Open Stack

Looking for recommendations on tuning of nova-placement-api. I have a 
few moderately-sized deployments (~200 nodes, ~4k instances), currently 
on Ocata, and instance creation is getting very slow as they fill up.

I discovered that calls to placement seem to be taking a long time, and 
even this is horribly slow:

$ time curl http://apihost:8778/
{"error": {"message": "The request you have made requires 
authentication.", "code": 401, "title": "Unauthorized"}}
real	0m20.656s
user	0m0.003s
sys	0m0.001s
$

nova-placement-api is running under mod_wsgi with the "standard"(?) 
config, i.e.:

...
   WSGIProcessGroup nova-placement-api
   WSGIApplicationGroup %{GLOBAL}
   WSGIPassAuthorization On
   WSGIDaemonProcess nova-placement-api processes=3 threads=1 user=nova 
group=nova
   WSGIScriptAlias / /usr/bin/nova-placement-api
...

Seems that the service is overwhelmed. Other WSGI services on the same 
host are responding in a timely manner.

I'm thinking of bumping up "processes" quite a bit (maybe like 16 or 
so). I'm assuming that increasing "threads" wouldn't help(?). Any 
lessons from experience in this area?

Other suggestions? I'm looking at things like turning off 
scheduler_tracks_instance_changes, since affinity scheduling is not 
needed (at least so-far), but not sure that that will help with 
placement load (seems like it might, though?)

TIA,

     ~iain