Indeed the VMware plugins were not using resource tracking (they know that my code should not be trusted!)
I think this bears however another question that we need to answer... it is likely that some change broke quota enforcement for plugins which do not use usage tracking. When I developed reservations & usage tracking we made an assumption that plugins should not be forced to use usage tracking. If they did not, the code will fallback to the old logic which just executed a count query. If we want to make usage tracking mandatory I'm fine with that, but we first need to make sure that every plugin enables it for every resource it handles. Salvatore On 17 March 2016 at 12:41, Gary Kotton <gkot...@vmware.com> wrote: > Thanks! > > Much appreciated. Will check > > From: Kevin Benton <ke...@benton.pub> > Reply-To: OpenStack List <openstack-dev@lists.openstack.org> > Date: Thursday, March 17, 2016 at 1:09 PM > To: OpenStack List <openstack-dev@lists.openstack.org> > Subject: Re: [openstack-dev] [Neutron] RBAC: Fix port query and deletion > for network owner > > After reviewing your logs[1], it seems that quotas are not working > correctly in your plugin. There are no statements about tenants being > marked dirty, etc. > > I think you are missing the quota registry setup code in your plugin init. > Here is the ML2 example: > https://github.com/openstack/neutron/blob/44ef44c0ff97d5b166d48d2ef93feafa9a0f7ea6/neutron/plugins/ml2/plugin.py#L167-L173 > <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_openstack_neutron_blob_44ef44c0ff97d5b166d48d2ef93feafa9a0f7ea6_neutron_plugins_ml2_plugin.py-23L167-2DL173&d=BQMFaQ&c=Sqcl0Ez6M0X8aeM67LKIiDJAXVeAw-YihVMNtXt-uEs&r=VlZxHpZBmzzkWT5jqz9JYBk8YTeq9N3-diTlNj4GyNc&m=0W0JwODAYJdy_tB4tQT0nd1zhv2OJ35zOHr5UM4KUSs&s=lXC43_6eXjZnhUzC2nnFkaD--k9FnCn8_IQK2_tFtcs&e=> > > > > http://208.91.1.172/logs/neutron/293483/1/check-tempest-vmware-nsx-v3/q-svc.log.txt.gz > <https://urldefense.proofpoint.com/v2/url?u=http-3A__208.91.1.172_logs_neutron_293483_1_check-2Dtempest-2Dvmware-2Dnsx-2Dv3_q-2Dsvc.log.txt.gz&d=BQMFaQ&c=Sqcl0Ez6M0X8aeM67LKIiDJAXVeAw-YihVMNtXt-uEs&r=VlZxHpZBmzzkWT5jqz9JYBk8YTeq9N3-diTlNj4GyNc&m=0W0JwODAYJdy_tB4tQT0nd1zhv2OJ35zOHr5UM4KUSs&s=yxUgpc6J-xovf-ZwLu-wFeU6dGj4Ne6P-vTR32UCjrI&e=> > > On Thu, Mar 17, 2016 at 1:30 AM, Gary Kotton <gkot...@vmware.com> wrote: > >> Hi, >> The review https://review.openstack.org/#/c/255285/ breaks our CI. Since >> this has landed we are getting failed tests with the: >> "Details: {u'message': u"Quota exceeded for resources: ['port'].", >> u'type': u'OverQuota', u'detail': uāā}" >> When I revert the patch and run our CI without it the tests pass. Is >> anyone else hitting the same or a similar issue? >> I think that for Mitaka we need to revert this patch >> Thanks >> Gary >> >> __________________________________________________________________________ >> 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 > >
__________________________________________________________________________ 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