Thanks for this, a few things inline. On 6/15/2018 10:04 AM, Chris Dent wrote:
The flip side of this is that it highlights that we have a growing documentation debt with the many features provided by placement and how to make best use of them in nova (and other services that might like to use placement). Before the end of the cycle we will need to be sure that we set aside a considerable chunk of time to address this gap.
I'm glad you mentioned this so I didn't have to. :) The two things off the top of my head that are going to be important for docs are:
1. How to deploy / model shared disk. Seems fairly straight-forward, and we could even maybe create a multi-node ceph job that does this - wouldn't that be awesome?!?!
2. Adding the placement database stuff to the install guide and the placement upgrade notes. I assume that for fresh installs, we want people using the placement database configuration so they don't have to move later during an upgrade, so that seems like a no-brainer. The upgrade docs for placement are per-release and should have a mention of the placement database and any notes / pointers (the spec?) on migration options. What we heard at the summit, for the most part, was 10 minutes of down time in the API for a DB copy and re-config was OK for most, and others that need zero downtime have ways of doing that as well. We don't need a definitive guide, just high level wording on options.
There is now a [heal allocations CLI](https://review.openstack.org/#/c/565886/) which is designed to help people migrate away from the CachingScheduler (which doesn't use placement).
Not only that, but we can use it to heal missing/sentinel consumers from the "consumer generations" blueprint. I've started something for that here:
https://review.openstack.org/#/c/574488/
Nova host aggregates are now magically mirrored as placement aggregates and, amongst other things, this is used to honor the [availability_zone hint via placement](https://review.openstack.org/#/c/546282/).
The aggregates mirror blueprint is not done until the 'nova-manage placement sync_aggregates' command is in place, which is started here:
https://review.openstack.org/#/c/575912/
# Extraction
All of the functional changes for the placement DB stuff are merged or approved, I think there are just some cleanup follow-ups and docs needed still, but otherwise that's done I think and could probably be removed from the runway slot it's in?
* <https://review.openstack.org/#/c/538498/> Convert driver supported capabilities to compute node provider traits
I (finally) got around to rebasing this yesterday. There are a few TODOs left but it shouldn't be too bad, hopefully I can close that out when I get back next week.
-- Thanks, Matt __________________________________________________________________________ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
