16:00:54 #startmeeting fuel 16:00:56 Meeting started Thu Jul 30 16:00:54 2015 UTC and is due to finish in 60 minutes. The chair is xarses. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:56 #chair xarses 16:00:57 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:59 Todays Agenda: 16:01:00 The meeting name has been set to 'fuel' 16:01:00 #link https://etherpad.openstack.org/p/fuel-weekly-meeting-agenda 16:01:01 Current chairs: xarses 16:01:02 Who's here? 16:01:05 hi 16:01:05 o/ 16:01:06 hi 16:01:06 hi 16:01:07 Hello! 16:01:10 рш 16:01:12 hi 16:01:37 o/ 16:01:38 hi 16:01:39 \~/ 16:01:56 hi 16:02:13 #topic Exceptions and status (xarses) 16:02:16 hi 16:02:29 So I wanted to get an update on the FF's that are done 16:02:43 hey - sorry I'm late 16:02:48 Hi guys 16:02:51 hi! 16:03:22 FFE for bug/1475759 ceph generators - was split into bug and feature, feature was moved to 8.0 16:03:36 FF Exception request for Fernet tokens support was declined 16:03:53 sorry 16:03:59 FFE for bug/1475759 ceph generators is a feature, and moved to 8 16:04:06 hi 16:04:16 FF Exception for LP-1464656 fix (update ceph PG calculation algorithm) - was split 16:04:34 FF Exception request for Deploy nova-compute (VCDriver) feature - was split 16:04:52 for the ones that where approved, what is their current merge status? 16:05:00 0/ 16:05:08 FF Exception request for Custom node attributes feature - is merged 16:05:16 FF Exception request for Templates for Networking feature 16:05:25 ^ akasatkin ? 16:05:29 everything is in? 16:05:38 All code is in master. 16:05:38 We've got working demployments with custom networks and templates. 16:05:38 But there are some bugs which need to be fixed. 16:05:38 And we need to land some fixes into master to get all stuff to work properly. 16:06:05 FF Exception request for Env Upgrade feature, was approved. Has everything landed? 16:06:18 So, it is OK, we work on bugs 16:07:07 xarses: not yet 16:07:24 ogelbukh: any blockers? what's your eta? 16:07:26 one patch out of 4 is merged, 1 is close to merge, 2 more are still in dev 16:08:08 no particular blockers, just not enough eng resource 16:08:27 ogelbukh: any specific area where you need help? 16:08:27 expect to be ready to merge all by tomorrow 16:09:10 just prompt responses from reviewers 16:09:23 sounds good 16:09:29 FF Exception request for Deploy nova-compute (VCDriver) feature 16:09:29 and I'm actually quite impressed by it 16:09:51 everything goes pretty quick 16:09:56 was approved in part, has that part landed? 16:11:40 holser bogdando aglarendil alex_didenko one of you could land only ^^ 16:11:57 thanks mihgen 16:12:45 ok, I will ask for a follow up on the ML 16:12:51 mihgen: VCDriver was landed 16:13:00 the rest is still in progress 16:13:17 sgolovatiuk: sorry I'm still looking for holser, you changed your nick 16:13:30 what is "the rest"? 16:13:32 I'll change it back 16:13:33 sorry 16:14:17 any other feature freeze related items to raise / talk about? 16:15:07 ok moving on 16:15:10 #topic Overall bugs situation (mihgen) 16:15:23 guys we are in a challenging situation 16:15:24 #link http://postimg.org/image/z81emkhop/ 16:15:41 this is grap for all bugs in fuel, start date is about when we started the release, end date is HCF 16:15:48 #link http://postimg.org/image/46ca5sglj/ 16:15:54 this one for critical & high only 16:16:32 of course, we just refocused to bugs from features, but still. Let's ensure that we are totally committed to work on bugs now 16:16:38 Also, we need to focus on most important 16:16:44 mihgen: We made Bug Squashing Day today 16:16:50 with aglarendil and bogdando 16:16:56 and alex_didenko 16:17:04 that's why let's continue to collect bugs to be moved to 8.0 16:17:22 I heard guys want to deal with mediums - they are travelling from release to release. we (UI team) also try to squash as much mediums as possible before SCF 16:17:25 and once leads agrees, those can be moved. 