When attempting to add a jenkins cartridge via the REST API when there is no jenkins server installed, the node fails and returns this chunk of error. DEBUG: [#<MCollective::RPC::Result:0x7f7e0db571c8 @results={:statusmsg=>"OK", :sender=>"ip-10-85-2-221", :data=>{:exitcode=>0, :output=>"/usr/libexec/stickshift/cartridges/embedded/jenkins-client-1.4/info/hooks/deconfigure: line 38: /var/lib/stickshift/46b611049ead4531832be7111d048151//.env/JENKINS_URL: No such file or directory\n/usr/libexec/stickshift/cartridges/embedded/jenkins-client-1.4/info/hooks/deconfigure: line 39: /var/lib/stickshift/46b611049ead4531832be7111d048151//.env/JENKINS_USERNAME: No such file or directory\n/usr/libexec/stickshift/cartridges/embedded/jenkins-client-1.4/info/hooks/deconfigure: line 40: /var/lib/stickshift/46b611049ead4531832be7111d048151//.env/JENKINS_PASSWORD: No such file or directory\nCLIENT_RESULT: Jenkins client 1.4 has been removed from: test\n"}, :statuscode=>0}, @action="cartridge_do", @agent="libra">] The site needs a unique exitcode to identify this failure and show the correct message to the user. The broker is currently configured to pass exitcode back to the client, but it looks like the node is returning insufficient info. Need this to make the UI report correctly in sprint 12.
Hold on - might be my problem... testing.... testing....
This is what jenkins client does on configure: if ! [ -f "$APP_HOME/.env/JENKINS_URL" ] then client_error "Jenkins server does not exist! Please create with rhc app create -t jenkins-1.4" exit 151 fi if [ -f "$APP_HOME/.env/OPENSHIFT_CI_TYPE" ] then client_error "Jenkins client already embedded in $application" exit 152 fi The output above appears to be from the deconfigure which is returning 0 because there is no client left to remove.
My bad, fixed.