Hi Li Ma, Great job on this!
I think Rally is the correct way to test something like that, what you could use to check the end to end solution is try to restart the DF DB server for timeouts during API configurations. At this point of time, i think this is really advance testing and i wouldn't bother too much right now, we need to make everything else more stable and then get back to this. On Wed, Feb 24, 2016 at 4:31 PM, Li Ma <skywalker.n...@gmail.com> wrote: > Hi all, today I submitted the working patch for review: > > https://review.openstack.org/#/c/282290/ > > I tested it by invoking many concurrent API calls to update a given > Neutron object, both for standalone and clustering of MySQL. The result is > that no reported bugs about inconsistency were triggered. > > At the current stage, I'd like to discuss about testing. I think it needs > to be tested at the gate by fullstack or rally. However, due to the > uncertainty of the problem, even without this patch, the bug is not always > triggered. How to make sure the test cases are reliable is the problem here. > > Any thoughts? Thanks a lot. > > -- > > Li Ma (Nick) > Email: skywalker.n...@gmail.com > > __________________________________________________________________________ > 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 > > -- Best Regards , The G.
__________________________________________________________________________ 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
