15:00:22 <tbarron> #startmeeting manila 15:00:23 <openstack> Meeting started Thu Apr 25 15:00:22 2019 UTC and is due to finish in 60 minutes. The chair is tbarron. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:24 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:26 <openstack> The meeting name has been set to 'manila' 15:00:41 <carloss> hey :) 15:00:43 <vhariria> hi 15:00:45 <jgrosso> Hi all! 15:00:47 <ganso> hello 15:00:58 <tbarron> manila courtesy ping: gouthamr xyang toabctl bswartz ganso erlon tpsilva vkmc amito jgrosso vhariria 15:01:02 <bswartz> .o/ 15:01:09 <_erlon_> hey 15:01:14 * bswartz is back from vacation 15:01:16 <lseki> o/ 15:01:26 <amito> o/ 15:01:27 <tbarron> bswartz: hope it was a good one :) 15:01:42 * tbarron waits a couple 15:02:29 <gouthamr> o/ 15:02:52 <tbarron> ok, we've got a pretty good group 15:03:00 <tbarron> Hi everyone! 15:03:12 <xyang> hi 15:03:19 <tbarron> #link Agenda: https://wiki.openstack.org/wiki/Manila/Meetings 15:03:41 <tbarron> We don't have a whole lot to cover today. 15:03:49 <tbarron> #topic Announcemenets 15:03:56 <tbarron> #topic Announcements 15:04:07 <tbarron> I don't have any. Anyone? 15:04:19 <bswartz> tbarron: we've successfully make "open discussion" last 55 minutes before 15:04:26 <bswartz> >_< 15:04:45 * tbarron can see bswartz is full of energy from his vacation 15:05:03 <tbarron> #topic Summit and PTG 15:05:05 <bswartz> I'm not suggesting we try again... 15:05:14 <tbarron> just reminders 15:05:26 <tbarron> We won't have *this* meeting next week 15:05:44 <tbarron> Monday-Wednesday Summit, Thurs-Saturday PTG 15:06:00 <tbarron> PTG will start at 9am Thursday mountain time US 15:06:11 <tbarron> That's UTC-6, right? 15:06:33 <tbarron> here are the manila summit topics 15:06:39 <bswartz> tbarron: yes 15:06:52 <tbarron> #link https://www.openstack.org/summit/denver-2019/summit-schedule/global-search?t=Manila 15:07:12 <tbarron> here's a link to PTG schedule: 15:07:23 <_erlon_> tbarron: do we have the schedule for the PTG topics? 15:07:26 <_erlon_> ow 15:07:26 <tbarron> #link https://www.openstack.org/ptg#tab_schedule 15:07:28 <_erlon_> :) 15:07:44 <tbarron> here is the etherpad for manila: 15:08:02 <tbarron> #link https://etherpad.openstack.org/p/manila-denver-train-ptg-planning 15:08:11 <tbarron> _erlon_: thanks for asking 15:08:27 <tbarron> I will turn this into a provisional schedule this week 15:08:58 <tbarron> so if you are remote and need a topic at a particular time please indicate that on the etherpad 15:09:07 <_erlon_> tbarron: provisional? 15:09:17 <tbarron> the schedule will always be somewhat provisional and fluid 15:09:40 <tbarron> _erlon_: we may make up time in some areas and find we need more time in others 15:09:44 <tbarron> _erlon_: make sense? 15:10:09 <tbarron> but we can make pretty firm commitments for remote attendees as needed 15:10:23 <xyang> tbarron, looks like Manila and Cinder PTG sessions will happen at the same time? 15:10:31 <_erlon_> hmm, ok, my concern is conflics with the Cinder discussions some of us might need to attend 15:10:35 <tbarron> xyang: there will be some overlap 15:10:58 <_erlon_> so, time slots would be nice to coordinate that 15:11:07 <tbarron> xyang: _erlon_ - so that is one reason we may discover that we need to move stuff around 15:11:37 <tbarron> _erlon_: xyang others - if there are cinder slots that you definitely want to attend and you know when they are please indicate that along 15:11:52 <tbarron> with stuff in manila that you really don't want to miss 15:12:31 <tbarron> if you see topics of general interest to manila in cinder then we may crash their session for those 15:13:11 <tbarron> The scheduling probably won't be perfect but it will be a lot better if people add information to the etherpad in advance. 15:13:19 <tbarron> Anything else on this topic? 15:13:25 <tbarron> Oh 15:13:40 <tbarron> Will FIT be able to bring the webcam/mic again? 