15:00:42 <ralonsoh> #startmeeting neutron_qos
15:00:43 <openstack> Meeting started Tue Jan 15 15:00:42 2019 UTC and is due to finish in 60 minutes.  The chair is ralonsoh. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:44 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:47 <openstack> The meeting name has been set to 'neutron_qos'
15:01:01 <ralonsoh> Hello everybody
15:01:03 <mlavalle> o/
15:01:03 <rubasov> o/
15:01:06 <lajoskatona> o/
15:02:02 <ralonsoh> let's start
15:02:08 <ralonsoh> #topic RFEs
15:02:15 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1578989
15:02:16 <openstack> Launchpad bug 1578989 in neutron "[RFE] Strict minimum bandwidth support (egress)" [Wishlist,In progress] - Assigned to Bence Romsics (bence-romsics)
15:02:17 <slaweq> hi
15:02:24 <ralonsoh> Let me write first the links
15:02:31 <ralonsoh> #link https://review.openstack.org/#/q/status:open+project:openstack/neutron+branch:master+topic:minimum-bandwidth-allocation-placement-api
15:02:31 <ralonsoh> #link https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/bandwidth-resource-provider
15:02:31 <ralonsoh> #link https://review.openstack.org/#/q/topic:minimum-bandwidth-allocation-placement-api+(status:open+OR+status:merged)+project:openstack/tempest
15:02:40 <ralonsoh> for nova, neutron and tempest
15:02:48 <ralonsoh> rubasov, lajoskatona please, updates?
15:02:57 <rubasov> yes we have some
15:03:26 <rubasov> if anybody is interested about this: https://etherpad.openstack.org/p/bandwidth-way-forward
15:03:34 <rubasov> feel free to join on Friday
15:03:51 <rubasov> the topics are mostly about the nova side of the feature
15:04:10 <mlavalle> rubasov: that doesn't affect our plans, right?
15:04:17 <rubasov> but depending on directions taken it may have a few consequences on neutron work too
15:04:18 <mlavalle> I mean Neutron plans
15:04:45 <rubasov> mlavalle: not too deeply iiuc
15:04:56 <mlavalle> yes, that's my understanding
15:05:19 <rubasov> there may be a need to control a neutron extension standalone
15:05:27 <rubasov> but not sure right now
15:06:12 <ralonsoh> what's the problem with bward compatibility with SRIOV ports?
15:07:03 <rubasov> ralonsoh: basically that what we had before (guarantee on dataplane but not in placement) allowed an inconsistent resource view
15:07:32 <rubasov> and now if we want to enforce the consistency of the resource view it's harder to not break what we had in neutron only
15:07:45 <lajoskatona> ralonsoh: here I think it is summarized quite well by gibi: https://etherpad.openstack.org/p/bandwidth-way-forward
15:07:46 <rubasov> not impossible but significantly harder
15:07:54 <ralonsoh> but you mean if you update the controller to newer version
15:08:09 <rubasov> ralonsoh: yes
15:08:12 <ralonsoh> ok
15:08:54 <ralonsoh> but I though that kind of updates were going to be send by neutron-server during the start process
15:09:47 <ralonsoh> the ml2 mech driver should be in charge of, during the boot process, to read the current status and update the current BW available
15:10:08 <ralonsoh> in this case, to read the already assigned BW and update the placement DB
15:10:22 <rubasov> one problem for example: what do we do if we already overallocated the guaranteed min bw on host when we upgrade to a control that supports guarantee in placement too?
15:10:42 <ralonsoh> rubasov, good question
15:10:47 <rubasov> s/control/control plane/
15:11:05 <ralonsoh> that's easy: you update placement DB and set available BW to negative number
15:11:22 <rubasov> ralonsoh: :-)
15:11:47 <rubasov> ralonsoh: I don't think placement allows a negative inventory
15:12:26 <ralonsoh> Or in this case, placement or neutron-server should raise an exception
15:12:52 <ralonsoh> if you want placement enforcement AND you over provisioned a host, this is your fault
15:12:59 <ralonsoh> (administrator fault)
15:13:17 <rubasov> that's definitely one option, but that's the case where we would break backwards compatibility
15:13:34 <rubasov> since this was working to some extent before (on the api at least)
15:13:49 <ralonsoh> well, IMO, this is not breaking backwards compatibility
15:13:59 <mlavalle> so in this context backs compatibility means the previous status quo
15:14:22 <rubasov> I also hope this will be the general consensus of the friday meeting
15:14:37 <ralonsoh> rubasov, you are right, I'll join this meeting
15:14:39 <mlavalle> ralonsoh: yeah, there is a bit of nomencalture wiggle room here
15:14:46 <rubasov> you're welcome
15:15:12 <rubasov> it'll be a hangouts call
15:15:25 <ralonsoh> is this the biggest architectural problem you have now with this feature?
