launched even if they target a
completely different branch and there are no changes at all in this branch.
This results in a lot of useless jobs running.
>From https://issues.jenkins-ci.org/browse/JENKINS-42161 and
, we concluded that polling is run for all the jobs targeting a given
repository URL so polling would be run for all branches and the jobs are
created only if the polling returns that there are some changes.
Based on this, we think that in our setup, polling probably doesn't work as
expected but we don't really know how it works in a master/slaves setup.
We have a couple of questions:
- does the master use the workspace dirs of the slaves for polling?
- is the polling run on the slaves instead of the master (i.e. do we need
to start a slave to run the polling and if so how does it choose the right
- is there a way we could get our setup working? Maybe by copying the
workspace dirs to the master and have the polling run on these copied
I suspect we are not alone with this kind of setup so any idea/workaround
You received this message because you are subscribed to the Google Groups
"Jenkins Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
To view this discussion on the web visit
For more options, visit https://groups.google.com/d/optout.