16:17:57 I suggest other teams (non-Fuel Library) also do triaging of their bugs and add corresponding non-release tags and set correct priorities 16:17:58 vkramskikh: focus now should be on unblocking extensive testing 16:18:06 #link https://bugs.launchpad.net/fuel/+bug/1479090 16:18:07 Launchpad bug 1479090 in Fuel for OpenStack "API locks up during serialization, Number of nodes to serialize impacts lockup time" [Critical,Confirmed] - Assigned to Sylwester Brzeczkowski (sbrzeczkowski) 16:18:07 Launchpad bug 1479090 in fuel "API locks up during serialization, Number of nodes to serialize impacts lockup time" [Critical,Confirmed] 16:18:09 Launchpad bug 1479090 in fuel "API locks up during serialization, Number of nodes to serialize impacts lockup time" [Critical,Confirmed] https://launchpad.net/bugs/1479090 16:18:32 this one, for example, blocks >=50nodes install. So we need to fix it quickly 16:18:59 we might want to move many mediums / lows to 8.0 but pick some of those if we can later 16:19:18 aglarendil: thank you for triaging session today 16:19:27 xarses: let's move to the aglarendil's update on triaging 16:19:28 I can provide some details 16:19:39 vkramskikh: you can only work on mediums if there's no high & critical bugs that you can fix, even now, before SCF 16:20:17 vkramskikh: we now have new criteria for UX defects for high and critical 16:20:42 angdraug: vkramskikh or you can't take high/critical... if you don't have subject matter expertise to work in there.. 16:20:43 We might want to assess if medium is actually high with UX criteria 16:20:58 but if its been moved from release to release, its probably not urgent 16:21:06 moving to the sub topic 16:21:14 #topic Fuel Library bugs triaging results (aglarendil) 16:21:27 so far we did triaging for Fuel Library bugs today for C/H/M bugs with core reviewers: 16:21:40 sgolovatiuk, alex_didenko and bogdando 16:21:56 brief results for bugs that were somehow changed are here: 16:22:04 #link https://etherpad.mirantis.net/p/7.0-fuellib-bugs-triage 16:22:19 oops, sorry, I will copy it to public location 16:22:22 after the meeting 16:22:53 so far, we added non-release tags and marked some old bugs as incomplete ones 16:22:57 as they were too old 16:23:05 #link https://etherpad.openstack.org/p/7.0-fuellib-bugs-triage 16:23:07 I just did 16:23:31 and could have been fixed by upstream merges or HA improvements or other features 16:23:40 also, we deduplicated some of them and decreased priorites of the bugs 16:23:52 and pushed them 16:23:57 to 8.0 16:24:07 list contains all bugs you guys processed? 16:24:18 please read this document and provide your objections if any 16:24:19 or only those which you have questions / need feedback? 16:24:27 nope, only those that we changed somehow 16:24:43 others we just did not touch saying "it is valid" 16:25:05 ok. Thanks. I actually checked a few, and milestone was still 7.0 16:25:16 so you wanted to get feedback first if any, and then move? 16:25:26 if there are no concerns? 16:25:31 mihgen: we did not move all of them 16:25:41 some of them were marked as wishlist 16:25:50 some of them were marked as low priority and won't fix 16:26:03 wow, you guys where busy 16:26:06 we will surely go over wishlist bugs and attach them to corresponding blueprints 16:26:14 ok understand now. 16:26:24 we wanna get feedback and revert our actions if needed 16:26:29 leaning to lazy consensus here 16:26:48 thanks aglarendil & team - great work, thank you for collecting all them in etherpad so other can review 16:27:33 xarses: moving on? 16:27:37 yep 16:27:41 #topic Change SCF date (mihgen) 16:28:11 question here guys is the following. We want to reach HCF date. and I don't see us converging to it yet 16:28:32 If we move SCF date 1 week earlier, would it allow us to focus on critical / highs more? 16:28:51 and as such to be sure in hitting HCF date? 16:29:17 mihgen i thought we're fixing high/critical now, and we have no time for medium bugs at all 16:29:31 i mean our "medium" bugs usually just