15:00:17 #startmeeting RDO meeting (2016-04-06) 15:00:17 Meeting started Wed Apr 6 15:00:17 2016 UTC. The chair is imcsk8. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:17 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:17 The meeting name has been set to 'rdo_meeting_(2016-04-06)' 15:00:23 \o/ 15:00:26 o/ 15:00:36 0/ 15:00:38 topic rollcall ? 15:00:45 #chair chandankumar saneax apevec rbowen 15:00:45 Current chairs: apevec chandankumar imcsk8 rbowen saneax 15:00:48 #topic do we put the dashboard (http://46.231.133.253:3030/rdo-dev ) under rdoproject.org ? 15:00:54 o/ 15:00:59 flepied, ^ ? 15:01:05 #chair dmsimard 15:01:05 Current chairs: apevec chandankumar dmsimard imcsk8 rbowen saneax 15:01:06 is that javascript only? 15:01:09 \o/ 15:01:14 o/ 15:01:17 #chair jruzicka flepied 15:01:17 Current chairs: apevec chandankumar dmsimard flepied imcsk8 jruzicka rbowen saneax 15:01:20 o/ 15:01:22 if so, could we just put it inside website? rbowen ? 15:01:31 #chair number80 15:01:31 Current chairs: apevec chandankumar dmsimard flepied imcsk8 jruzicka number80 rbowen saneax 15:01:38 o/ 15:01:43 #chair jpena 15:01:43 Current chairs: apevec chandankumar dmsimard flepied imcsk8 jpena jruzicka number80 rbowen saneax 15:01:55 Well, if we do some front page redesign, then, sure. 15:02:12 Which would be awesome, and pkovar has been giving that some thought 15:02:12 it would be nice to have it in a very visible place 15:02:18 rbowen, no, that's status of builds pipeline 15:02:19 And working on that with tigert 15:02:24 o/ 15:02:31 #chair trown 15:02:31 Current chairs: apevec chandankumar dmsimard flepied imcsk8 jpena jruzicka number80 rbowen saneax trown 15:02:34 front page redesign +111 15:02:38 Oh, right. 15:02:38 rbowen, it wouldn't be on front page 15:02:39 o/ 15:02:44 #chair amoralej 15:02:44 Current chairs: amoralej apevec chandankumar dmsimard flepied imcsk8 jpena jruzicka number80 rbowen saneax trown 15:02:55 rbowen: i can help with that 15:03:01 status.rdoproject.org? 15:03:09 heh, I guess we should've stayed on rollcall few mins 15:03:18 pkovar++ 15:03:46 number80, yeah, but that's usually all systems status for projects 15:04:01 cf http://status.fedoraproject.org/ 15:04:10 or http://status.openstack.org/ 15:04:34 so more like monitoring? 15:04:39 that dmsimard is working 15:04:44 ack 15:04:45 As we add new things to our vast galaxy of RDO websites, we need to be sure to document this somewhere. 15:04:51 It's kind of hard to keep track of it all. 15:04:58 Both for sanity and for active monitoring. 15:05:07 *nods8 15:05:20 Even just a page on rdoproject.org that lists all of the various websites would be helpful to me. 15:05:30 'Cause we're up to 8 or 10 right now. 15:05:32 could it be also in a frame on the main page? or is it too much info for newbies? 15:05:37 yeah, so an idea: flepied could that dashboard be just part of monitoring ? 15:05:42 that sounds like a task for status.rdoproject.org - list all RDO websites and whatnot along with status 15:05:44 or we really need that fancy JS ? 15:05:58 * imcsk8 likes the fancy js :) 15:06:04 it could be a dashboard link somewhere 15:06:13 yeap, we should have pointing to all available pages, but i think the status page is different 15:06:29 and also be integrated into the one dmsimard is working on later 15:06:45 flepied, it is just for looking i.e. your dashboard won't trigger any alarms or something? 15:06:52 Having a live status feed on the front page would be awesome. 15:07:24 pulse.rdoproject.org, watch RDO pulse live :) 15:07:28 but we should not mix the target consumers: sys admin of our platforms and packages/testers 15:07:33 packagers 15:07:34 extra points if it would also answer the "is rpmfactory down?" question ;) 15:08:01 flepied has a point, the community at large might not need to know there are cpu spikes or stuff like that 15:08:28 we could curate the ongoing alarms and display them, maybe with something like https://demo.cachethq.io/ 15:08:32 it's API driven and stuff. 