Past ops summits included hypervisor tuning sessions that were generally well attended and the beginnings of some upstream docs were subsequently distilled (https://wiki.openstack.org/wiki/Documentation/HypervisorTuningGuide) from the various resulting etherpads. However subsequent contributions to the wiki-centralised docs have been almost non-existent despite many deployers being concerned with tuning their clouds and numerous summit presentations show-casing the results of various tuning choices.
In this session we would have a recap of prior recommendations and welcome new input. However the primary goal would be to map out a way to turn this knowledge into upstream documentation (e.g. a section in the ops and perhaps also architecture guides) with some discussion around repeatable regression tests for and result tracking of tuning recommendations.
Moderators: Blaire Bethwaite, Joe Topjian