move to next release 16:29:54 due to lack of time, so i don't think it change something 16:30:15 let's get some data if it's the case 16:30:43 moving the scf date doesn't really change what we're supposed to be working on other than disrupting if there are any pending reviews for medium bugs out there 16:30:56 mwhahaha: +1 16:31:08 rather than being able to close bugs we are preventing us from doing so by artificial dates 16:31:23 i think 1) discourage working on medium bugs 16:31:26 pending reviews still require attention on review test & merge and introduce risk of regression 16:31:34 that attention is better spent on high priority bugs 16:31:59 2) Re-triage medium bugs against UX criteria to see if they deserve high 16:31:59 while i wait for a bug to reproduce, i can work on other things 16:32:18 just because it's high/crit doesn't mean i don't have time for a minor fix 16:32:19 +1 to move HCF week. it will send clear message that gates for medium bugs will be closed soon 16:32:20 yes, but other things should be other high priority bugs 16:32:32 we don't have that many bugs 16:32:36 3) There should only be a small set of medium bugs that people want to work on 16:32:37 and we will have better focus on CRIT/HIGH 16:32:37 +1 to move 16:33:04 also see my point about medium review increasing unnecessary risk of regressions 16:33:25 #link http://lp-reports.vm.mirantis.net/reports/custom_report?milestone=7.0&status=Fix%20Committed&status=Fix%20Released&importance=Medium&importance=Low&importance=Wishlist&fix_committed_from=2015-07-01&fix_committed_to=2015-09-03&assignee=fuel-python&assignee=fuel-qa&assignee=fuel-provisioning&assignee=fuel-ui&assignee=fuel-library&assignee=fuel-astute 16:33:35 sorry for internal link I hope it will be open soon 16:33:50 but it's 125 medium / low fixed over July 16:34:10 mihgen: ya, I was about to say, I thought we where open sourcing that 16:34:33 xarses: I thought too, authz in there, I don't know why it can't be deployed public 16:34:42 mihgen: ran the same report, got even bigger numbers 16:34:46 so guys still there are many being closed 16:34:56 that's good from one hand 16:35:00 mihgen: if we will move SCF, it can affect some criteria for it - test cases/CI readiness 16:35:23 from another - this time might could be spent on high/critical 16:35:33 nurla: it's SCF, not HCF, why CI is affected at all? 16:35:37 nurla: I don't think any tests should be blocked for code that is already in master 16:35:38 what about test cases? 16:35:53 regardless of the bug priority 16:36:03 we have to implement jobs for test before SCF actually 16:36:11 268 medium and lower bugs w/o customer-found tag fixed in fuel since july 1 based on the incoming/outgoing report 16:36:16 so I think we can still land medium and low for couple more weeks, but the focus should stay around high and critical 16:36:29 87 of those in the last week 16:36:53 angdraug: In Fuel LP project? As my data was about dev teams 16:36:58 not including qa / infra 16:36:58 yes 16:37:20 mihgen: should we vote then or move to the ML 16:37:37 xarses: let's vote 16:38:11 223 excluding docs and infra 16:38:19 76 last week 16:38:47 #startvote move SCF closer 1 week? Yes, No 16:38:48 Begin voting on: move SCF closer 1 week? Valid vote options are Yes, No. 16:38:49 Vote using '#vote OPTION'. Only your last vote counts. 16:39:09 #vote yes 16:39:10 #vote no 16:39:13 #vote yes 16:39:16 #vote yes 16:39:17 I don't like it, but I see the need 16:39:17 #vote yes 16:39:19 #vote yes 16:39:40 #vote yes 16:39:44 irc://fuel-irc.mirantis.com:7777/#vote yes 16:39:44 #vote yes 16:39:45 #vote yes 16:39:46 #vote yes 16:39:48 #vote yes 16:39:51 #vote yes 16:39:56 #vote yes 16:40:05 #vote yes 16:40:05 #vote yes 16:40:06 IvanKliuk:  yes is not a valid option. Valid options are Yes, No. 16:40:29 #vote yes 16:40:34 IvanKliuk: you need to re-try, as there was a wrong message pasted before yes 16:40:42 oh you did, sorry 16:40:44 are you really sure that we haven't some critical in medium issues? 