16:00:17 #startmeeting Cinder 16:00:19 Meeting started Wed Aug 23 16:00:17 2017 UTC and is due to finish in 60 minutes. The chair is smcginnis. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:20 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:23 The meeting name has been set to 'cinder' 16:00:25 ping dulek duncant eharney geguileo winston-d e0ne jungleboyj jgriffith thingee smcginnis hemna xyang1 tbarron scottda erlon jbernard _alastor_ bluex karthikp_ patrickeast dongwenjuan JaniceLee cFouts Thelo vivekd adrianofr mtanino karlamrhein diablo_rojo jay.xu jgregor lhx_ rajinir wilson-l reduxio wanghao thrawn01 chris_morrell watanabe.isao,tommylikehu mdovgal ildikov wxy viks ketonne 16:00:31 abishop sivn 16:00:32 hi 16:00:34 Hi 16:00:35 <_alastor_> o/ 16:00:36 hey 16:00:40 hi 16:00:47 Hello :) 16:00:49 hi 16:01:02 o/ 16:01:06 Hi 16:01:13 o/ 16:02:01 o/ 16:02:26 Hi. 16:02:31 OK, looks like probably enough to get going. 16:02:36 #topic Announcements 16:02:43 The end is near! 16:02:54 Still 16:03:19 We will need to cut a Cinder RC2 today, or early tomorrow at the latest. 16:03:26 * DuncanT passes Jay a sandwich board 16:03:40 hi 16:03:42 any bugs we need to land ? 16:03:57 hemna: Nothing I see out there right now. 16:03:58 https://review.openstack.org/#/q/project:openstack/cinder+branch:stable/pike+status:open 16:04:09 So if you have something, speak now or forever hold your peace. 16:04:26 Or wait until the first stable release. :) 16:04:40 that's at the ptg iirc 16:04:57 Speaking of which... 16:05:15 Normal reminders, register for the PTG: 16:05:17 #link https://etherpad.openstack.org/p/cinder-ptg-queens Planning etherpad for PTG 16:05:22 Oops 16:05:37 ACtually, first add topics and your name to the PTG planning etherpad. ^^ 16:05:45 #link https://www.eventbrite.com/e/project-teams-gathering-denver-2017-tickets-33219389087 PTG registration 16:05:50 Then register ^ 16:05:51 hi 16:06:30 The topic etherpad is starting to get filled up, which is great. 16:06:30 Going to do team photos at the PTG again so sign up :) 16:06:41 diablo_rojo_phon: Is there a sign up for that yet? 16:06:48 diablo_rojo_phon: good to know. 16:07:06 smginnis yes, I will send it out again but I put it in the 4 wk out update 16:07:31 I have to send it to the new PTLS but I was hoping the old would communicate to the new as well ;) 16:07:42 diablo_rojo_phon: OK, great. 16:07:48 @! 16:07:48 <_pewp_> jungleboyj (。Ő▽Ő。)ノ゙ 16:07:53 I'll let jungleboyj worry about then. :) 16:07:58 What? 16:08:04 diablo_rojo_phon: I haven't heard :P 16:08:25 Also, those that need ptgbot power need to poke me in the ptg channel so I can hook them up. 16:08:52 :) 16:08:53 prometheanfire: evrardjp I'll send something out to the new PTLS today or tomorrow. 16:09:02 k 16:09:17 jungleboyj: No "I have the power!" from giphy? :) 16:09:43 smcginnis: Too many things going on to play. ;-) 16:09:50 haha 16:09:53 Off your game today. 16:10:05 smcginnis: :-p 16:10:44 On that note, next week is final GA for Pike. jungleboyj and I will figure out who's running the meeting next week, but I suppose that's a good time for hand off. 16:11:10 * jgriffith hands jungleboyj a beer 16:11:19 hand off complete... oh wait 16:11:24 ;) 16:11:30 #topic Video interviews at the PTG 16:11:40 * smcginnis starts slacking off 16:11:45 * jungleboyj chugs the beer. 16:12:06 * jungleboyj gets the feeling that smcginnis is all too happy to put me in charge. ;-) 16:12:12 #link http://lists.openstack.org/pipermail/openstack-dev/2017-August/121369.html 16:12:24 So, I think this is where I take over. How is that for timing? 16:12:31 ;) 16:12:55 So, there was an e-mail to the mailing list from Rich Bowen, RDO Community Liaison asking for volunteers to do video interviews. 16:13:12 Looking for updates on what happened in Pike and to highlight any cross project work that happened. 16:13:30 I think Sean and I could do a project update. 16:14:10 From a cross project standpoint wondering if jgriffith and ildikov would be willing to talk about the work with Cinder/Nova APIs ? 16:14:32 jungleboyj: like we would have any other choice :) 16:14:41 :) 16:14:59 ildikov: Well, you do. :-) 16:15:09 jungleboyj: I think Nova folks could use an update too 16:15:20 Just letting you know the opportunity is there. 