15:13:57 <tbarron> I think ganso used to do that ... 15:14:29 <_erlon_> tbarron:sure we can take it 15:14:44 <tbarron> We'll set up a bluejeans session for remote and announce it on #openstack-manila (meeting topic) that morning 15:14:51 <tbarron> _erlon_: awesome, thanks! 15:15:04 <bswartz> Bluejeans? 15:15:06 <bswartz> Not zoom? 15:15:25 <tbarron> bswartz: anyone is welcome to set up zoom session if they want 15:15:26 <_erlon_> zoom++ 15:15:32 <_erlon_> :) 15:15:37 <tbarron> I was just repeating the past. 15:15:49 <_erlon_> tbarron: zoom is the new hipe now ;) 15:15:50 <bswartz> Oh I remember the bad old days of webex 15:15:59 <bswartz> Zoom makes me happy 15:16:08 <bswartz> Bluejeans is okay too 15:16:09 <_erlon_> bswartz: +1 15:16:16 <carloss> bswartz: +1 15:16:47 <lseki> zoom++ 15:17:08 <tbarron> some people don't like open source solutions I guess :) 15:17:37 <tbarron> I'm gonna delegate the VC setup so I don't have a vote, will use whatever ... 15:17:43 <_erlon_> tbarron: can I install BJ in my server? 15:18:11 <lseki> actually I'm ok with anything but webex 15:18:18 <tbarron> _erlon_: in principle I guess, for our purposes that's not needed though 15:18:32 <tbarron> bluejeans is a memory hog 15:18:38 <tbarron> the client 15:18:40 <bswartz> I'd prefer open source, but I wasn't aware that zoom was less open source than bluejeans 15:18:55 <bswartz> I just note that they have excellent Linux support 15:19:15 <tbarron> bswartz: I haven't researched it but I think that's the case, the main thing is that we have something reliable. 15:19:34 <gouthamr> yeah bluejeans has pretty messy browser support 15:19:37 <_erlon_> tbarron: just joking about the 'opensourceness' of it 15:19:54 <jgrosso> gouthamr: !! 15:20:03 <tbarron> _erlon_: gotcha 15:20:15 <tbarron> #topic Stein retrospective 15:20:52 <tbarron> We'll start Thursday morning with a Stein retrospective since doing Rocky retrospective last time seemed worthwhile. 15:21:01 <tbarron> I want to keep it to a half hour. 15:21:18 <tbarron> But we can spin off topics if we need to and add them to the agenda. 15:21:41 <tbarron> It will help for people to think in advance of things that went well and especially 15:21:43 <gouthamr> jgrosso: :P hey, but it's got infinite recording quota, which we'll need... 15:21:49 <tbarron> things that we should do differently. 15:22:08 <tbarron> So we have an etherpad for this 15:22:22 <tbarron> #link https://etherpad.openstack.org/p/manila-stein-retrospective 15:22:42 <tbarron> You'll likely think of some things that need improvement in the coming week. 15:23:04 <tbarron> So please bookmark that so you can take 30s to put such items on the list. 15:23:21 <tbarron> Questions? Comments? 15:23:49 <tbarron> #topic Train Schedule 15:23:56 * tbarron has fun saying that 15:24:19 <tbarron> I've put up a WIP review with manila deadlines here: 15:24:35 <tbarron> #link https://review.opendev.org/#/c/655667/ 15:24:42 <tbarron> Please take a look now. 15:24:57 <tbarron> Maybe we don't need to spend PTG time on it. 15:25:40 <tbarron> Here is the rendered schedule from that review: 15:25:47 <bswartz> tbarron: I imagine a lot of people voted for "train" because of the possible puns 15:25:53 <tbarron> #link http://logs.openstack.org/67/655667/1/check/openstack-tox-docs/553c2e7/html/train/schedule.html 15:26:06 <tbarron> bswartz: +1 15:26:24 <gouthamr> We have a spec proposal freeze? :P 15:26:38 <gouthamr> noooo 15:26:40 <tbarron> This dev cycle is shorter than last time but w/o as many holidays 15:26:56 <tbarron> gouthamr: I added that to see if anyone was paying attention 15:27:10 <tbarron> gouthamr: keystone had it, so it seemed worth thinking about 15:27:11 * gouthamr passed 15:27:42 <tbarron> The advantage would be that we don't have people like gouthamr dropping specs on us at the last minute. 15:27:52 <tbarron> The disadvantage is that ^^^^^ 15:28:00 <tbarron> What do people think? 