16:40:51 #vote Yes 16:41:00 #vote Yes 16:41:14 nurla: we should check them for the new criteria in UX, but in general they wont be 16:41:14 nurla: we are not, and there is another procedure to move bugs to another release 16:41:22 where we would ask your team to approve it 16:41:38 ok, looks like we cant go the otherway, so I'll end it 16:41:42 #endvote 16:41:43 Voted on "move SCF closer 1 week?" Results are 16:41:44 Yes (15): warpc__, xarses, angdraug, mattymo_, xenolog13, vkramskikh, dpyzhov, akislitsky, evgenyl, IvanKliuk, akasatkin, amaksimov, prmtl, alex_didenko, mihgen 16:41:45 No (1): mwhahaha 16:41:49 so of course if there is anything critical, it must not go to 8.0 16:42:14 Ok, so we will move SCF closer by 1 week 16:42:17 moving on? 16:42:31 #topic Bugs stalled in "in progress" state: http://bit.ly/1Iu6RsH (mihgen) 16:42:32 sorry mwhahaha. It's about focus now.. 16:42:44 #link http://bit.ly/1Iu6RsH 16:42:48 that's fine 16:42:52 fokls I came across 16:43:01 and it's ton of bugs stalled actually 16:43:13 And I started to dig in thinking why 16:43:30 the reason can be lp-reports 16:44:01 where we look only for particular teams of our interest. But if someone unaffiliated with the team takes a bug 16:44:08 it's being shown in Uknown bugs 16:44:09 So how should we attempt to address it? 16:44:14 it's related with review queues, as I remember this week we've got lower review activity and thus growing amount of in progress bugs 16:44:31 angdraug: btw whole your fuel-infra is unkown now I believe ... 16:44:54 shouldn't be, it now has a team in lp-reports 16:45:12 xarses: to address queues, we've got to check SLAs and poke people around, so I'd like to ask leads to ensure that it's the case 16:45:31 for unknown, I'd like someone to take an action item and clean it up 16:45:35 there are alot of in-progress high and even a critical 16:45:41 seems wrong... 16:45:59 i'd expect to see a good amount of medium / low 16:46:07 a bunch of these bugs are in a weird state if you compare the comments vs status, https://bugs.launchpad.net/fuel/+bug/1430376 In progress but was updated to won't fix according to the comments 16:46:08 Launchpad bug 1430376 in Fuel for OpenStack 7.0.x "Duplicate server parameter in dnsmasq.template.erb" [Low,In progress] - Assigned to Bartlomiej Piotrowski (bpiotrowski) 16:46:10 Launchpad bug 1430376 in fuel/7.0.x "Duplicate server parameter in dnsmasq.template.erb" [Low,In progress] 16:46:11 Launchpad bug 1430376 in fuel/7.0.x "Duplicate server parameter in dnsmasq.template.erb" [Low,In progress] https://launchpad.net/bugs/1430376 16:46:40 mwhahaha: LP doesn't always show status updates in comments, check activity history 16:46:43 #link https://bugs.launchpad.net/fuel/+bug/1430376/+activity 16:46:47 that's where you look for it 16:47:17 yea that's not confusing 16:47:26 #link http://lp-reports.vm.mirantis.net/reports/custom_report?milestone=7.0&assignee=unknown 16:47:39 154 unknown bugs in the internal tool we use 16:47:49 so those are getting lost quickly if we filter by teams 16:48:11 angdraug: mos-infra is actually in there 16:48:26 #action mihgen to follow up on lp-reports.vm.mirantis.net still being closed 16:48:37 xarses: thanks I will 16:48:48 is sbanka here? 16:49:14 alex_didenko: can you please follow up with teams and ask them to fill their people in corresponding groups? 16:49:22 mihgen: can we move along so we can get the three other topics? 16:49:22 in order to clean it up? 16:49:54 xarses: moving on 16:49:57 mihgen: ok 16:50:01 #topic librarian wiki & patches (mwhahaha) 16:50:07 alex_didenko: thanks! 16:50:07 I sent out a note yesterday to openstack-dev about the wiki page with an explanation of the librarian work and workflows. Please take some time to read the thread and the wiki page and provide feedback. 