16:15:32 jungleboyj: I mean besides the 1.5 people who's reviewing 16:15:44 mriedem loves giving updates. 16:15:49 We had also put on the agenda to do an overview of what has happened during Pike during the PTG. 16:15:51 jungleboyj: sure thing, thank you 16:15:59 smcginnis: :) 16:16:15 jgriffith: Has suddenly gotten quiet. 16:16:37 I think they just wanted 1-3 people for those updates. I nominate jungleboyj and ildikov. 16:16:38 :-) 16:16:50 smcginnis: :-p 16:17:02 jgriffith and I will sit back and have a beer and make faces behind the camera. 16:17:11 +1 16:17:20 * jungleboyj laughs 16:17:25 What have I gotten myself into. 16:17:25 smcginnis: jgriffith has to be there, I can't be blamed for the whole thing :) 16:17:40 I wanted to make faces at ildikov and jgriffith 16:17:46 jungleboyj ohhh..... you have no idea!! moua-moua-hahah 16:17:51 ildikov: Well, if he's not there then it makes it easy to blame all the problems on him. :) 16:18:03 jungleboyj: be the boss and do that :) 16:18:09 smcginnis nice try, I perfected that trick remember :) 16:18:09 smcginnis: I can only imagine how release "collaboration" interview session will go then :p 16:18:21 smcginnis: you know how people associate things to faces... 16:18:22 *smh* 16:18:41 Ok, I guess I am going to be the boss and say that Sean and I will do a general update. 16:19:20 ildikov: Lets talk offline and figure out the right people to do an update on Nova/Cinder and what all should be done at the PTG with regards to that and make that happen. 16:19:40 I also need to coordinate with mriedem on the when we are going to do a joint Nova/Cinder session. 16:19:55 jungleboyj: Reminds me, should also coordinate with mriedem on a joint cinder/nova timeslot. 16:20:00 jungleboyj: sure, I will repeat multi-attach every 10 seconds and someone else will do the update and we're done, awesomeness :) 16:20:21 Anyone else have anything related to Cinder they think we would want to share? 16:20:25 smcginnis: Jinx. 16:20:30 https://review.openstack.org/#/c/496284/ 16:20:30 You lagging up there? 16:20:44 Yeah, I think so. 16:20:50 prometheanfire: Hold on. 16:21:02 jungleboyj: Next topic? 16:21:14 Yeah, I think we are done with that one. 16:21:23 At least have a plan. 16:21:25 #topic Oslo cleaning up deprecated stuff 16:21:32 And it is me again. 16:21:52 So, Oslo is going to be working during queens to clean up things that they have deprecated. 16:22:03 #link http://lists.openstack.org/pipermail/openstack-dev/2017-August/121345.html 16:22:18 I haven't had a chance to look and see what impact that has on Cinder. 16:22:40 Sharing here for awareness. If there is anyone bored and looking to get commits I would appreciate help looking at that. 16:22:55 If not, I will try to find some time to work on that during the release. 16:23:03 jungleboyj: I've started looking at a few deprecation warning. I can try to spend a little time on it. 16:23:16 that's a pretty long list 16:23:26 jgriffith: Yeah, that part makes me a little nervous. 16:23:31 smcginnis: I can help you with it 16:23:40 e0ne: ++ 16:23:40 * e0ne likes to remove code 16:23:43 e0ne: smcginnis Awesome, thank you. 16:23:49 grep to the rescue 16:24:13 e0ne: Me too. :) 16:24:32 jungleboyj: Move on? 16:24:32 :) 16:24:41 smcginnis, plus me ;) 16:24:52 Yep, I think so. Appreciate help we can get there. 16:25:04 Let's see how many lines of code we can remove in Queens. :) 16:25:15 replication..... 16:25:15 smcginnis: Removal only Fridays? 16:25:17 w00t! 16:25:19 all of them!!! 16:25:27 jgriffith: Hah! :) 16:25:45 #topic cinder-manage output 16:25:55 prometheanfire: Not on the agenda, but I know you wanted to discuss this. 16:25:59 replication should die in a fire. 16:26:01 #link https://review.openstack.org/#/c/496284/ 16:26:02 a whole topic! 16:26:11 we are lucky out there 16:26:16 :) 16:26:44 looks like the review are not overwhelmingly negative, so I guess if we fix we are good to go... 16:26:50 hemna: Comes in from left field. 