14:00:48 <ihrachys> #startmeeting networking
14:00:49 <johndperkins> hey
14:00:50 <openstack> Meeting started Tue Jun 21 14:00:48 2016 UTC and is due to finish in 60 minutes.  The chair is ihrachys. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:50 <mlavalle> o/
14:00:50 <bcafarel> hi
14:00:51 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:54 <openstack> The meeting name has been set to 'networking'
14:00:55 <john-davidge> hi
14:00:56 <andreas_s> o/
14:00:56 <hichihara> hi
14:00:57 <amuller> hiya
14:00:57 <haleyb> hi
14:01:03 <hoangcx> hi
14:01:04 <akamyshnikova> hi
14:01:04 <annp> Hi
14:01:08 <namnh> hi
14:01:10 <HenryG> o/
14:01:10 <sbelous_> hi!
14:01:25 <ihrachys> I will lead the meeting, but note that armax is probably somewhere around :)
14:01:35 <ihrachys> #link https://wiki.openstack.org/wiki/Network/Meetings Agenda
14:01:45 <yamamoto> hi
14:01:55 <korzen> hi
14:02:02 <ivc_> hi
14:02:10 <ihrachys> let's run thru the agenda really quick, we may have the meeting packed
14:02:20 <ihrachys> #topic Announcements
14:02:35 <ihrachys> just repeating what HenryG said the prev meeting...
14:02:35 <ihrachys> The mid-cycle is August 17-19 in Cork, Ireland.
14:02:39 <ihrachys> #link https://etherpad.openstack.org/p/newton-neutron-midcycle
14:02:43 <ihrachys> please register if you go
14:02:51 <ihrachys> any more announcements?
14:03:18 <ihrachys> nope, moving on :)
14:03:30 <ajo> o/
14:03:37 <obondarev> o/
14:03:49 <ihrachys> I expect blueprints to take a vast slot in the meeting, so let's move them to later.
14:03:56 <ihrachys> #topic Bugs
14:04:11 <ihrachys> the prev week the deputy was dasm
14:04:26 <ihrachys> dasm: any highlights?
14:04:27 <dasm> yep.
14:04:41 <dasm> week was pretty calm.
14:04:47 <dasm> we had two gate issues
14:05:05 <dasm> one with neutron-lbaas which broke neutron gates
14:05:15 <dasm> and other, fixed by ihrachys
14:05:29 <dasm> right now everything seems to be under control
14:05:50 <ihrachys> thanks
14:05:53 <dasm> this week, bug deputy is john-davidge
14:05:54 <ihrachys> we have more gate failures
14:05:59 <ihrachys> #link https://bugs.launchpad.net/neutron/+bugs?field.tag=gate-failure
14:06:00 <john-davidge> o/
14:06:11 <john-davidge> Yes, there are a couple of gate failures so far this week
14:06:32 <john-davidge> #link https://bugs.launchpad.net/neutron/+bug/1594796 Is the most recent
14:06:33 <openstack> Launchpad bug 1594796 in neutron "test_api_extension_validation_with_good_dns_names fails with 500 error" [Critical,Confirmed]
14:06:55 <ihrachys> right. that one I think is looked at by pavel-bondar right now. but other l3 folks are welcome to assist.
14:07:09 <carl_baldwin> I just noticed it.  I'm looking in to it also.
14:07:18 <john-davidge> We also have #link https://bugs.launchpad.net/neutron/+bug/1594376 which appears to be more intermittent
14:07:18 <openstack> Launchpad bug 1594376 in neutron "Delete subnet fails with "ObjectDeletedError: Instance '<IPAllocation at 0x7f6abc3abe50>' has been deleted, or its row is otherwise not present."" [High,Confirmed]
14:07:18 <ihrachys> thanks carl_baldwin
14:08:18 <ihrachys> carl_baldwin: are you aware of that one?
14:08:36 <ihrachys> in logstash, I see 33 hits in last 7 days
14:08:37 <mlavalle> ihrachys, carl_baldwin: also looking
14:08:45 <carl_baldwin> I am now.  :)
14:08:59 <ihrachys> ok good. let's own the gate.
14:08:59 <carl_baldwin> I will look in to it once the first is resolved.
14:09:19 * carl_baldwin owning it. :)
14:09:24 <ihrachys> :)
14:09:39 <ihrachys> #topic Docs
14:09:55 <ihrachys> I haven't seen Sam-I-Am for a while
14:10:07 <ihrachys> I wonder whether he moved to greener pastures ;)
14:10:25 <ihrachys> anyone willing to update on behalf of docs subteam?
14:10:35 <haleyb> he does moo a lot :)
14:11:49 <ihrachys> ok, I guess we skip, but we need to solve the issue of no docs representation during the meeting. maybe armax will want to follow up on that.