15:08:44 and open source :p 15:08:45 dmsimard, +1 15:09:13 ok, so for now, let's include flepied's dashboard as is somewhere in website? 15:09:21 w/o creating new subdomain 15:09:33 rbowen, ^ can that work w/ middleman? 15:09:41 we need then, better CSS 15:09:49 flepied, please link to dashboard source 15:10:08 apevec: Yes. We can do content in Markdown, but we can also embed raw HTML anywhere we want. 15:10:15 #link https://github.com/rdo-infra/rdo-dashboards 15:10:24 number80, it's just red and green, what can be bettered :) 15:10:43 apevec: it feels foreign compared to the rest of the website 15:11:05 flepied, ah it needs crontab? 15:11:14 apevec: yes 15:11:27 number80: for what i see it could be easily adapted to the RDO website CSS 15:11:28 so it's not accessing datasource directly from JS, it's prebuilt 15:11:39 If it requires more than just putting something into a web page, we'd need to work with misc/Duck to make that happen. 15:11:58 rbowen, yeah, that's the case... 15:12:17 ok, then we should probably take this to email and figure out what the requirements are. 15:12:21 ack 15:12:30 As misc is out of office today, and that'll probably take a while anyways. 15:12:51 so i guess this is an action right? 15:13:00 Yes 15:13:27 #action Discuss monitoring website requirements on the ML 15:13:31 #undo 15:13:31 Removing item from minutes: ACTION by imcsk8 at 15:13:27 : Discuss monitoring website requirements on the ML 15:13:37 you need someone :) 15:13:39 #action flepied, rbowen, pkovar to sync on what's involved in getting the dashboard onto the front page of the site. 15:13:43 How's that. 15:13:46 ohhh damn! :( 15:13:53 rbowen, ack 15:14:11 rbowen: i would like to help with that 15:14:12 imcsk8, #action 15:14:19 #undo 15:14:19 Removing item from minutes: ACTION by rbowen at 15:13:39 : flepied, rbowen, pkovar to sync on what's involved in getting the dashboard onto the front page of the site. 15:14:26 #action flepied, rbowen, pkovar, imcsk8 to sync on what's involved in getting the dashboard onto the front page of the site. 15:14:40 However, if we do that on rdo-list, anybody can jump in. 15:14:41 should we proceed with next topic? 15:14:57 +1 15:15:04 +1 15:15:23 #topic rpm factory agenda 15:15:36 fbo or flepied ^ ? 15:16:10 we worked on a retro planing to be able to put in rpm factory in production 15:16:27 after the Mitaka release and before the summit 15:16:49 which is exactly week of Mar 18 15:16:52 so it must work, or we'll be publicly shamed :) 15:17:04 #link http://softwarefactory-project.io/etherpad/p/Retro_Planning_RDO_-_RPMF 15:17:07 yes, better fix that Ceph :) 15:17:08 for details 15:17:12 Apr 18? 15:17:21 yes Apr 18 15:17:24 or are we using a time machine :p 15:17:27 err Apr ! 15:17:33 I'm so last month 15:17:38 back in time 15:17:56 too much delorean 15:18:05 yeah :) 15:18:08 hehe 15:18:37 are we done with this topic? 15:19:01 flepied, fbo - please post on rdo-list 15:19:27 not everybody reads meeting minutes 15:19:35 #action fbo to post a summary of the plan on rdo list 15:19:55 ok, next topic 15:20:04 #topic Any last minute RDO Mitaka GA blockers? 15:20:21 that's mine 15:20:41 I summarized etherpad where we're collecting RC issues 15:21:01 https://etherpad.openstack.org/p/rdo-mitaka-stable-prep 15:21:27 one of not crossed is about missing new *ui packages 15:21:27 looks good except the selinux issue 15:21:50 so what's the feeling, should that be a blocker? 