15:01:24 #startmeeting manila 15:01:24 Meeting started Thu Mar 10 15:01:24 2016 UTC and is due to finish in 60 minutes. The chair is bswartz. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:25 hola 15:01:26 hello o/ 15:01:26 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:01:28 The meeting name has been set to 'manila' 15:01:28 hi 15:01:32 hi 15:01:33 hello 15:01:34 hi 15:01:34 Hi 15:01:40 hi 15:01:49 Hi 15:01:49 hi 15:01:52 hey 15:02:09 \o 15:02:11 Nobody added anything to the agenda today 15:02:18 so I'm just going to talk about the RC 15:02:19 and bugs 15:02:27 hi 15:02:34 agenda + notifications please 15:02:34 bswartz: I had something in the agenda 15:02:48 dustins: oh, that was under the March 3rd meeting -- I got confused 15:03:11 Oops, I just did the "Next Meeting" thing and updated it 15:03:21 okay dustins I'll put your topic back in and we'll hit it in one second 15:03:30 w00t 15:03:40 #topic announcements 15:03:45 #link #link https://launchpad.net/manila/+milestone/mitaka-rc1 15:03:54 err 15:03:59 #link https://launchpad.net/manila/+milestone/mitaka-rc1 15:04:21 we have a huge number of bugs targeted to RC1 15:04:46 and there will probably be more coming, as people continue to test and file bugs 15:05:01 reviewers should be using this list to prioritize their reviews 15:05:10 keystone v3 support is borken btw :( 15:05:17 s/borken/broken/ 15:05:23 mkoderer: indeed 15:05:30 mkoderer: I tagged that bug critical 15:05:52 anyways our target date for RC1 is exactly 1 week from today 15:05:54 bswartz: I have a fix running in my env but it need much more polishing 15:06:16 and honestly I'll be shocked if every one of these targeted bugs gets fixed 15:06:17 mkoderer: fix for generic driver? 15:06:29 mkoderer: and usage of nova-, cinder- and neutronclients? 15:07:02 so let's get the important ones fixed and merged ASAP and do our best to get the rest fixed 15:07:25 vponomaryov: so baiscally I started with neutron.. and know I wrote a general solution 15:07:27 starting next week if you own a bug and there is no fix posted I'm going to start pinging you and asking if we can retarget the bug to newton 15:07:59 RC1 doesn't get tagged until we have ZERO open bugs 15:08:30 I would like to add a keystone v3 gate job after the fix is done 15:08:52 bswartz: +1 15:08:54 mkoderer: why add? switch existing one fo rgeneric driver 15:09:01 no more gate jobs 15:09:06 just update the current ones 15:09:12 bswartz: ok 15:09:31 unless there's a good argument to continue testing on keystone v2 15:09:43 then we need to ensure that at least one voting job also covers old keystone API 15:09:50 vponomaryov: basically disabling v2 for at least on job 15:09:57 s/on/one/ 15:10:32 so I have 2 more things to say about bugs 15:10:56 if you file a bug and you think it needs to be fixed in Mitaka -- make sure you find a bug supervisor and get it targeted! 15:11:48 for reviewers: don't even look at fixes for bugs that aren't targeted to RC1 15:11:49 #link https://launchpad.net/~manila-bug-supervisors 15:11:57 we have too much to review without looking at untargeted stuff 15:11:58 https://launchpad.net/~manila-bug-supervisors 15:13:16 okay that's all I have to say about bugs 15:13:34 btw we had bug smash event this week 15:13:36 https://etherpad.openstack.org/p/OpenStack-Bug-Smash-Mitaka 15:13:42 #agenda https://wiki.openstack.org/wiki/Manila/Meetings 15:13:54 manila was one of the hot topics at least in the german team 15:14:08 mkoderer: that's awesome 15:14:30 #topic Common Capabilities Matrix 15:14:40 okay dustins it's your turn 15:15:00 Thanks, bswartz! 15:15:25 #link https://review.openstack.org/#/c/288504/ 15:15:30 #link http://docs-draft.openstack.org/04/288504/6/check/gate-manila-docs/c696e6f//doc/build/html/devref/share_back_ends_feature_support_mapping.html#mapping-of-share-drivers-and-common-capabilities 15:16:06 I'm adding a common capabilities table to the developer reference for things like QoS, Thin Provisioning, DHSS, etc 15:16:30 I'd like to have all of the drivers' capabilities in there before it merges 15:16:35 cknight: woah. great find. one table's not displayed. 15:16:55 dustins: in case of ZFSonLinux driver, "compression" can be set in config option for ZFS dataset, so answer is yes and no 15:17:02 I've got a good chunk of them in there, but I'd like to have all of them 15:17:12 vponomaryov: then it's supported 15:17:30 bswartz: but not by manila/scheduler 15:17:36 dustins, can you enumerate the ones you need help with? 15:18:07 vponomaryov: we should only have those that reported as capabilities currently.. 15:18:11 vponomaryov: if the driver doesn't use the common capability, then it's not supported -- but it sounds like it could be supported 15:18:19 I recommend a feature in Newton to add it 15:18:27 bswartz: so, we can consider it a bug 15:18:37 no that's a feature 15:18:45 bswartz: support exists but not reported as capability 15:19:04 even if we pretend it's a bug, it's a low priority one 15:19:12 bswartz: a bugfix i think; we had this issue with the hnas driver.. tpsilva fixed it.. 15:19:25 that's not the kind of thing we should waste time on when we have very broken things 15:19:25 Looks like I need ZFSonLinux, Generic, EMC VNX, EMC Isilon, GlusterFS, GlusterFS Native, HDFS, LVM, LXD, Windows SMB, IBM GPFS, and Oracle ZFSSA 15:19:42 Those are the ones 15:19:57 dustins Looks pretty small list ;) 15:20:02 cknight: Strange, when I look at the docs-draft, I see the table I added 15:20:04 dustins: you just listed like 75% of the drivers 15:20:13 )) 15:20:17 dustins: try removing the note. 15:20:36 dustins: I didn't believe it either, but I looked in 2 different browsers. 