Thanks Edward, your python rewrite generally looks good at a glance.
I confirmed the test failures reported by automated testing: jenkins-slave is not being found because it is in the precise series only.
Changing test/100-deploy-trusty:line 19 to "d.add('jenkins-slave', 'cs:precise/jenkins-slave')" remedies this issue until there is a trusty version of jenkins slave.
The tests also hit an error in "master-relation-changed" due to what appears to be a race condition with the configuration and restart of the jenkins slave and adding a node to master. Running the hook via debug hook runs fine.
Thanks Edward, your python rewrite generally looks good at a glance.
I confirmed the test failures reported by automated testing: jenkins-slave is not being found because it is in the precise series only.
Changing test/100- deploy- trusty: line 19 to "d.add( 'jenkins- slave', 'cs:precise/ jenkins- slave') " remedies this issue until there is a trusty version of jenkins slave.
The tests also hit an error in "master- relation- changed" due to what appears to be a race condition with the configuration and restart of the jenkins slave and adding a node to master. Running the hook via debug hook runs fine.
See logs: https:/ /gist.githubuse rcontent. com/anonymous/ 0067138ce2cc697 b8c88/raw/ 3f4c03688400b32 f3aa66e1bd3bad5 b7398f80a5/ jenkins- race-condition
I confirmed this as an issue in the merge targets bash implementation also. Adding some retry logic to add_node should fix this issues.
-1 for test fixes, but otherwise looks good. Thanks again!