16:26:55 :D 16:27:25 ya, I was more curious how far back that can be backported, as a user facing change I'm not expecting much 16:27:25 so I guess it's a thank you ... and a question 16:27:40 prometheanfire: what a SYNC! 16:27:45 prometheanfire, evrardjp: So the tl;dr is the output is currently truncated making it hard to match up output in scripting> 16:27:46 evrardjp: :P 16:28:02 impossible in some cases 16:28:03 yup. 16:28:40 prometheanfire: I don't know that this is really a bug so I don't see it having backport potential. 16:28:45 Hmm, that does change the result returned. I wonder if that would be considered an API change? 16:28:52 i think it is a bug 16:29:03 * smcginnis would rather not microversion it, but needs to bring it up at least 16:29:03 smcginnis: that's kinda what I was wondering 16:29:21 Oy. 16:29:23 microversion? this is cinder-manage 16:29:32 eharney: Yeah. 16:29:37 or is it not? 16:29:38 Oh, it's a cmd. Yeah, nevermind. 16:29:51 Could someone possible be using the output as it is now, in a script? That's the only important question... 16:30:06 that's not a specific enough question 16:30:19 I doubt it 16:30:27 Could we break someone by changing this? 16:30:28 I don't think the output is useful as is 16:30:32 More specific - is fixing this for this case going to screw anyone else? :) 16:30:33 I agree. 16:30:36 it has the same output after the fix, it's just not truncating the hostname, right? 16:30:45 eharney yes 16:30:47 someone could be logging status via it though 16:30:56 so that'd change values 16:31:01 I'm not sure about backporting output changes to cinder-manage though 16:31:10 smcginnis: THat is my concern. 16:31:13 besides, they could just use "cinder service-list" 16:31:25 that call in cinder-manage should probably be deprecated IMO 16:31:28 ya, but cinder-manage is more direct, should be faster 16:31:38 prometheanfire ? 16:31:48 cinder-manage is also useful if running offline right? 16:31:55 jgriffith: inconsistency between cinder service-list and cinder-manage is a concern 16:32:00 that's where we ran into it 16:32:13 prometheanfire can you define "running offline" 16:32:16 I don't know what that means 16:32:20 but I think the use cases that lead to noticing it are very much low. 16:32:20 cinder-api is stopped 16:32:21 sorry :( 16:33:09 I suppose it can be considered a bug, just not really a bug directly affecting Cinder, but users of cinder-manage. 16:33:14 I meant there is not many use cases. In our case, for example, it's very specific. And we can workaround it by checking in the DB. But that's ugly. We were just looking into a more user friendly way forward 16:33:18 eharney: Any distro related concerns about backporting it? 16:33:37 not from gentoo 16:33:43 evrardjp: I'd rather not have an answer that involves going directly to the db. :/ 16:33:44 haha prometheanfire ;) 16:33:48 prometheanfire: :) 16:33:53 :P 16:33:56 smcginnis: agreed with you there. 16:34:03 is the "vs" example in the commit message correct? 16:34:20 smcginnis: but I also understand the risks of touching stable code, so it's all about tradeoffs and risks. 16:34:26 evrardjp: should be able to confirm 16:34:31 in other words: no my call! 16:34:34 i.e. is it supposed to be including the "@lvm" ? 16:34:54 no it's FQDN@something vs hostname@something 16:35:06 IIRC 16:35:36 I just ran both cinder service-list and cinder-manage service list and both output foo@bar for c-vol 16:36:00 that's added for multi-backend 16:36:02 This only happens if it is set up with a FQDN, right? 16:36:08 yes. 16:36:10 smcginnis yes 16:36:39 I'm not inclined to support a backport of this FWIW just in order to offer up ability to query and offline service 16:36:43 my vs from ocata here 16:36:44 | cinder-volume | storage.mthode.org@lvmdriver-1 | nova | enabled | up | 2017-08-23T16:36:26.345706 | - | 16:36:50 cinder-volume storage nova enabled :-) 2017-08-23 16:36:36.351278 3.10 1.21 16:36:55 I'm also not inclined to continue to have both versions of the same cmd FWIW 16:37:17 prometheanfire that what I get as well on current master 16:37:19 well they don't operate the same way IIRC 16:37:27 evrardjp nope, they don't 16:37:28 I think whether we get rid of cinder-manage is another discussion we should probably have, but while we do have it, we should make sure the output matches. 16:37:33 but I understand the position, and I have fine with that. 16:37:41 thanks for the consideration! 