14:00:55 #startmeeting sahara 14:00:56 Meeting started Thu Aug 27 14:00:55 2015 UTC and is due to finish in 60 minutes. The chair is SergeyLukjanov. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:57 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:59 The meeting name has been set to 'sahara' 14:01:08 o/ 14:01:17 #link https://wiki.openstack.org/wiki/Meetings/SaharaAgenda 14:01:27 ^_^ 14:01:28 let's wait for a few mins for more folks 14:01:45 i added couple last minute items to the agenda 14:01:48 \o 14:01:53 hi 14:02:08 o/ 14:03:07 o/ 14:03:12 hello/ 14:03:29 #topic sahara@horizon status (crobertsrh, NikitaKonovalov) 14:03:35 #link https://etherpad.openstack.org/p/sahara-reviews-in-horizon 14:03:38 folks, please 14:03:50 Looks like another patch or two are on the path to merging, which is good. 14:04:10 not much updates from me, the event-log patch may land soon I think 14:04:11 folks, please review event-log change 14:04:17 I've had a few people mentioning "very strange" things happening in their horizon related for form submission. I plan to take a look at that today. 14:04:55 crobertsrh: Any specific flows to watch out for to confirm/deny issues? 14:05:22 I think that elmiko reported an issue when running a spark job. 14:05:34 yea 14:05:39 Job launching was failing at form submission time. 14:06:05 crobertsrh, elmiko looks strange 14:06:15 probably client issues? 14:06:27 I had that issue where java jobs were failling, because the filtering in horizon for empty input/output ids was failing 14:06:28 Not sure. I have not yet managed to reproduce it. 14:06:36 i was getting errors submitting the form, it wasn't even making it to the api endpoint 14:06:44 It was checking for None or "", but I was getting "None" (as a string) 14:06:51 this was F21 14:06:53 i have now backed off to a stable/kilo devstack and am much happier =) 14:07:06 tmckay: i got some of those too, on f21 14:07:06 I hacked it by adding a check for "None" along with [None, ""] 14:08:00 vgridnev: I think you've told me about this kind of issue 14:08:32 NikitaKonovalov, there were same issues, but I was unable to reproduce that on master 14:09:05 not sure 14:09:27 tmckay, this is not called hacked in horizon, it's almost a normal flow (adding a contstat to a list of existing workarounds) 14:10:43 NikitaKonovalov, I meant in the sense that I modified my horizon and restarted in devstack :) 14:11:09 tmckay: you should propose the change I think 14:11:19 probably 14:11:46 anything else to discuss in this topic? 14:12:01 nothing from me 14:12:39 #topic News / updates 14:13:10 It looks like several Sahara-related talks are on tap for Summit. 14:13:30 \o/ 14:14:19 Remember to check out https://etherpad.openstack.org/p/sahara-liberty-bug-fix-day if you're looking for action items: we have a few high-prio bugs that haven't been beaten down still (though a good number are in late progress; thanks everyone!) 14:14:29 i have one more patch(docs) for the keystone sessions, then they will be finished. also i'm preparing the castellan changes and will probably need a FFE for them. 14:15:39 I'm working on bug with using Swift via SSL endpoint 14:16:27 Also i've updated patches with new HDP plugin 14:16:43 I'll be on PTO next two weeks (Aug 31 - Sept 11), I'll handle 0.11.0 release and liberty-3 as well, but most probably will not join next two IRC meetings, the best way to communicate with me - email slukjanov@mirantis.com 14:16:48 I'm investigating the usage of Tempest External Plugin Interface for the CLI clients scenario tests in sahara (which currently requires you to copy tempest code...) 14:16:54 Helping out with various bugs; in very late stages of TripleO/Sahara integration. It's installing; last bits are MQ and balancer integration. 14:17:03 not sure I can make it for the freeze, I think I will ask for an exception 14:17:17 SergeyLukjanov: how do you feel about FFE for the improved secret storage? 