Created attachment 944486[details]
Jenkins log
Description of problem:
The jenkins connection with the builder that gets lost after 15 minutes and is not reconnected when launching a new build (it tries to re provision which is not needed, and not supported by the broker).
Version-Release number of selected component (if applicable):
2.1
How reproducible:
Always
Steps to Reproduce:
1 create a new domain
2 create a new application
3 enable jenkins
4 modifiy the time to live to 60 minutes
5 modify the build : add a "sleep 20m" in the script which will block the build for 20 minutes
6 launch
Actual results:
It's jenkins that looses the connection. However, the Openshift plugin doesn't seem to cope very well with this situation : jenkins requests a new slave to be added, the broker detects that it still exists and build is cancelled.
Expected results:
When you detect that the slave already exists, the conclusion should be to trigger a reconnect (you know the slave exists and that jenkins is not connected)... this would at least fix the situation where the disconnect happens after the work is done.
Additional info:
Verified this bug with 2.2/2015-03-18.2 puddle, PASS.
In initial reproduce steps, actually no need step 4, because when jenkins build is running, whatever time-to-live is 15m or 60m, the jenkins builder would never be terminated, time-to-live is for "idled slaves". In old builds, in my case, when the time for jenkins building is longer that 5 mins, the jenkins slave would be offline.
Steps:
1. Create an app with jenkins-client embedded.
2. Add a "build" action hooks script in app's git repo: add a "sleep 20m" in the script which will block the build for 20 minutes
Result:
After 20 mins, the jenkins build is completed successfully.
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.
For information on the advisory, and where to find the updated
files, follow the link below.
If the solution does not work for you, open a new bug report.
https://rhn.redhat.com/errata/RHBA-2015-0779.html
Created attachment 944486 [details] Jenkins log Description of problem: The jenkins connection with the builder that gets lost after 15 minutes and is not reconnected when launching a new build (it tries to re provision which is not needed, and not supported by the broker). Version-Release number of selected component (if applicable): 2.1 How reproducible: Always Steps to Reproduce: 1 create a new domain 2 create a new application 3 enable jenkins 4 modifiy the time to live to 60 minutes 5 modify the build : add a "sleep 20m" in the script which will block the build for 20 minutes 6 launch Actual results: It's jenkins that looses the connection. However, the Openshift plugin doesn't seem to cope very well with this situation : jenkins requests a new slave to be added, the broker detects that it still exists and build is cancelled. Expected results: When you detect that the slave already exists, the conclusion should be to trigger a reconnect (you know the slave exists and that jenkins is not connected)... this would at least fix the situation where the disconnect happens after the work is done. Additional info: