On Thu, Jun 27, 2019 at 12:19 AM Ben Pfaff <[email protected]> wrote: > On Wed, Jun 26, 2019 at 07:46:25PM +0530, [email protected] wrote: > > From: Numan Siddique <[email protected]> > > > > Running the command "ovn-nbctl set logical_switch_port foo > external_ids:foo=bar" > > results in the incremetal processing engine to recompute the flows on the > > chassis where the logical port 'foo' is claimed. > > > > This patch avoids this unnecessary recomputation by omitting the alert > for the > > Port_Binding.external_ids column. > > > > Signed-off-by: Numan Siddique <[email protected]> > > Acked-by: Han Zhou <[email protected]> > > --- > > > > v2 -> v3 > > ------ > > * Based on the review comments from Han, dropped p2 and p3 from the > series. > > This only omits the alert for external-ids from one table. I think that > ovn-controller should not look at the external_ids in most tables. The > only ones I see it dealing with are: the one on Chassis, which seems > legitimate (it is only setting keys there, and the behavior is > documented) and external_ids:dns_id in the DNS table, which makes me > suspicious because it appears that it's an undocumented communication > channel from ovn-northd to ovn-controller. I think that all the others > can omit alerts. >
To sync the dns entries from nb db to sb db, we store the uuid of nb db dns row in sb db. I will submit a doc patch for this. Thanks for pointing this out. > However, why just omit alerts? If ovn-controller does not read these > (it should not), then they may be omitted entirely. > Right? > That's right. I don't understand the alert APIs properly. I will check it and update the patch to omit tracking the external_ids of all tables. Thanks Numan _______________________________________________ dev mailing list [email protected] https://mail.openvswitch.org/mailman/listinfo/ovs-dev