15:22:05 the ui packages arrived quite late, so they should be 0 day updates 15:22:07 selinux has not been a blocker for RDO in the past 15:22:12 it is not covered by CI but it wouldn't break other things 15:22:12 trown: true 15:22:39 that port 5000 is also not really reproducible afaict 15:22:44 I kind of see selinux as an enterprisy thing 15:23:03 i just tested the repo yesterday and didn't get the 5000 error 15:23:05 apevec: 0day updates, I'm reviewing/fixing trove/sahara-ui to fasten the process 15:23:24 trown: it's enabled by default on CentOS 15:23:24 quack 15:23:36 #agreed push new *ui packages as 0day updates, not blocking Mitaka GA 15:23:42 #chair Duck 15:23:42 Current chairs: Duck amoralej apevec chandankumar dmsimard flepied imcsk8 jpena jruzicka number80 rbowen saneax trown 15:24:15 number80: how can you reproduce the selinux issue? I've tried a few times, with packstack and tripleo, without luck 15:24:16 number80: sorry i didn't get the quack 15:24:16 trown, that's actualy subtopic later, about DoD 15:24:31 imcsk8: Duck is the new sysadmin working with misc 15:24:41 it worked for me with selinux enabled too 15:24:47 number80: hmm, fair, and I learend recently that Cent is Comunnity Enterprise... but maybe we should have it as an explicit blocker early in Newton rather than last minute blocker at the end of Mitaka 15:24:52 how nice, I've got a nice chair to sit on :-) 15:25:12 ack, let's move to the next topic we have plenty today 15:25:16 just when I went back home and was tired, that's sweet 15:25:22 ok, so can anyone reproduce selinux issues? 15:25:26 ok, next topic 15:25:29 imcsk8, no 15:25:38 we're going through subtopics :) 15:25:42 ohh ok 15:26:05 #info selinux port 5000 issue is not reproducible 15:26:19 ^ unless someone speaks up w/ steps to reproduce 15:26:19 this is kind of a tough meeting for a newer chair :), you are doing great though imcsk8 15:26:19 apevec: I have a vagrant file that automate it for me on fresh CentOS image 15:26:36 trown: thanks :) 15:26:42 number80, so you can consistently reproduce? 15:26:45 yup 15:26:59 just launching packstack --allinone 15:27:04 and wait 15:27:09 ok, but before that? 15:27:22 number80: once you get it, could you give me access to the machine, so I can do some checks? 15:27:25 I can help on topics like migrating to Middleman4 is a pain in the ass and how macha icecreams really kicks ass 15:27:29 I need all steps incl. image etc 15:27:42 Duck, append in https://etherpad.openstack.org/p/RDO-Meeting 15:27:45 apevec: setting packages 15:27:46 and we'll try to get to it 15:27:48 jpena: ack 15:27:55 #undo 15:27:55 Removing item from minutes: INFO by apevec at 15:26:05 : selinux port 5000 issue is not reproducible 15:28:15 #action jpena to look at number80's machine which produces selinux issue 15:28:53 next, subtopic? 15:29:03 yes, packstack GA release 15:29:06 #info packstack GA release 15:29:14 we had trouble merging upstream 15:29:25 hitting 0 swap nodes 15:29:44 yes, I will try to ask pabelanger for help today 15:30:01 as it turns out -infra might want a cleaner approach to making the virtual machines consistent 15:30:02 there are few patches which I consider blockers 15:30:18 stuck between a rock and a hard place 15:30:21 so should I try dice w/ rechecks? 15:30:41 there might be no other way, I don't see such a large-reaching change to be implemented quickly 15:30:59 or we could keep current RC2 with patches in RPM 15:31:01 alternatively, we can make the jobs non-voting 15:31:04 then update to GA tarball as 0day 15:31:10 but we have to be very careful 15:31:27 dmsimard, yeah, let's not do that just yet 15:31:39 I like recheck, then fallback to patches and GA tarball as 0day 15:31:51 apevec, dmsimard let me know if you need anything for packstack 15:31:54 apevec: if RC2 works then let's give time to fix gating 15:32:10 #agreed push packstack RC as is then update to packstack upstream GA tarball later 15:32:19 ok next subtopic 15:32:30 #link https://review.gerrithub.io/#/q/status:open+projects:openstack-packages+branch:rpm-master 15:32:44 ^ is any of that mitaka GA blocker? 