On 10/10/16 11:11, Spyros Trigazis wrote:
4. Finally, a thought under investigation is replacing the nodes one
by one using a different image. e.g. Upgrade from fedora 24 to 25
with new versions of packages all in a new qcow2 image. How could
we update the stack for this?
This should work now - you just set an update_policy for rolling
updates. That allows you to configure a batch size for changes (e.g.
batch_size: 1 for one-by-one replacement; you can also configure a
minimum number of nodes to remain in service). Then do a stack update
with the new image name and Heat will handle the rest.
The trickier part is disabling whatever is running on the server before
deleting it. You can do this with a SoftwareDeployment with action:
DELETE inside the scaled unit (if you're scaling out nested stacks).
Alternatively, you can configure pre-delete hooks on the servers,
configure a Zaqar queue to send hook notifications to, and use Zaqar
subscriptions to have the notifications to that queue trigger a Mistral
OpenStack Development Mailing List (not for usage questions)