A year and a half ago I did some analysis on how nova uses placement.
I've repeated some of that analysis today and here's a brief summary of the results. Note that I don't present this because I'm concerned about load on placement, we've demonstrated that placement scales pretty well …
Back in February I wrote up a review of the issues involved with extracting placement from nova. Many of the tasks associated with that have been addressed but there's definitely plenty left to do so I'm writing up a second version.
Most code which is within nova.api.openstack.placement …
This is an appendix to my Placement Container Playground series. The last one was Playground 6. That was going to be the end of the series because "woot, it works" and that was enough.
Since then the container has proven useful in a variety of ways, mostly as the host …