What's best practices for starting, bouncing, stopping a deployment?

[Min Zhou]

Hi all,

The operations workload of a distributed system often include initializing, upgrading, bouncing, stopping, restarting the system. It seems that the official  examples of ansible (https://github.com/ansible/ansible-example) just showed how to initialize a deployment, no further operations.
I think ansible can do the other things as well. What's the best practices to achieve those tasks?  By different directory or different roles? Or just different ansible project?
Is there any existing example for this?


Thanks,
Min

[Henry Finucane]

Separate playbooks for special kinds of rollouts seem to be the way to go. Everyone should have a "rolling openssl upgrade" playbook, for instance ;).

--
You received this message because you are subscribed to the Google Groups "Ansible Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ansible-proje...@googlegroups.com.
To post to this group, send email to ansible...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ansible-project/249c93ee-b6a0-4dba-8463-57df3e2f5d13%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.