OpenStack Upgrades - HA with Pause/resume begging to be scripted

[Ryan Farrell]

Steps to perform ha cluster upgrade with pause + resume are very user unfriendly and likely to result in user error because of the care needed to parse leader / non-leader unit numbers for both the master application and the hacluser subordinate.

- This is begging for a script that can read the juju status and provide exact steps tailored to a specific cloud.

- [X] This is a doc addition request.

If you have a troubleshooting or support issue, use the following resources:

 - Ask OpenStack: http://ask.openstack.org
 - The mailing list: http://lists.openstack.org
 - IRC: 'openstack' channel on Freenode

-----------------------------------
Release: 0.1 on 2019-04-08 13:55
SHA: 65630c0bf5114d2b1682917d5502bfa4f830f0d8
Source: Can't derive source file URL
URL: https://docs.openstack.org/project-deploy-guide/charm-deployment-guide/stein/app-upgrade-openstack.html

[Peter Matulis (petermatulis)]

Thank you very much for this report. I agree that it can be "tricky". However, scripting cloud upgrades in official documentation would then make it "risky". We will consider adding a note about the idea of a script and a possible example snippet. If you have any code on your end that would help expedite this it would be very welcome.

[Peter Matulis (petermatulis)]

The guide now makes reference to automation scripts [1] as something optional that users may consider for non-production clouds. I am marking this bug as (incidentally) Fix Released.

[1]: https://docs.openstack.org/project-deploy-guide/charm-deployment-guide/latest/upgrade-series-openstack.html#automation