15:15:57 <rubasov> one of them, I think all currrent problems are listed in the etherpad
15:16:37 <rubasov> the other is: if the nova support is not complete in stein can we still turn it on somehow, so people can get to try it
15:17:08 <rubasov> likely server boot and stop will work, but migrate may not in stein
15:18:00 <ralonsoh> but you almost have the APIs between nova and neutron working
15:18:36 <rubasov> yes, I think the neutron part is much more sure to get completed in stein
15:18:46 <rubasov> and the apis between them too
15:19:31 <ralonsoh> so far, all placement resources are correctly passed between projects
15:19:49 <ralonsoh> and in Nova there are some "corner cases" to be handle
15:20:04 <rubasov> that's a good summary
15:20:23 <ralonsoh> regarding to neutron/tempest patches
15:20:33 <rubasov> I have some fresh code
15:20:37 <rubasov> this patch series
15:20:39 <rubasov> https://review.openstack.org/630999
15:21:00 <rubasov> is now getting close to be able to correct transient errors of the placement sync process
15:21:30 <rubasov> if you have time please review the agent extension patches
15:21:48 * mlavalle added it to his pile
15:22:01 <ralonsoh> and this one too: https://review.openstack.org/#/c/588319/
15:22:02 <rubasov> soon we'll also try lajoskatona's end-to-end test against this patch series
15:22:23 <ralonsoh> this one is mandatory
15:22:33 <rubasov> ralonsoh: yes, that's also reviewable
15:22:37 <lajoskatona> yeah, to see if the tempest scenario really works :-)
15:22:54 <rubasov> the agent extension has its neutron and neutron-lib side of course
15:23:09 <ralonsoh> sure, on the depends-on
15:23:13 <rubasov> lajoskatona: do you have some patches to add?
15:24:26 <lajoskatona> yeah, I think nothing special, the tempest part is under review
15:24:48 <rubasov> then that's all from me now
15:24:58 <lajoskatona> I have to clear it with tempest folks, how they expect the patches
15:25:22 <lajoskatona> yeah, I think so, on Friday we will see how to continue
15:25:47 <rubasov> one more bit
15:25:59 <rubasov> there's a mail thread about the friday meeting
15:26:01 <rubasov> http://lists.openstack.org/pipermail/openstack-discuss/2019-January/001710.html
15:26:22 <rubasov> if there's any change in the plans (I hope not anymore) it'll be advertised there
15:26:30 <ralonsoh> sure!
15:26:52 <ralonsoh> something else?
15:26:57 <rubasov> nope
15:27:06 <ralonsoh> OK, thanks guys! Of course in https://etherpad.openstack.org/p/neutron_qos_meeting_chair we have the links.
15:27:20 <rubasov> thanks
15:27:31 <ralonsoh> next one
15:27:33 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1560963
15:27:34 <openstack> Launchpad bug 1560963 in neutron "[RFE] Minimum bandwidth support (egress)" [Wishlist,In progress] - Assigned to Rodolfo Alonso (rodolfo-alonso-hernandez)
15:27:41 <ralonsoh> patches:
15:27:42 <ralonsoh> #link https://review.openstack.org/#/q/topic:bug/1560963+(status:open)
15:27:54 <ralonsoh> Two parts:
15:28:11 <ralonsoh> first the non-tunneled traffic: https://review.openstack.org/#/c/406841/
15:28:22 <ralonsoh> second one: the rest of the patches
15:28:38 <ralonsoh> (the last patch is still not upstream)
15:28:45 <ralonsoh> reviews are welcome
15:29:01 <mlavalle> so they are ready for reviews then
15:29:27 <ralonsoh> all of them. The pyroute2 patches are failing because a missing patch in neutron-lib
15:29:34 <mlavalle> ack
15:30:02 <ralonsoh> and that's all for this one
15:30:18 <ralonsoh> next one
15:30:19 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1796925
15:30:20 <openstack> Launchpad bug 1796925 in neutron "[RFE] port forwarding floating IP QoS" [Wishlist,New] - Assigned to LIU Yulong (dragon889)
15:30:37 <ralonsoh> the RFE was approved
15:30:46 <ralonsoh> waiting for the spec/patches
15:31:20 <ralonsoh> I don't see LIU here
15:31:25 <mlavalle> slaweq: do you have a meeting with liyyulong today?
15:31:37 <slaweq> mlavalle: tomorow morning my time
15:31:50 <mlavalle> slaweq: would you ask him about this RFE?
15:31:54 <slaweq> sure
15:31:58 <mlavalle> thanks
15:32:11 <ralonsoh> thank you
15:32:29 <ralonsoh> next one
15:32:30 <ralonsoh> https://bugs.launchpad.net/neutron/+bug/1727578
15:32:31 <openstack> Launchpad bug 1727578 in neutron "[RFE]Support apply qos policy in VPN service" [Wishlist,Triaged]
15:32:38 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1727578
15:32:59 <ralonsoh> we have the spec
15:33:03 <ralonsoh> #link https://review.openstack.org/#/c/531074/
15:33:12 <ralonsoh> but there is no progress on this one
15:33:30 <mlavalle> I thought he had re-started working on it
15:33:39 <ralonsoh> that's perfect
15:33:40 <mlavalle> in late December
15:34:03 <mlavalle> that's waht we reported internally
15:34:17 <ralonsoh> so we'll see updates soon (maybe)
15:34:22 <mlavalle> yeah
15:34:26 <ralonsoh> thanks!
