Ben Pfaff <b...@ovn.org> wrote on 05/23/2016 12:34:27 PM:
> From: Ben Pfaff <b...@ovn.org> > To: Ryan Moats/Omaha/IBM@IBMUS > Cc: ovs dev <dev@openvswitch.org> > Date: 05/23/2016 12:34 PM > Subject: Re: [ovs-dev] Trying to decide of the following clean up > code in a test case is a bug or not > > On Fri, May 20, 2016 at 05:12:12PM -0500, Ryan Moats wrote: > > I'm looking at the cleanup portion of the "ovn -- send gratuitous arp > > on localnet" test case at lines 2591-2593 in tests/ovn.at: > > > > # Delete the localnet ports. > > AT_CHECK([ovs-vsctl del-port localvif1]) > > AT_CHECK([ovn-nbctl lport-del ln_port]) > > > > What appears to be happening is that the port is removed from the > > openvswitch, but the corresponding Port_Binding row isn't automatically > > removed. This doesn't cause a problem when rebuilding everything in > > each iteration, but it is causing me some headaches when trying to > > persist. > > One would expect that, following the ovs-vsctl, ovn-controller would set > the chassis column in the appropriate Port_Binding row to the empty set. > Only "ovn-nbctl lport-del" should actually remove the row. I came to the same conclusion over the weekend and that is what the incremental processing patch now does. Thanks for the confirmation! Ryan _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev