We've noticed the load on the Nova and Keystone APIs. It may well be that there is an increased load on the hypervisor but this is distributed and thus we might not have seen it. The API impact was significant.... a factor of 14x increase in API call rate (http://openstack-in-production.blogspot.de/2014/03/cern-cloud-architecture-update-for.html)
Tim > -----Original Message----- > From: gordon chung [mailto:g...@live.ca] > Sent: 27 May 2015 20:43 > To: OpenStack Development Mailing List not for usage questions > Subject: Re: [openstack-dev] [ceilometer][all] Scalable metering > > > > > > So I agree doing lots of naive polling can lead to issues on even the > > fastest of APIs, but are there any bugs about this that were opened > > against nova? At the very least nova should investigate why the > > specific calls are so slow and see if we what we can do to make them > > at least a little faster and lighter weight. > > > > > > good question. Tim, i was also wondering, is the big issue the load on the > nova > api or the hypervisor? i'm just trying out a solution that might help > regarding the > former. > > cheers, > _________________________________________________________________ > _________ > 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 __________________________________________________________________________ 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
