Two weeks ago, the patch https://review.openstack.org/#/c/164329/ was merged to stable/juno. This patch fixes bug https://launchpad.net/bugs/1414218 on the stable/juno branch. This patch removed LOG.debug statements from the for-loop of method _output_hosts_file() in file neutron/agent/Linux/dhcp.py. Without the LOG.debug statements in the for-loop, the performance of dhcp agent updates improves significantly.
As it stands currently, neutron on the stable/juno branch behaves as if https://review.openstack.org/#/c/164329/ was never merged. The merge of patch https://review.openstack.org/#/c/153181 puts some LOG.debug statements in the for-loop of the same function. The net result is the unintentional revert of patch https://review.openstack.org/#/c/164329/. According to review.openstack.org, the patch https://review.openstack.org/#/c/153181 merged on April 1st at 1:56 pm. Patch https://review.openstack.org/#/c/164329/ merged on the same day at 2:21 pm. After 153181 merged, the review of 164329 should have triggered a merge conflict error which would have compelled the owner to do a rebase. The merge conflict wasn't reported and the patches merged anyway. Does this point to a problem with the Jenkins CI system? Note that this is only a problem on the neutron stable/juno branch. I filed bug https://bugs.launchpad.net/neutron/+bug/1445202 to have the LOG.debug statements taken out. The review is https://review.hpcloud.net/#/c/174596.
__________________________________________________________________________ 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