14:11:56 <sc68cal> I'm around.
14:12:09 <ihrachys> sc68cal: stage is yours.
14:13:05 <sc68cal> I think probably the only thing to report is that Sam-I-Am is away so there's slack to be taken up
14:13:58 <sc68cal> i'll see if I can reach out to edgar since he's our liaison?
14:13:59 <carl_baldwin> The weather here in Colorado is so good he might just be out enjoying it.
14:14:18 <ihrachys> sc68cal: thanks!
14:14:31 <ihrachys> yeah, summers are counter productive
14:14:41 <ihrachys> #topic Transition to OSC
14:15:00 <ihrachys> dasm: you have an update for OSC right?
14:15:58 <dasm> ihrachys: no.. i just added it to agenda, on behalf of amotoki
14:16:01 <rtheis> Progress continues in OSC to enhance and add networking commands
14:16:26 <ihrachys> dasm: ouch. I always assume that's you, since you are apparently everywhere!
14:16:32 <rtheis> floating ip and rbac support out for review
14:17:06 <amuller> rtheis: can you link a resource tracking this work?
14:17:13 <ihrachys> rtheis: would neutron reviewers be helpful there, or you think it's better to leave it to OSC core team?
14:17:27 <rtheis> https://etherpad.openstack.org/p/osc-neutron-support
14:18:22 <amuller> rtheis: thank you
14:18:22 <rtheis> ihrachys: I think OSC team is doing okay to keep up, but reviews are always welcome
14:18:48 <ihrachys> great to hear.
14:19:12 <ihrachys> #topic Moving to Keystone v3 API in Neutron
14:19:20 <ihrachys> dasm: now that should be your slot
14:19:21 <dasm> ihrachys: that's me :D
14:19:38 <dasm> yes. docs were changed and merged.
14:19:52 <ihrachys> user docs or dev docs?
14:19:59 <dasm> devref
14:20:34 <dasm> it will probably require additional lookup, after db and internal code will be changed.
14:20:41 <ihrachys> dasm: I bet we have work to do in networking guide. it's on radar right?
14:20:51 <dasm> to update it, and notify all developer changes
14:20:57 <dasm> ihrachys: yes, i have this also in my mind.
14:21:18 <ihrachys> dasm: do we have mergeable reviews in neutron repos for the feature?
14:21:37 <dasm> not yet. i'm working on offline db migrations
14:21:50 <dasm> to change column names
14:21:55 <ihrachys> ok, enjoy :)
14:22:05 <dasm> :D
14:22:20 <ihrachys> now, I'd like to get back to the blueprints section we skipped before, but before I do...
14:22:25 <ihrachys> #topic Review velocity
14:22:43 <ihrachys> armax: wanna cover that one?
14:23:18 <ihrachys> :D
14:23:27 <ihrachys> ok, I guess I will need to communicate that one myself
14:24:21 <ihrachys> basically, armax have some concerns about how we do reviews, both quantitatively as well as qualitatively
14:24:58 <ihrachys> for the former, it's as simple as: neutron core reviewers don't seem to keep up with the patch load, and some don't seem to try
14:25:49 <ihrachys> for the latter, it's more about focus on targeted features, not quality of particular reviews per se
14:25:50 <hichihara> :p
14:26:27 <ihrachys> the team seems to be quite distributed (or scattered, depending on your perspective) and does not seem to have a common focus
14:26:51 <njohnston> Hopefully the mid-cycle will help focus everyone
14:26:59 <ihrachys> that's in drastic comparison to e.g. nova team that is very focused (maybe more than you would like)
14:27:05 <amuller> The dashboard rossella initiated should probably be mentioned at this point, linked from http://status.openstack.org/reviews/
14:27:07 <dasm> njohnston: yeah, but midcycle is pretty late.
14:27:43 <ihrachys> njohnston: midcycle is a single week long event only, it doesn't tackle the culprit. also it's targeted to IRL participants
14:27:50 <njohnston> true
14:28:05 <ihrachys> amuller: that's a great notice, and yes, I actually wanted to mention it
14:28:18 <john-davidge> Perhaps this can be articulated in detail on the ML?
14:28:39 <ajo> that's right, we should focus on making use our gerrit dashboard as a primary source
14:28:40 <ihrachys> john-davidge: that's part of the deal. but it would be great if we run it here first.
14:29:11 <ihrachys> I would ask current core reviewers, what could help us to stick to common causes, and overall, bump review stats?
14:30:16 <amuller> We try to go through new RFEs in the drivers meeting
14:30:22 <ihrachys> there is a significant number of patches in the review queue that does not seem to be taken care by review team as much as you could expect. is it because we target wrong things that no one cares? or everyone is too busy with specific bits they already have on plates?
