15:00:26 <ralonsoh> #startmeeting neutron_qos
15:00:30 <openstack> Meeting started Tue Jan 14 15:00:26 2020 UTC and is due to finish in 60 minutes.  The chair is ralonsoh. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:31 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:33 <openstack> The meeting name has been set to 'neutron_qos'
15:00:33 <ralonsoh> hello!
15:00:37 <slaweq> hi :)
15:00:40 <slaweq> ;)
15:00:40 <maciejjozefczyk> hello!
15:00:42 <lajoskatona> Hi
15:00:56 <ralonsoh> davidsha is here
15:01:12 <davidsha> Hey!
15:01:21 <ralonsoh> I think we have qorum
15:01:28 <ralonsoh> let's start
15:01:28 <ralonsoh> #topic RFEs
15:01:38 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1476527
15:01:38 <openstack> Launchpad bug 1476527 in neutron "[RFE] Add common classifier resource" [Wishlist,Triaged] - Assigned to Igor D.C. (igordcard)
15:01:46 <ralonsoh> 1) the spec
15:02:05 <ralonsoh> #link https://review.opendev.org/#/c/678865/
15:02:09 <ralonsoh> IMO, we can merge it
15:02:23 <ralonsoh> if slaweq 's comments are solved
15:02:39 <ralonsoh> are we going to use RBAC and shared?
15:02:57 <davidsha> I responded, I think QoS has both on qos polices?
15:03:29 <ralonsoh> yes, and yes, QoS has both
15:03:46 <davidsha> I have the RBAC already in the service plugin, so should I remove shared?
15:04:20 <ralonsoh> slaweq, ^
15:05:00 <slaweq> sorry, I was in the kitchen for a while
15:05:16 <ralonsoh> if ClassificationGroup is going to have a RBAC policy
15:05:23 <ralonsoh> and this policy is going to be shared or not
15:05:24 <slaweq> can be rbac for me
15:05:33 <ralonsoh> then "shared" is not necessary
15:05:49 <ralonsoh> so we agree on this, only RBAC
15:06:01 <davidsha> kk, I'll remove it
15:06:09 <slaweq> ok, thx
15:06:17 <ralonsoh> cool, we should merge the spec this week
15:06:30 <slaweq> but does QoS policy has "shared" flag too?
15:06:32 <slaweq> or not?
15:06:44 <slaweq> if it has, maybe we should add it here too?
15:06:46 <ralonsoh> yes
15:06:53 <ralonsoh> qos has shared
15:07:16 <slaweq> so You can share qos policy via 2 mechanisms RBAC and shared flag, right?
15:07:34 <ralonsoh> yes and this is something we should avoid
15:07:51 <ralonsoh> because eventually we can have a clash
15:08:01 <ralonsoh> between both params
15:08:04 <slaweq> we have it like that for networks also
15:09:51 <slaweq> in fact if we want to be consistent with other resources we should add both possibilities to share
15:09:58 <slaweq> so shared flag would be needed than
15:10:04 <davidsha> ok
15:10:13 <slaweq> but if we want to do it only with RBAC, that probably this flag is not needed
15:10:21 <slaweq> as You want :)
15:11:23 <davidsha> Cool, I'll leave both in!
15:11:30 <slaweq> ok
15:11:39 <ralonsoh> perfect
15:11:54 <slaweq> so the spec is ready to go now
15:12:08 <ralonsoh> cool
15:12:16 <davidsha> \o/
15:12:21 <ralonsoh> 2) the n-lib patch
15:12:29 <ralonsoh> #link https://review.opendev.org/#/c/670049/
15:12:41 <slaweq> +W'ed spec patch
15:13:02 <ralonsoh> you need to address lajoskatona 's comments
15:13:07 <davidsha> slaweq: thanks!
15:13:09 <ralonsoh> about the API definition
15:13:37 <ralonsoh> but apart from this, I'm ok with the n-lib patch
15:13:51 <davidsha> Yes, I'll have those up by the end of the day.
15:14:00 <davidsha> kk
15:14:05 <lajoskatona> davidisha and all: good to see this progressing :-)
15:14:05 <ralonsoh> cool
15:14:48 <davidsha> lajoskatona: thanks for the reviews, helps the progress!
15:14:54 <davidsha> :D
15:15:13 <ralonsoh> 3) and the last patch to be discussed, the neutron patch with the DB definiton and the extension
15:15:17 <ralonsoh> #link https://review.opendev.org/#/c/670050/
15:15:51 <ralonsoh> apart from waiting for the n-lib patch and then deleting the duplicated exceptions, the patch is OK for me
15:16:19 <ralonsoh> well, of course
15:16:22 <ralonsoh> we need testing
15:16:34 <davidsha> I'll port over the unit and functional tests we have in neutron classifier
15:16:49 <slaweq> one nitty nit for this last patch, please remove "POC" from the title in next revision if it's going to be merged :)
15:17:02 <slaweq> I think we shouldn't merge something called POC
15:17:02 <davidsha> Will do!
15:17:46 <ralonsoh> cool, the classification/plugin.py need some extensive testing (and reviews), it's a hard nut to crack
15:18:22 <davidsha> Anything I can do to make the reviews easier?
15:19:03 <ralonsoh> I don't know, I just spent 1 hour reading and understanding the code...
15:19:13 <lajoskatona> Perhaps an api-ref, I suppose there's something for the original project?
15:19:14 <davidsha> Lol, comments it is ;P
15:19:15 <ralonsoh> is just a matter of time
15:19:48 <davidsha> there should be, I'll try to find it.
