I see many examples in nova of where we still read rows with read_deleted="yes". I think we need to see a plan for how to remove all of those before we can progress this.
Michael On Tue, Mar 11, 2014 at 9:06 AM, Johannes Erdfelt <johan...@erdfelt.com> wrote: > On Mon, Mar 10, 2014, Joshua Harlow <harlo...@yahoo-inc.com> wrote: >> Sounds like a good idea to me. > > I generally think this is a good idea too. > >> I've never understood why we treat the DB as a LOG (keeping deleted == 0 >> records around) when we should just use a LOG (or similar system) to >> begin with instead. >> >> Does anyone use the feature of switching deleted == 1 back to >> deleted = 0? Has this worked out for u? > > This isn't the only potential use. It's possible that code depends on > being able to still access deleted records. > > For instance, in the past we could delete an instance_type, but if an > instance is still referencing it, code would still try to fetch it from > the database some times. > > This particular example probably isn't an issue anymore since I think all > of that has been moved to instance metadata specifically to avoid > problems like this. > > That said, I think it's well worth the effort to simplify the code and > make operators lives easier. > > JE > > > _______________________________________________ > OpenStack-dev mailing list > OpenStack-dev@lists.openstack.org > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev -- Rackspace Australia _______________________________________________ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
