what is the standard Ansible rolling-deploy with continuous integration testing workflow?

[Gallagher_Polyn]

Hey all,

Happily (and with help from folks on this forum) I am able to do rolling deploys a la the lamp_haproxy example on all of the providers covered by Ansible’s cloud modules: rackspace, aws, do, linode and gce – yay!

Next up: how would github.com/ansible/ansible-examples/tree/master/lamp_haproxy be modified or used in context with a CI component?

I can’t now find the few Ansible configuration references I may have seen shown in online presentations by Michael or Tim, but I think I recall they each showed a CI step directly in the context of a rolling deploy playbook? Is that right? If not, I might expect the flow to be something like: git pull, CI server tests pass, CI server kicks off rolling deploy.

Any references to latest-greatest web app CI-Ansible strategies greatly appreciated. Here are the two that most interest me:

• developer.rackspace.com/blog/move-fast-and-dont-break-things-testing-with-jenkins-ansible-and-docker.html

• www.activestate.com/blog/2014/01/using-docker-run-ruby-rspec-ci-jenkins

Thanks,

G

[Michael_DeHaan1 (Michael DeHaan)]

Usually you would have something like:

Jenkins build
if successful, call rolling update in stage
if successful, call rolling update in prod