15:34:47 <ralonsoh> Ok, no more approved or active RFEs
15:34:49 <mlavalle> he finsihed most of what he's been doing in Ocatvia
15:35:19 <ralonsoh> Am I missing any other one?
15:35:37 <mlavalle> hang on
15:35:40 <ralonsoh> sure
15:36:07 <mlavalle> I'll re-start this one this week: https://review.openstack.org/#/c/613820/
15:36:33 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1777627
15:36:34 <openstack> Launchpad bug 1777627 in neutron "RFE: QoS – rule's management (set,delete, show…) requires <qos-policy> parameter in addition to <rule-id>" [Wishlist,In progress] - Assigned to Miguel Lavalle (minsel)
15:36:49 <ralonsoh> any update on the patch?
15:37:05 <mlavalle> I need to deploy it in my dev system and address the comments
15:37:16 <mlavalle> shouldn't be too difficult
15:37:42 <mlavalle> rebasing + deploying + working on comments
15:38:09 <ralonsoh> added to my review pile and to the etherpad (https://etherpad.openstack.org/p/neutron_qos_meeting_chair)
15:38:14 <mlavalle> thanks
15:38:58 <ralonsoh> ok, let's move on
15:39:04 <ralonsoh> #topic Bugs
15:39:11 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1810025
15:39:12 <openstack> Launchpad bug 1810025 in neutron "L2 agent do not clear QoS rules after restart" [Medium,In progress] - Assigned to Chengqian Liu (liuchengqian90)
15:39:22 <ralonsoh> patch: https://review.openstack.org/#/c/627779/
15:40:19 <ralonsoh> there is one small concern about using a class member
15:40:29 <ralonsoh> but the patch looks good
15:40:35 <ralonsoh> reviews are welcome
15:41:04 <ralonsoh> next one
15:41:05 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1809497
15:41:06 <openstack> Launchpad bug 1809497 in neutron "_get_filterid_for_ip can generate an UnboundLocalError" [High,Confirmed]
15:41:32 <ralonsoh> is anyone here reviewing this one?
15:41:54 <ralonsoh> if not, I'll take a look at this one tomorrow
15:41:55 <slaweq> I think I did patch for it
15:42:10 <slaweq> https://review.openstack.org/#/c/630773/
15:42:20 <slaweq> but I wasn't aware of bug report so I didn't link it
15:42:34 <ralonsoh> OK, so this was the problem
15:42:42 <ralonsoh> can you update the patch link?
15:42:46 <mlavalle> so, let's link it
15:43:14 <slaweq> mlavalle: yep, I'm doing it right now
15:43:21 <mlavalle> thanks
15:43:25 <ralonsoh> perfect, thanks
15:43:49 <ralonsoh> next one
15:43:52 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1784006
15:43:53 <openstack> Launchpad bug 1784006 in neutron "Instances miss neutron QoS on their ports after unrescue and soft reboot" [Medium,Confirmed] - Assigned to Miguel Lavalle (minsel)
15:43:54 <slaweq> done
15:44:24 <ralonsoh> mlavalle, ?
15:44:38 <mlavalle> didn't make progress here
15:44:46 <ralonsoh> no problem
15:45:26 <ralonsoh> last one
15:45:34 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1785189
15:45:35 <openstack> Launchpad bug 1785189 in neutron "Floatingip and router bandwidth speed limit failure" [Undecided,In progress] - Assigned to Brian Haley (brian-haley)
15:45:55 <ralonsoh> patch: https://review.openstack.org/#/c/596637/
15:46:16 <ralonsoh> waiting for Lu replies
15:46:57 <ralonsoh> mlavalle, if next time (in two weeks) there are no updates, I can take it
15:47:05 <mlavalle> ok
15:47:15 <ralonsoh> perfect
15:47:35 <ralonsoh> I don't have more bugs in my list
15:47:41 <ralonsoh> Am I missing something?
15:48:18 <ralonsoh> ok, next section
15:48:21 <ralonsoh> #topic Open Discussion
15:48:32 <ralonsoh> please, any update/comment?
15:49:29 <ralonsoh> 3 2 1
15:49:42 <ralonsoh> thank you all for attending
15:49:46 <mlavalle> Thanks
15:49:48 <rubasov> thank you guys
15:49:49 <mlavalle> o/
15:49:55 <ralonsoh> see you in two weeks (and this friday)
15:49:59 <ralonsoh> #endmeeting