14:30:24 <amuller> but we don't have recurring meetings to track ongoing work
14:30:46 <ihrachys> amuller: in a way, the meeting like the one today are supposed to be used for that, too
14:31:24 <amuller> ihrachys: Aye, it's not a crazy idea IMO to add a section to this meeting, or to start a new recurring meeting for cores and anyone interested to track ongoing work, see if there's high priority items that are stuck and not getting reviews
14:31:30 <ajo> may be we should keep a section on the meeting for this purpose
14:31:48 <ihrachys> ajo: there is actually one for blueprints.
14:32:00 <ajo> and ping the people responsible for the things that are starving reviews
14:32:05 <ihrachys> but maybe the time slot we have for neutron team meetings is just not enough to cover anything in details.
14:32:19 <ihrachys> ajo: how do you determine who is responsible?
14:32:34 <ajo> yeah, first one, is the assigned reviewer to an RFE,
14:32:37 <amuller> ihrachys: armax said a while ago that each blueprint should have an approver or a responsible core
14:32:40 <ihrachys> we have approvers in LP blueprints, but no approvers for RFEs and bugs.
14:32:56 <ajo> if it's not an RFE, and it's a bug, then , the core reviewers of the specific area?
14:33:00 <ajo> maY be we could automate that
14:34:09 <ajo> may be we could add some sort of tag to bug description to specify who may be doing reviews on that bug ?
14:34:29 <ihrachys> I would suggest we don't target anything that does not have at least two cores + backup to drive the feature. it may reduce the number of targeted features, but that would just reflect the reality where 80% of things we target right now will not probably happen this cycle.
14:34:30 <ajo> it's a bit of micromanagement may be, but, it's a way to make sure people get a more prioritized review inbox
14:34:53 <ihrachys> ajo: a comment in LP may be good enough to have people responsible.
14:35:25 <ihrachys> ajo: or you mean, we generate personalized gerrit dashes?
14:35:33 <ajo> ihrachys, I personally cope very bad with all the amount of email my inbox receives, specially bug notifications, even if I try to keep up with reading them all...
14:35:37 <ihrachys> that would probably be hard to maintain
14:35:41 <ajo> ihrachys, personalized gerrit dashes
14:36:05 <ajo> well... to be fair
14:36:07 <ihrachys> is the current dash that we have from rossella_s too big to digest?
14:36:10 <ajo> we can already add people to a review
14:36:16 <ajo> and it will show in the standard gerrit dash
14:36:29 <ajo> (the standard : no dash)
14:37:29 <ihrachys> yeah. the only problem I have with that is people adding random people to their patches just because they are on core team
14:38:18 <ihrachys> any more thoughts from other team members? do you think the way it goes is good enough?
14:38:41 <john-davidge> It sounds like a well documented, well focused custom dash for core reviewers could go a long way to solving some of this
14:39:19 <ihrachys> john-davidge: but there is already one, that shows just targeted pieces, and there are actually not that many patches there
14:39:26 <ajo> yes, may be per-core customized links, published in shorteners.
14:39:30 <ihrachys> who uses the dashboard from rossella_s?
14:39:31 <ihrachys> o/
14:39:40 <ajo> we tag by using bug comments, or in the description
14:39:46 <ajo> reviewers: a,b,c,d
14:40:25 <ajo> I tend to forget using it, I only do it when my default queue drains..
14:40:47 <hichihara> me too ;)
14:41:02 <ihrachys> ajo: does it happen? I mean, the queue drain?
14:41:14 <ajo> ihrachys, almost never TBH, just sometimes
14:41:26 <ajo> so, this is, why in my case, I should probably switch the priority around
14:41:37 <ihrachys> so that's the point. unless you say no to untargeted stuff, it just sneaks into your mind
14:41:51 <ajo> true
14:42:39 <HenryG> easier said than done
14:42:45 <ihrachys> ok, I guess other cores are too shy to speak up or smth. I guess armax will need to follow up on the topic in ML anyway.
14:42:51 <ajo> HenryG, true
14:43:43 <ihrachys> let's run thru some blueprints really quick
14:43:49 <ihrachys> #topic Blueprints
14:43:52 <ihrachys> #link https://launchpad.net/neutron/+milestone/newton-2
14:44:00 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/troubleshooting
14:44:08 <ihrachys> amuller: does it happen in N?
14:44:20 <ihrachys> afaik it's still in spec discussion mode
14:44:22 <amuller> ihrachys: it's waiting for reviews on the spec
14:44:32 <amuller> ihrachys: Hynek added a new revision today
14:44:41 <hynekm_> it would appreciate reviews
14:45:00 <ihrachys> amuller: you are an approver. do you think it's ready?