14:18:08 (if an FFE makes sense for this, it's more a bugfix, but let's see) 14:18:12 tosky, I think it's not for the freeze - it's not so needed in release to have FFE, but probably we don't need FFE for it at all due to the fact that it's not in the code itself 14:18:34 elmiko, if it'll be on review before FF - O think so 14:19:13 SergeyLukjanov: not sure about that, i cut put it up but i'm still waiting on castellan to hit pypi. is it worth posting it even if all the tests will fail? 14:20:27 i hate to have it slip another cycle, but the timing just might not work out 14:22:30 is Andrey Pavlov here? 14:22:45 huichun, he is on PTO till the end of next week 14:22:56 so, we've decided to push the updates / ACL patches w/o him :) 14:23:07 ok thx sergey 14:23:38 elmiko, yeah, I understand, let's evaluate FFE for it when L3 happen 14:23:52 SergeyLukjanov: ok, maybe i'll just put it as a draft review 14:23:52 huichun, I updated your patch to client, and put separate change with schema in sahara 14:24:55 vgridnev: ok I will update it after meeting, there is something little to fix ,thx vgridnev 14:24:58 huichun, please rebase your patch with suspend / resume on https://review.openstack.org/#/c/217635/ 14:25:25 vgridnev: ok i will rebase it after meeting 14:26:57 #topic Bug triage / fix days / doc days progress 14:27:09 elmiko, egafford, anything you want to share? 14:27:15 and thx for organising it :) 14:27:17 SergeyLukjanov: High prios are going well. 14:27:26 Some remaining ones: 14:27:35 [HDP] EDP Hive jobs do not work: https://bugs.launchpad.net/sahara/+bug/1279360 14:27:36 Launchpad bug 1279360 in Sahara "[HDP] EDP Hive jobs do not work" [High,Confirmed] 14:27:46 Tempest client tests fails because register image failed: https://bugs.launchpad.net/sahara/+bug/1430252 14:27:47 Launchpad bug 1430252 in Sahara "Tempest client tests fails because register image failed" [High,In progress] - Assigned to Jacob Bin Wang (dongliang0531) 14:27:52 elmiko: scheduler edp patch has been updated,need your review 14:28:00 [HBASE] Service has only 0 RegionServer, when HBASE_REGIONSERVER on every node: https://bugs.launchpad.net/sahara/+bug/1467922 14:28:01 Launchpad bug 1467922 in Sahara "[HBASE] Service has only 0 RegionServer, when HBASE_REGIONSERVER on every node" [High,Triaged] 14:28:06 huichun: ack 14:28:20 Create big cluster failed because token expired: https://bugs.launchpad.net/sahara/+bug/1486653 14:28:21 Launchpad bug 1486653 in Sahara "Create big cluster failed because token expired" [High,In progress] - Assigned to Sergey Reshetnyak (sreshetniak) 14:28:59 There's also a high-prio bug about having multiple concurrent periodic task drivers with multiple engines, but methinks that's a larger architectural issue than bugfix week is likely to tackle. 14:29:19 yea, that's a biggie 14:29:39 as far as docs go, i put the etherpad up, we are set for next week. 14:29:43 If we have time to hit any of these, that'd be excellent. 14:29:52 #link https://etherpad.openstack.org/p/sahara-liberty-doc-update-day 14:30:28 i will be at the security mid-cycle next week, but i will keep up on the doc reviews. so, everyone grab a doc or two next week! 14:31:48 #topic Client release plans 14:31:53 elmiko: If there is known missing documentation, how should we record/sign up to fix those issues? 14:32:03 Ha, take that new topic. 14:32:44 egafford: are we talking about missing from a current doc, or new doc to be created? 14:32:50 do we have something else on previos topic? we could return to it 14:33:00 #undo 14:33:01 Removing item from minutes: 14:33:05 just egafford had a question about new docs 14:33:21 Either/both. Likely missing from a current doc, on average. 14:33:31 actually, the idea fo the doc day that we have each cycle I think is to document missing parts and update existing 14:33:49 missing from a current doc, just follow the proceedure and take the doc and add your content in a review 14:33:55 because things are changing and we'd like to keep docs up to date for releases 14:33:59 if its a new doc, add the doc to the list and make a review 14:34:15 SergeyLukjanov: +1 14:34:32 SergeyLukjanov: Yup, of course. Okay, cool. So if we know there are missing bits, might make sense to sign up to review those docs, otherwise same process. 14:34:38 egafford: i will add a note about this to the pad 14:36:36 i think we're set on this topic SergeyLukjanov 14:36:41 :) 14:36:54 #topic Client release plans 14:37:11 so, I think you've already read my email in os-dev 14:37:22 \o/ 14:37:37 +1 to \o/ 14:38:20 #note python-saharaclient release 0.11.0 planned for 8/31-9/1 14:38:24 #info python-saharaclient release 0.11.0 planned for 8/31-9/1 14:38:41 so, let's keep pushing and merge all features support till this dates 14:39:47 any questions or issue so far 14:39:48 ? 