15:20:52 Oh whoa, the one I added is there, but the one before is gone 15:21:00 Weird... 15:21:12 * dustins really needs to stop committing when on airplanes 15:21:23 dustins: the ascii table borders don't match up 15:21:47 dustins: yeah, stop doing that. we've a lot of things in-flight already. :P 15:22:04 lol 15:22:07 okay I think this discussion belongs in the code review not the meeting 15:22:09 Yup, since I moved some things around 15:22:26 so docs fixes are also fair game to review and merge 15:22:30 Yeah, I'll fix that once this is done and have one that's decidedly unbroken :) 15:22:49 dustins: +1 -> thanks for making the list, it's helpful 15:23:10 Happy to help! 15:23:20 dustins: use the ML for wider attention if you need responses and aren't getting them 15:23:30 bswartz: Wilco 15:23:39 some of our driver maintainers don't come to these meetings at all 15:23:57 bswartz: who do we have for HDFS? 15:24:20 bswartz dustins: on that topic, could we please request them to update the driverlog? 15:24:34 the driver is unmaintained at the moment, but rraja has been kind enough to fix the CI system 15:24:47 gouthamr: Driverlog? 15:25:00 gouthamr: yes 15:25:21 dustins: https://github.com/openstack/driverlog/blob/689c897a2c35d72ae0b6748f3ef11be2b87ad824/etc/default_data.json 15:25:21 dustins: that all? 15:25:39 bswartz: Yup, that's all I got for this topic 15:25:45 Thanks 15:25:45 #topic open discussion 15:25:54 alright anything else for today? 15:26:02 I wanted to point out that it seems manila doesnt really support notifications 15:26:07 there is basically one place 15:26:16 rpc.get_notifier('shareType').info( 15:26:17 context, 'share_type.create', notifier_info) 15:26:24 but other than that not really 15:26:33 ameade: multiple places.. 15:26:42 hi, we're looking implement a new driver in the Newton release. What are the deadlines to submit a new driver? 15:26:45 for .error there are more 15:27:06 If anyone's interested, the RDO Test Days are today and tomorrow 15:27:16 jseiler: for Mitaka we did Feature Freeze date minus 3 weeks 15:27:16 #link https://www.rdoproject.org/testday/mitaka/milestone3/ 15:27:43 I have some good news from HPB support :) 15:27:52 jseiler: we haven't discussed the deadline for Newton yet 15:28:15 I had a deep dive session with rkukura and we have a plan to support HPB without any API changes :) 15:28:29 I will give an update on the ML 15:28:37 jseiler: but unless we make a change, it will be the same for Newton 15:28:43 mkoderer: nice! 15:28:59 jseiler: however earlier is always better! 15:29:16 mkoderer: without any API changes in which project? 15:29:26 bswartz: not in Neutron and not in Manila :) 15:29:28 were you looking at changes the neutron API? or the manila API? or both? 15:29:32 bswartz: what have you done for past releases? I'm familiar with cinder, they require new drivers for the 1 milestion (i.e., m1). Have you done anything similar for manila drivers? 15:29:59 jseiler: did you see what I wrote? 15:30:03 jseiler: we merged drivers in m3 15:30:04 bswartz: there were some thought have a api change in manila to notify with the used segementation_id 15:30:32 * bswartz wonders if his IRC is on the fritz again 15:30:32 bswartz: but all good.. I will modify my prototype and let you know 15:30:45 bswartz: yes. I was trying to get a feel for what was done in the past. thank you. 15:31:08 vponomaryov: thank you 15:31:25 jseiler: by the deadline we expect a complete driver with high unit tests coverage and working (and passing) CI 15:31:48 we weren't explicit about the date for the CI requirement in Mitaka so we can fix that for Newton 15:32:03 the CI should be passing by the driver deadline date IMO 15:32:10 bswartz: thank you. so it sounds similar to the requirements for a cinder driver. 15:32:20 jseiler: "complete" means required set of features is implemented, see #link http://docs.openstack.org/developer/manila/devref/driver_requirements.html 15:32:42 vponomaryov: thank your for the link 15:33:06 mkoderer: I'd prefer if you worked on bugs for the next week and not the HPB stuff 15:33:30 bswartz: and my boss kicks me if I don't work on HPB :P 15:33:41 I think this is the highest number of open bugs we've ever had against an RC 15:33:49 bswartz: but keystone v3 has higher prio than HPB 15:33:49 I'm still looking at open bugs and targeting some of them, so the number could go higher 15:34:26 mkoderer: so, you are going to upload v3-fixes, right? 15:34:46 vponomaryov: yeah I hope I have something working till tomorrow 15:35:01 thanks mkoderer 15:35:08 mkoderer: ok 15:35:27 alright I think we're done for today 15:35:42 bswartz: and btw I have two other team member that are focusing on manila bug fixing (dgonzalez and tpatig) 15:35:51 thanks everyone and happy bug hunting/fixing! 15:36:10 cya 15:36:11 #endmeeting