16:37:43 smcginnis okie dokie 16:37:46 But I'll leave that up to our Queen PTL. Er, I mean our Queens PTL. :{ 16:37:49 :P 16:37:53 haha 16:37:54 jgriffith: I think having it in both is good as cinder-manage does not need cinder-api to be running 16:38:04 *SMH* again 16:38:05 as an operator that's useful 16:38:27 prometheanfire Yeah, I get it; but I wish operators would speak up a bit more before we introduce madness :) 16:38:30 prometheanfire: definitely more convenient than checking in the DB ;) 16:38:33 I am not in a hurry to remove cinder-manage. 16:38:43 :D 16:38:52 I do agree that what we get out of cinder-manage and cinder service-list should be consistent. 16:38:58 jungleboyj we can table that. but to be clear I'm not recommending complete removal of cinder-manage 16:39:16 k, so anything else to discuss about this? 16:39:33 We should maybe discuss cinder-manage at the PTG. 16:39:34 for this bug, probably not :) 16:39:41 jgriffith: Just removing the ability to show the running services? 16:39:43 * DuncanT wishes cinder service list (and cinder manage) listed API services... 16:39:44 Let's leave any opinions on the review. 16:39:47 if you have multiple cinder nodes with the same name but on different domains you're kinda asking for trouble IMO but whatever 16:39:50 smcginnis: ++ 16:40:01 smcginnis: agreed 16:40:17 DuncanT: There's something you can work on. :] 16:40:32 Thanks evrardjp and prometheanfire 16:40:36 it does 16:40:37 sigh 16:40:38 * jgriffith shuts up 16:40:43 thanks for having us! 16:40:53 jgriffith: I thought so. 16:40:59 #topic Open discussion 16:41:17 I did want to bring up one thing I learned about reno usage. 16:41:47 It might be changed, but as it is right now reno will scan through and pick up any changes to release notes between tags and show it based on that. 16:42:10 So if an Ocata release note is updated to fix a type or URL in pike, it will end up showing up in the pike release notes. 16:42:36 So proper handling for that situation is to edit the release note directly in the stable/ocata branch. 16:42:45 has dhellmann been told (or a bug been made)? 16:43:17 prometheanfire: He's thinking of changing the way reno handles that, but for now I have a couple patches for the reno usage docs and the project team guide. 16:43:42 Apparently it's a little tricky to handle in reno, so that may take a little time. 16:44:20 perhaps a version override header or something 16:45:07 There is an ignore-note directive that can be added to the release rst file to explicitly tell it to not display specific release notes in the output. 16:45:26 That's the workaround for the cases where the notes have already been edited. 16:45:43 But just wanted to raise it here so cores are aware of it in reviewing patches. 16:46:02 Try not to let release notes updates through on master if it is for a stable release. 16:46:06 smcginnis: Thanks for pointing that out. That is good to know. 16:46:22 It was news to me, so I figured it would be good to get out there. 16:46:40 smcginnis, any patch to see about this? 16:47:25 Reno one was approved: https://review.openstack.org/496391 16:47:34 thanks 16:47:40 And update to project team guide is pending: https://review.openstack.org/496318 16:48:03 smcginnis: There was a request to add notes to the documentation page about Reno. I can include that info as well. :-) 16:48:37 jungleboyj: Perfect! Probably good to have some basic info and reference over to the reno guide in case that changes at all. 16:48:46 Anything else? 16:48:51 smcginnis: ++ 16:49:47 Oh, we should probably also get ocata patches through before that transitions stable phases. 16:49:50 https://review.openstack.org/#/q/project:openstack/cinder+branch:stable/ocata+status:open 16:50:14 smcginnis: Good note. I think I looked a week ago or so and there wasn't much sitting around. 16:50:23 Good reminder to people to get anything proposed. 16:50:46 ++ 16:51:00 OK, I think that's all I've got. Anything else? 16:51:10 Bueller? 16:51:35 Cool, thanks everyone. 16:51:48 thanks! 16:51:52 thanks 16:51:53 o/ 16:51:56 Thanks! 16:51:58 @! 16:51:58 <_pewp_> jungleboyj ( ・ω・)ノ 16:52:01 thanks! 16:52:14 #endmeeting