14:40:25 nop 14:40:28 none from me 14:40:33 the Liberty-3 will be somewhere on the week, most probably Wed - Thu 14:40:54 (/me planning to push the client release from the beach with some good whiskey :) ) 14:42:21 hi sergey do I have time to submit a spec about improving EDP job log details? currently when job failed, we just let user know failed,but not why, I want to show more error info, so he can know where is detail error in his jar, like class not found error 14:42:25 SergeyLukjanov: 14:42:42 huichun, nope, the FF is next week 14:42:48 ok 14:42:55 sadly, kinda late for a new spec =( 14:43:01 #topic Liberty-3 14:43:03 huichun, but it can be submitted for M cycle anway 14:43:18 it's a good idea, I have thought about it too 14:43:21 Yeah, definitely a good M cycle item 14:43:23 huichun, yeah, really sorry about that but we're already going into FF 14:43:25 tmckay: ok will submit at that cycle 14:43:33 We briefly talked about it in Vancouver. 14:43:37 ^_^ 14:43:45 huichun, it can be submitted now, it will just get bumped 14:43:46 sounds like a good spec though, we need better error logs 14:43:58 tmckay: ok 14:44:27 definitely, better error logs++ 14:44:28 I think the FFEs could be good discussed on the next IRC meeting 14:44:33 (and errors--) 14:44:34 * SergeyLukjanov adding item to the next meeting 14:44:47 #topic Ceilometer at the gate 14:44:53 #link http://lists.openstack.org/pipermail/openstack-dev/2015-August/072958.html 14:45:01 i added this 14:45:13 I've assigned bug to myself and have patch prepared to fix this (saw this email yesterday) 14:45:16 cdent reached out to us, to let us know about the changes to ceilometer 14:45:22 #link https://bugs.launchpad.net/congress/+bug/1489436 14:45:23 Launchpad bug 1489436 in Sahara "ceilometer becoming a devstack plugin will impact gate job config" [Low,Triaged] - Assigned to Sergey Lukjanov (slukjanov) 14:45:32 ah, excellent. SergeyLukjanov++ 14:45:53 it's very simple, just returning back ceilo when it'll be extracted from devstack 14:45:59 yea 14:46:08 just wanted to make sure we were on top of it 14:47:50 cool 14:48:00 any news re Sahara / Zaqar? 14:48:16 i've been doing some research about zaqar 14:48:19 * flaper87 here if you guys need help or have questions 14:48:30 for the guest client, but also for a possible experiment 14:48:37 flaper87: yea, we still need to talk sometime! 14:48:41 #topic Sahara / Zaqar integration for guest agents implementation 14:49:16 not much to report yet, but i am starting to play with the zaqar api, mainly learning the ropes so far 14:49:32 elmiko, flaper87, I think we should find some time on design summit for the Zaqar-based unified agent talk 14:49:40 i also want to try something with dumping log data to a queue, but this is very experimental 14:49:47 SergeyLukjanov: +1 14:49:55 SergeyLukjanov: sounds good, what if we talk a bit earlier and we do a hacking session ? 14:50:27 we're also going to talk about zaqar a little at the security mid-cycle 14:50:31 yeah, sure, hopefully, we should have a proposal at least for summit :) 14:50:44 flaper87: i'm down with that 14:50:47 flaper87: That'd almost certainly make discussion time more valuable, yeah. 14:51:18 * flaper87 writes all that down 14:51:52 that's all i had 14:53:13 Future of Sahara code inside the Horizon 14:53:16 #topic Future of Sahara code inside the Horizon 14:53:28 do we have something to chat about here today? 14:53:52 Not sure anything has really changed. 14:53:55 I think we can leave this for the summit 14:54:09 Yes, at summit, we can agree to go back to our own repo :) 14:54:16 lol 14:54:26 NikitaKonovalov, and crobertsrh can you give me a hand to get some reviews in the storm ui patch? 14:54:40 Absolutely. 14:54:43 it would be great to have that done so the storm talk can be better presented 14:54:47 tellesnobrega: we'll try 14:54:54 #topic Open discussion 14:55:01 thanks 14:55:10 tellesnobrega: even if you have to use experimental code for the talk, i think its ok 14:55:12 tellesnobrega: I've already givenit a +1 :) 14:55:26 NikitaKonovalov: we need to talk sometime after next week about the sparkhara talk too 14:55:36 elmiko, sure, but it sounds better to say it is already there 14:55:37 i'd like to go over our plans and figure out how we can integrate 14:55:40 List of topics for summit? 14:55:41 tellesnobrega: +1 14:55:52 elmiko: yes we can talk in #sahara channel 14:55:59 NikitaKonovalov: definitely 14:56:39 crobertsrh, i saw the +1, maybe you can ping some horizon people to take a look, you have more influence there 14:57:04 Sure, I'll give my magic wand a try :) 14:57:11 thanks :) 14:57:38 2 mins left 14:58:21 have fun on vacation SergeyLukjanov =) 14:58:41 thx :) 14:59:00 here is an etherpad for FFEs proposals http://etherpad.openstack.org/p/sahara-liberty-ffes 14:59:38 #endmeeting