15:28:36 <bswartz> I'm neutral 15:28:48 <tbarron> There are otherwise no new project specific deadlines added. 15:29:01 <tbarron> gouthamr: I'm happy to drop that one. 15:29:14 <tbarron> I don't think specs at the last minute are that big of an issue. 15:29:32 <tbarron> People who do it better have damn good specs or they won't be approved. 15:29:36 <gouthamr> tbarron: yeah, it doesn't make too much sense to me to separate spec proposal and spec freeze 15:29:49 <tbarron> Submitting right before a deadline is always risky. 15:30:04 <tbarron> We have finite reviewer bandwidth, etc. 15:30:27 <tbarron> OK, unless someone wants to keep the spec proposal deadline I will re-submit with that gone. 15:30:44 <tbarron> So then we have Spec Freeze at Train-2 15:30:56 <tbarron> Also driver submission deadline at Train-2 15:31:05 <tbarron> by 26 July. 15:31:39 <tbarron> Feature proposal freeze R-9 week, 16 August 15:31:40 <gouthamr> ack, we've gotten better at tracking ongoing work, so i think it's worth just having one deadline 15:32:25 <tbarron> Feature Freeze matches the general freeze, at Train-3 milestone, 13 Sept. 15:32:54 <tbarron> Anyone have issues with these dates? 15:33:01 <gouthamr> awesome 15:33:31 <tbarron> OK, I'll make the revision as we discussed and remove the WF -1. 15:34:00 <tbarron> If upon reflection any of you want to make a change then add it to the PTG agenda and we can revisit. 15:34:28 <tbarron> But I'd like to go on and merge our plan as of today so that we can present that to the community at Summit. 15:34:54 <tbarron> #topic Bugs 15:35:00 <tbarron> jgrosso: got any for us? 15:35:17 <jgrosso> thanks tbarron I have a few for us 15:35:26 <jgrosso> https://bugs.launchpad.net/manila/+bug/1713062 15:35:27 <openstack> Launchpad bug 1713062 in Manila "Missing ability to automatically build configuration reference artifacts" [High,Triaged] 15:36:32 <gouthamr> been High for a few cycles now 15:36:36 <tbarron> jgrosso: what outcome are you looking for on this one? It's a genuine issue. 15:36:51 <jgrosso> is anyone looking at it 15:37:07 <gouthamr> we need an owner, oslo-config has some new tooling we can consider using 15:37:08 <jgrosso> been sitting at high for a year 15:37:33 <gouthamr> cinder added a "config table" sphinx extension to render options cleanly for different drivers 15:38:15 <tbarron> gouthamr: if you happen to have that review then pls. add a pointer to it to this bug 15:38:24 <gouthamr> this thing: https://github.com/openstack/cinder/blob/master/doc/ext/driver_opts.py 15:38:41 <tbarron> jgrosso: we're going to have to figure how to manage our backlog but 15:38:50 <gouthamr> ack 15:39:00 <tbarron> jgrosso: you are still working to reduce it so that we can see it clearly 15:39:13 <tbarron> jgrosso is doing an awesome job on that front 15:39:26 <tbarron> jgrosso: how many bugs are we down to? 15:39:27 <jgrosso> thanks doing my best :) 15:39:37 <jgrosso> 202 open 15:39:42 <tbarron> he's clearing out old probably unreproducible bugs 15:40:06 <tbarron> jgrosso: do you have a sense of how many of those are driver bugs? 15:40:18 <jgrosso> looks like 30 ish 15:40:22 <tbarron> jgrosso: those we should make separate backlog queues for 15:40:40 <tbarron> jgrosso: so we still have around 170 "common core" bugs? 15:40:42 <jgrosso> Yeah I know we spoke about separate queues 15:41:09 <jgrosso> we have 32 doc 15:41:56 <tbarron> we're getting close to the point where we can see stuff well enough to have some meetings where we re-prioritize and devise 15:42:07 <tbarron> some dashboards with the separate backlogs 15:42:18 <jgrosso> that would be great 15:42:41 <tbarron> I'm relucatant to just volunteer for a bug because no one else is raising their hand. 15:43:04 <tbarron> That just gives an appearance that we're taking care of stuff unless I really have bandwidth to work on it. 15:43:25 <tbarron> So we didn't get a volunteer for this one. 15:43:35 <jgrosso> understood and that is a good thing no reason to assign it if you cant work on it 15:44:05 <jgrosso> I am trying to get the high ones looked at first 15:44:18 <tbarron> jgrosso: yes, thanks. 