Open Stack

Kai-

Looking at the code the 60 second interval is hard coded and not changeable via a config parameter.

Having said that, and given that it’s a fairly expensive operation to update the DB for `every` compute node in the system, maybe the better question is why do you want to increase the update frequency?  Do you really expect to be creating/deleting instances so fast on a single node that a 60 second latency in the update of the stats is really a problem?

--
Don Dugger
"Censeo Toto nos in Kansa esse decisse." - D. Gale
Ph: 303/443-3786

From: Kai Brennenstuhl [mailto:kai.brennenstuhl at gmail.com]
Sent: Thursday, June 26, 2014 6:56 AM
To: openstack at lists.openstack.org
Subject: [Openstack] [openstack][nova] Compute node update interval

Hi,

how can I specify the interval in which the (free) resources of my compute nodes are updated and written into the DB?
I looked into nova.conf doc and already changed the following attributes:

heal_instance_info_cache_interval = 20

service_down_time = 20

periodic_fuzzy_delay = 20

qpid_heartbeat = 20

scheduler_driver_task_period = 20



call_timeout = 20

db_check_interval = 20

But the update interval in which the resources are written into the database still is about one minute.

Which parameters should I change? Or is this configuration done somewhere else than in nova.conf?

Kind regards
Kai
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20140630/a80bae0d/attachment.html>