15:20:05 <lajoskatona> davidisha: thanks
15:21:00 <ralonsoh> ok, something else to be commented in this spec?
15:21:06 <ralonsoh> *RFE
15:21:50 <ralonsoh> ok, let's move on
15:22:02 <ralonsoh> next one, #link https://bugs.launchpad.net/neutron/+bug/1851362
15:22:02 <openstack> Launchpad bug 1851362 in neutron "[RFE] ports do not inherit their associated network's policy" [Low,In progress] - Assigned to Rodolfo Alonso (rodolfo-alonso-hernandez)
15:22:17 <ralonsoh> I made a huge mistake in n-lib 1.30.1
15:22:27 <ralonsoh> redefining qos_policy_id in port
15:22:46 <ralonsoh> so now we have n-lib 2.0.0 (without py2 support, btw)
15:23:03 <ralonsoh> once we have the neutron patch #link https://review.opendev.org/#/c/701517/
15:23:18 <ralonsoh> bumping the n-lib version, I'll continue with the neutron patch
15:23:32 <ralonsoh> #link https://review.opendev.org/#/c/693244/
15:23:38 <ralonsoh> and the OSsdk one
15:23:47 <ralonsoh> #link https://review.opendev.org/#/c/693247/
15:24:04 <ralonsoh> that's all, any other comment?
15:24:27 <davidsha> None from me.
15:24:39 <ralonsoh> thanks!
15:24:46 <ralonsoh> ok, next one
15:24:50 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1843165
15:24:50 <openstack> Launchpad bug 1843165 in neutron "RFE: Adding support for direct ports with qos in ovs" [Wishlist,In progress] - Assigned to waleed mousa (waleedm)
15:25:17 <ralonsoh> A "trivial" patch due to the full support of QoS for offloaded ports, in newer OVS versions
15:25:27 <ralonsoh> of course, the patch should make those kind of checks
15:25:34 <ralonsoh> #link https://review.opendev.org/#/c/611605/
15:25:56 <ralonsoh> please, review the patch
15:26:23 <ralonsoh> the main goal of this patch, apart form adding portbindings.VNIC_DIRECT to the ovs driver
15:26:31 <ralonsoh> is to check is this is supported
15:26:41 <ralonsoh> any comment?
15:27:01 <davidsha> no, I'll review later
15:27:18 <ralonsoh> and that's all form my list of RFEs
15:27:24 <ralonsoh> any other one?
15:27:44 <slaweq> IMO this last patch is almost ready to go
15:27:51 <slaweq> I had only some comments to the docs
15:27:54 <ralonsoh> I think so
15:28:05 <ralonsoh> I need to push my comments
15:28:50 <ralonsoh> ok, let's move to next section
15:28:51 <ralonsoh> #topic Bugs
15:29:00 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1845176
15:29:00 <openstack> Launchpad bug 1845176 in neutron "Removing of QoS queue in neutron-ovs-agent fails due to existing references" [Medium,In progress] - Assigned to Rodolfo Alonso (rodolfo-alonso-hernandez)
15:29:09 <ralonsoh> and the patch
15:29:15 <ralonsoh> #link https://review.opendev.org/#/c/687922/
15:29:45 <ralonsoh> what can I say, IMO is ready to be merged!
15:29:57 <ralonsoh> please, add it to your review piles
15:30:19 <ralonsoh> ok, next one
15:30:24 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1853840
15:30:24 <openstack> Launchpad bug 1853840 in neutron "Neutron fails to create bandwidth providers if CONF.host is set" [High,Fix released] - Assigned to Bence Romsics (bence-romsics)
15:30:32 <ralonsoh> closed
15:30:37 <ralonsoh> #link https://review.opendev.org/#/c/696600/
15:30:54 <ralonsoh> thanks to bence
15:31:08 <ralonsoh> I'll remove it from the list
15:31:26 <ralonsoh> next one
15:31:34 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1853171
15:31:34 <openstack> Launchpad bug 1853171 in neutron "Deprecate and remove any "ofctl" code in Neutron and related projects " [Medium,In progress] - Assigned to David Shaughnessy (david-shaughnessy)
15:31:39 <ralonsoh> this one will take more time
15:32:04 <ralonsoh> because the ocftl command is used in many parts of the code
15:32:05 <davidsha> The neutron side is done correct?
15:32:13 <ralonsoh> #link https://review.opendev.org/#/c/695479/
15:32:16 <ralonsoh> this patch?
15:32:33 <davidsha> yup, or is there more to be done?
15:32:38 <ralonsoh> the firewall
15:32:41 <davidsha> Ah
15:32:42 <ralonsoh> it's using it
15:33:00 <ralonsoh> so, because this is not a real bug
15:33:11 <ralonsoh> the priority is not high
15:33:37 <ralonsoh> once you have the classifier , you can spend time on this
15:33:41 <ralonsoh> if you have it
15:33:51 <davidsha> kk, I'll try
15:33:55 <ralonsoh> thanks!
15:34:06 <davidsha> np!
15:34:36 <ralonsoh> any other bug that I missed?
15:35:21 <ralonsoh> ok, let's move to the next section
15:35:23 <ralonsoh> #topic Open Discussion
15:35:56 <ralonsoh> I have nothing in the agenda
15:36:07 <ralonsoh> everything is good in the QoS world
15:36:14 <davidsha> :D
15:36:18 <slaweq> \o/
15:36:40 <ralonsoh> perfect, thanks folks! I'll end the meeting here
15:36:44 <ralonsoh> see you online!
15:36:48 <ralonsoh> #endmeeting