15:44:23 <jgrosso> maybe we should table this until after PTG 15:44:31 <jgrosso> I know people are very busy 15:44:38 <tbarron> And I agree that this one is high priority. 15:44:50 <tbarron> But it's not *urgent*. 15:45:13 <jgrosso> ok 15:45:43 <jgrosso> some I just see HIGH and unassigned 15:45:47 <jgrosso> https://bugs.launchpad.net/manila/+bug/1666547 15:45:48 <openstack> Launchpad bug 1666547 in Manila "poor api performance with postgresql backend" [High,In progress] 15:45:50 <jgrosso> like this one 15:46:01 <tbarron> Why is this high? 15:46:34 <bswartz> Someone cares a lot about postgres no doubt 15:46:35 <tbarron> Is there a manila user reporting that this is hurting them? 15:46:58 <tbarron> Maybe SAP 15:47:03 <bswartz> mkoderer says so 15:47:10 <tbarron> SAP has resources, they should propose a fix. 15:47:10 <jgrosso> Look like a fixed was released for Cinder 15:47:31 <bswartz> The issue is that indices are not created 15:47:33 <jgrosso> I just saw it was unassigned 15:47:43 <bswartz> If that's true, it could be an easy fix with high positive impact 15:47:52 <tbarron> jgrosso: I'm not challengin you, thanks for bringing it to our attention! 15:47:57 <bswartz> https://review.opendev.org/#/c/437346/ 15:48:15 <jgrosso> no worries 15:48:18 <tbarron> I'm just speaking as PTL from a business perspective, it may be a low hanging fruit and we'd welcome a fix. 15:48:52 <jgrosso> completely understand :) 15:49:12 <tbarron> cinder has a much simpler DB schema than we do 15:49:23 <jgrosso> cool I had no idea 15:49:31 <tbarron> we have tons of joins 15:49:44 <tbarron> but I digress 15:50:15 <tbarron> I don't think this fix is rocket science but I do think it requires care. 15:51:01 <gouthamr> needs a new tag on launchpad: 'mid-tree-hanging-fruit' 15:51:07 <jgrosso> :) 15:51:14 <tbarron> If no one objects I'm going to downgrade this to Medium with the remark that if someone is affected by it they can bump it back up. 15:51:25 <tbarron> For all I know SAP has fixed it in their fork. 15:51:32 <jgrosso> tbarron: sounds good to me 15:52:00 <jgrosso> That is all I had for now 15:52:20 <tbarron> jgrosso: awesome, thanks so much for what you are doing 15:52:30 <gouthamr> jgrosso++ 15:52:31 <jgrosso> tbarron: your very welcome ! 15:52:36 <jgrosso> :) 15:52:51 <tbarron> jgrosso: would you add that config doc bug as a PTG topic? 15:52:59 <tbarron> maybe we can get a plan for it there. 15:53:04 <jgrosso> sure 15:53:25 <tbarron> #topic open discussion 15:53:53 <tbarron> So as you can see we like other openstack projects are resource challenged 15:54:23 <tbarron> much of my work, and of manila colleagues from my company 15:54:41 <tbarron> is with deployments and bug fixes that customers have reported vs 15:55:00 <tbarron> bugs that someone has found upstream and is no longer making any noise about 15:55:39 <tbarron> So we're doing our best to get bugs and planned work in better focus since we can't just 15:56:06 <tbarron> rely on the fact that OpenStack is popular and developers are clamoring to be a part of it and will fix 15:56:22 <tbarron> arbitrary stuff in order to make their way into the community. 15:56:49 <tbarron> Fixing bugs with the generic driver is a huge time sink. 15:57:01 <jgrosso> ill bet it is 15:57:28 <tbarron> But other open source back ends that people are actually invested in, like Ceph 15:57:45 <tbarron> have feature gaps so that they can't yet serve as a replacement reference driver. 15:58:13 <tbarron> Just sayin ..... :) 15:59:09 <tbarron> looks like we managed to spend a perfectly good hour of your time today! 15:59:16 <jgrosso> :) 15:59:20 <tbarron> See you at PTG and on #openstack-manila 15:59:30 <gouthamr> o/ 15:59:30 <tbarron> Bye all, and thanks! 15:59:36 <tbarron> #endmeeting