Hi Everyone, So we were having a discussion in #openstack-qa about the amount of logs we have to process in our elastic search cluster and that we frequently have issues with the amount of RAM available in the cluster.
In general, it would be good for projects to take a look at how verbose their logging is, especially on successful tempest runs, and see if that's really necessary and see if they can make things a bit more concise and targeted (this will help operators too). However, one thing that stood out to me is that we're getting a lot of deprecation warnings from all our gate jobs. Using this query to see how many deprecation warning we emit on jobs running on master in the past 7 days: http://logstash.openstack.org/#dashboard/file/logstash.json?query=message:%5C%22deprecated%5C%22%20AND%20loglevel:%5C%22WARNING%5C%22%20AND%20build_branch:%5C%22master%5C%22 it found 17576707 hits. I don't really see any reason for us to be running dsvm jobs on master using deprecated options. We're looking for some people to volunteer to help with this, it isn't something that's very difficult to fix. It'll likely just require changing things in devstack to avoid configuring services with deprecated options. But, a big part is also going to be going through the logs to find out where we're using deprecated options. Thanks, Matthew Treinish
signature.asc
Description: PGP signature
__________________________________________________________________________ OpenStack Development Mailing List (not for usage questions) Unsubscribe: [email protected]?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