15:33:07 there might be more actually, I think we need full reqcheck sync otherwise per-service upgrades won't work 15:33:20 due to too old min deps in spec 15:33:27 that should be CI job, eventually 15:34:42 looks like nobody sees a blocker? 15:34:49 nope 15:34:55 no blockers for me 15:34:55 nope 15:35:09 (I can do a reqcheck pass though for services though) 15:35:31 number80, ok, let's do that during GA respins 15:35:41 we can split that between two of us 15:36:00 wfm 15:36:02 #action number80 and apevec to do reqcheck during GA respins 15:36:12 rdopkg is helping, yay 15:36:17 ok, and now DoD, take #3 ! 15:36:35 #info defnition of done 15:36:51 we just mentioned selinux, do we have it in the release check list? 15:36:54 lol I would love to complete this task... only had it for a quarter year 15:36:59 I have to step out, girlfriend is not able to pick kid up from school.. brb 15:37:28 trown, so last proposal was to declare done when both packstack and tripleo-nonha pass CI 15:37:31 did I get that right? 15:37:52 and we just need to write that in stone somewhere 15:37:56 probably website page 15:38:23 +1 15:38:30 +1 15:38:41 apevec: I had tripleo HA also 15:38:42 +1.00001 15:39:02 apevec: just no tempest in the HA case because of resource constraints 15:39:05 hey guys I just installed https://repos.fedorapeople.org/repos/openstack/openstack-mitaka/rdo-release-mitaka-1.noarch.rpm and I'm getting a 404 from the mirror that's listed in th rpoe file 15:39:12 ah, then I misunderstood 15:39:23 failure: repodata/repomd.xml from openstack-mitaka: [Errno 256] No more mirrors to try. 15:39:23 radez: we're in a meeting 15:39:24 http://mirror.centos.org/centos/7/cloud/x86_64/openstack-mitaka/repodata/repomd.xml: [Errno 14] HTTP Error 404 - Not Found 15:39:25 so, nonha full tempest 15:39:29 apevec: RDO - bunch of python stuff 15:39:29 https://www.redhat.com/archives/rdo-list/2016-March/msg00118.html is the last attempt 15:39:35 and HA = pingtest ? 15:39:41 imcsk8: ok 15:39:55 nonha with tempest smoke, ha with pingtest 15:40:04 radez, buildlogs only for now 15:40:16 radez, details later 15:40:38 ok, and for packstack, whatever weirdo is doing, smoke I guess? 15:40:39 apevec: buildlogs? 15:41:10 radez, just use rdo-release-mitaka.rpm it will enable testing repo 15:41:16 but details later please :) 15:41:38 trown, ok, let's put that on vote finally 15:41:50 +1 15:41:53 +1 15:42:01 apevec: oh gotcha, thx 15:42:22 +1 15:42:34 packstack aio tested with weirdo job, scenario001--003, tripleo-quickstart nonha with tempest smoke, ha with pingtest 15:42:50 ^ that'd be current RDO definition of done 15:43:02 nice 15:43:03 trown, that is what runs in promotion pipeline right? 15:43:07 I would just add 'technical' definition of done 15:43:12 yes 15:43:20 since there are also non-technical artifacts 15:43:27 rbowen has the list for that other stuff 15:43:44 apevec: pretty close, need to switch nonha back to tempest (it was broken for a while) 15:44:17 #agreed RDO 'technical' definition of done: packstack aio tested with weirdo job, scenario001--003, tripleo-quickstart nonha with tempest smoke, ha with pingtest 15:44:30 next topic is non-technical part :) 15:44:41 #topic RDO Mitaka release announcement, Tuesday April 12 - https://etherpad.openstack.org/p/mitaka-rdo-release - Speak up now if you have any objections to either the phrasing or the timing. 15:44:57 so timing is tight, considering we need to: 15:45:10 1) respin after GA tarballs are available on Thu 15:45:15 We're planning to push this release notice out on the 12th. The date is necessary because we are coordinating with other news outlets. 