[Openstack-operators] Fwd: [openstack-dev] [nova][placement] Upgrade placement first!
Matt Riedemann
mriedemos at gmail.com
Mon Mar 26 20:40:11 UTC 2018
FYI
-------- Forwarded Message --------
Subject: [openstack-dev] [nova][placement] Upgrade placement first!
Date: Mon, 26 Mar 2018 15:02:23 -0500
From: Eric Fried <openstack at fried.cc>
Reply-To: OpenStack Development Mailing List (not for usage questions)
<openstack-dev at lists.openstack.org>
Organization: IBM
To: OpenStack Development Mailing List (not for usage questions)
<openstack-dev at lists.openstack.org>
Since forever [0], nova has gently recommended [1] that the placement
service be upgraded first.
However, we've not made any serious effort to test scenarios where this
isn't done. For example, we don't have grenade tests running placement
at earlier levels.
After a(nother) discussion [2] which touched on the impacts - real and
imagined - of running new nova against old placement, we finally decided
to turn the recommendation into a hard requirement [3].
This gives admins a crystal clear guideline, this lets us simplify our
support statement, and also means we don't have to do 406 fallback code
anymore. So we can do stuff like [4], and also avoid having to write
(and subsequently remove) code like that in the future.
Please direct any questions to #openstack-nova
Your Faithful Scribe,
efried
[0] Like, since upgrading placement was a thing.
[1]
https://docs.openstack.org/nova/latest/user/upgrade.html#rolling-upgrade-process
(#2, first bullet)
[2]
http://eavesdrop.openstack.org/irclogs/%23openstack-nova/%23openstack-nova.2018-03-26.log.html#t2018-03-26T17:35:11
[3] https://review.openstack.org/556631
[4] https://review.openstack.org/556633
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
More information about the OpenStack-operators
mailing list