16:50:07 #link https://wiki.openstack.org/wiki/Fuel/Library_and_Upstream_Modules 16:50:08 The reviews have been updated and are passing CI as well. 16:50:09 #link https://review.openstack.org/#/q/status:open+project:stackforge/fuel-library+branch:master+topic:bp/fuel-puppet-librarian,n,z 16:51:03 what do we want to land from those? 16:51:03 comments/concerns/etc? 16:51:06 So are we going to have a victim, erm... first module? 16:51:11 stdlib 16:51:12 :D 16:51:22 cinder was dropped? 16:51:28 oh it's there 16:51:40 we need to test in custom iso if we want to land it 16:51:41 but if you look at the ordering of the patches, stdlib is the first 16:51:55 and ensure that people can figure out later where the code is coming from 16:52:22 Also, our upgrade tarball should take the right code 16:52:39 kozhukalov_: ^ 16:52:40 mihgen: nurla can we designate 1 or 2 QA to be trained by mwhahaha so there are no gaps? 16:52:50 sgolovatiuk: any objections? 16:53:02 xarses: 1 should be over than enough 16:53:24 can we land it today? tomorrow? 16:53:36 I'd like to see it landed today 16:53:38 sgolovatiuk: alex_didenko Looks like you'd need to look into it guys, and let's socialize it 16:53:51 the longer we wait, the less likely it's to happen 16:53:55 if we want to merge, then let's do carefully 16:54:15 so what do we need to do before merge to be sure we are ok? 16:54:25 test custom iso? how BVT? Swarm? 16:54:33 there is a discussion in ML let's continue there and move to another topic 16:54:38 custom ISO should be enough 16:54:40 so you have to pick the modules you want to start with 16:54:44 plus what I mentioned 16:55:05 the CI shows that it's still functioning even without core peices like stdlib 16:55:05 moving on 16:55:06 While merging, I'd start with one and then see 16:55:06 we can use 'diff -r' :) 16:55:17 all we need is to confirm that puppet code in fuel-library rpm remains exactly the same as before merge, no? 16:55:29 alex_didenko: +1 16:55:32 and compare /etc/puppet/modules from old library and from new one (with librarian) 16:55:45 i did that as part of the migration of each module 16:55:52 which is where you see some have upgraded versions 16:56:04 so this is something you do during the migration from our repo to librarian 16:56:12 ok, nice 16:56:14 so i'd be happy to talk to someone about it and go do some diffs 16:56:18 grep -r won't work for people on the hook to support 16:56:20 ok, lets move on, unless there are objections on ML we'll land it today 16:56:21 but this is why i picked the modules out there 16:56:24 it has to be easy 16:56:25 #topic Ceph-related bug fixes (mihgen) 16:56:31 #link https://review.openstack.org/#/c/204814/ 16:56:40 I've heard that we don't pay enough attention to this one 16:56:54 just wanted to know what's up in there 16:57:17 looks like we need ceph SMEs to review this one, so xarses please take a look 16:57:25 yep 16:57:35 ok, as far as I know a bugfix contains of two parts - fuel-web and library 16:57:43 alex_didenko: correct 16:57:45 and fuel-web has some backwards compatibility 16:57:52 Ryan and Andrew are the main persons who should review patch. 16:58:03 #link https://review.openstack.org/#/c/204811/ 16:58:06 so we should merge it first, update CI iso and only then merge library part 16:58:06 Also I've added fuel-cores, because there were no any 16:58:07 this is library part 16:59:04 these commits needs more details as to what is being done 16:59:16 is koder here? 16:59:19 these commit messages don't provide enough info on what/why it's changing and what is needed 16:59:20 mattymo: sorry, won't get to you please raise on the ML 16:59:51 i looked at the commits but since i don't know what is supposed to be there, I don't feel comfortable +1'ing them (at least the python one) 17:00:04 ok let's review it guys then and leave comments 17:00:05 mwhahaha: there is a BP that explains it 17:00:13 which is not linked to 17:00:15 in the commit message 17:00:16 So I think we would need to ask for clarifications then 17:00:17 but ya, it needs to be linked 17:00:20 time 17:00:21 if it's not clear 17:00:27 #endmeeting