14:45:19 <amuller> ihrachys: I haven't read the new revision from today, I last talked to Hynek last week for the new direction
14:46:08 <ihrachys> note that we approach N2 in several weeks. which makes chances of anything with no code patches in review quite low
14:46:27 <ihrachys> amuller: I see armax is on review list for the spec. is there agreement?
14:47:09 <HenryG> Should this BP be 'High' priority?
14:47:58 <HenryG> I mean in relation to some of the others?
14:49:05 <HenryG> Perhaps in the next drivers meeting the priorities can be discussed.
14:49:40 <ihrachys> HenryG: that's a good idea. there is little reason to approve more RFEs while we struggle to deliver half or what we targeted.
14:50:14 <amuller> ihrachys: we'll see :)
14:50:26 <ajo> I just ask to make a 2nd-pass on the qos-rules extended validation RFE, as it's a blocker for the other QoS RFEs
14:50:52 <ihrachys> ajo: you mean, you have more to add to what was discussed the prev drivers meeting?
14:51:09 <ajo> ihrachys, I added on the bug itself, I agree on the reduced scope of the RFE
14:51:14 <ihrachys> ack
14:51:18 <ihrachys> next in the list is
14:51:22 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/multi-l3-backends
14:51:43 <ihrachys> iyamahat: that's on you. what's the status here.
14:52:22 <ihrachys> there is a link to a gerrit topic in LP, but it does not seem to work
14:52:54 <ihrachys> ok, I guess we go next
14:52:57 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/push-notifications
14:52:57 <hichihara> The feature is managed by Kevin now
14:53:03 <ihrachys> hichihara: I see
14:53:08 <ihrachys> that one is also on kevinbenton
14:53:23 <ihrachys> I am approver. we landed some first patches for revision numbers already, more to come.
14:53:40 <ihrachys> that and subnet object that korzen is working on should unblock the progress there.
14:53:48 <ihrachys> I think it's on good track.
14:53:57 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/vlan-aware-vms
14:54:24 <ihrachys> rossella_s is not here, but maybe someone else wants to update us.
14:54:54 <ihrachys> there's a huge queue for reviews:
14:54:55 <ihrachys> #link https://review.openstack.org/#/q/topic:bp/vlan-aware-vms+status:open
14:55:48 <ihrachys> I believe we closed the nova interaction concern in the ML thread, so it's a matter of reviews.
14:56:27 * ihrachys feels lonely
14:56:32 <ihrachys> ok, the last one for today
14:56:33 <ihrachys> #topic https://blueprints.launchpad.net/neutron/+spec/fwaas-api-2.0
14:56:34 <hichihara> It seems there is progress but many -1
14:56:44 <ihrachys> sc68cal: do you want to update on that one?
14:56:46 <hichihara> I'm here :)
14:56:52 <ihrachys> I see some patches at https://review.openstack.org/#/q/status:open+project:openstack/neutron-fwaas+branch:master+topic:fwaas_v2_api
14:57:10 <ihrachys> but nothing that would land so far
14:57:12 <njohnston> The prerequisite for that is the l3 agent extensions work
14:57:13 <njohnston> https://review.openstack.org/#/c/315745/16
14:57:23 <sc68cal> and a devstack plugin
14:57:43 <sc68cal> basically are people serious about fwaas? because the resources haven't really been put forward
14:57:48 <ihrachys> njohnston: it's weird, fwaas v2 is High, but your thing is Low
14:57:50 <njohnston> sc68cal: yes; there is work on the devstack plugin currently
14:57:55 <ihrachys> I believe we should bump l3 extension thing
14:58:22 <ihrachys> sc68cal: I am good to ditch it. I was good to do it at the summit. but armax is the king.
14:59:09 <ihrachys> we need to decide when we do the call on removal of those repos from neutron
14:59:18 <ihrachys> okay, that's all for bp
14:59:22 <ihrachys> one small thing before we end the meeting
14:59:30 <ihrachys> #topic Other stuff
15:00:04 <ihrachys> some of you may want to know that we landed a patches for governance lately that changed the way we manage governance tags for neutron bits
15:00:07 <ihrachys> #link https://review.openstack.org/#/c/323522/
15:00:18 <ihrachys> that removed some tags like supports-upgrade from *aas repos
15:00:28 <hichihara> I will check
15:00:32 <ihrachys> which is fine since it only reflects reality (of missing grenade gating for those repos)
15:00:36 <ihrachys> so just a heads up
15:00:54 <ihrachys> ok, that's all. we will follow up on remaining bps next week.
15:00:57 <ihrachys> cheers
15:01:00 <ihrachys> #endmeeting