15:45:26 So if we need to push it later, I need to know soon-ish 15:45:33 It's fine if we do, I just need to know. 15:45:37 2) rerun CI pipeline using testing repo (in construction by trown and dmsimard today) 15:46:02 rbowen, what's the latest you need to know? 15:46:21 number80 and I will try to respin tomorrow asap 15:46:28 It would be best to know by Friday if we need to push it back. 15:46:33 ok 15:46:38 ack 15:46:44 We'll be on LWN, OpenSource.com and ... I forget. A couple other places. 15:46:50 trown, dmsimard - do you think we can have that new CI pipeline by tomorrow ? 15:47:03 rbowen, so ... no pressure, right :) 15:47:04 rbowen: in my twitter! :) 15:47:06 apevec: dmsimard had to leave 15:47:10 Although, if we need to go ahead and give it a couple more days right now, that's fine. 15:47:15 rbowen, what about from redhat community blog? 15:47:24 It's not that it has to be the 12th, that was a date that was picked. 15:47:39 chandankumar: Yes, there, too, and RedHatStack, and rdoproject.org and seven.centos.org 15:47:47 ok, I'll check for ETA w/ dmsimard and trown later 15:48:04 apevec: worst case I can setup a one off job for tripleo and dmsimard already has one off jobs for weirdo 15:48:17 #info apevec to give GO to rbowen by Friday 15:48:52 trown, yes, I was using generic-weirdo to qualify RCs 15:48:56 Thanks. 15:49:09 trown, and relying on your manual testing of RC1 for 3o 15:49:32 ok, next topic? 15:49:36 yes 15:49:39 apevec: yep, I can at least turn my manual test into a public job by tomorrow 15:49:51 trown, thanks! 15:49:55 * chandankumar reminds we have 11 mins left. 15:49:58 #topic OpenStack Summit (April 25-29, Austin) - Want to hang out with Rich in the RDO booth? https://etherpad.openstack.org/p/rdo-austin-summit-booth 15:50:10 That's the whole message. Sign up if you want to do a demo on the RDO booth at Summit. 15:50:12 /EOL 15:50:42 rbowen, are all RDO ducks lined up? :) 15:50:47 Yes. :-) 15:50:49 We have DUCKS 15:50:50 rbowen: I registered few slots for RDO contributors (current and new) to present our workflow 15:50:54 rbowen, photo! 15:51:04 I haven't seen them yet, but they're wearing red cowboy hats. 15:51:06 i would like to go, but i don't have money to go to Austin, dollars are too expensive 15:51:17 Next topic. :-) 15:51:25 #topic Mitaka Test Day - April 13, 14 - https://www.rdoproject.org/testday/mitaka/ga/ 15:51:29 I have a meeting at the next hour. 15:51:36 That topic, also, is all in the bullet point. 15:51:50 Although if you want specific things tested, please look at the tests page and update it. 15:51:54 imcsk8, let me know when i proceed with my topic 15:51:59 /EOL 15:52:01 rbowen: I'll add upstream install guide 15:52:07 Awesome. 15:52:13 chandankumar: which one is yours? 15:52:18 imcsk8, DLRN 15:52:27 well is the next one 15:52:30 should we proceed? 15:52:38 I'm done. :-) 15:52:44 #topic DLRN review queue cleanup 15:53:03 so i have complete all the codes and docs changes for DLRN renaming 15:53:12 I'll do quick reviews after the meeting 15:53:26 afaict all are good to merge now 15:53:27 and we have some reviews pending to complete the tasks 15:53:33 https://review.gerrithub.io/271220 15:53:33 https://review.gerrithub.io/271544 15:53:33 https://review.gerrithub.io/270799 15:53:33 https://review.gerrithub.io/271026 15:53:34 https://review.gerrithub.io/269944 15:53:34 https://review.gerrithub.io/265986 15:53:55 should this ones go as links? 15:54:06 yes 15:54:06 #action apevec to review and merge pending DLRN reviews 15:54:19 #link https://review.gerrithub.io/271220 15:54:26 #link https://review.gerrithub.io/271544 15:54:31 #link https://review.gerrithub.io/270799 15:54:33 i will be doing some code cleanup in DLRN by next week 15:54:36 I think those are automatically linked 15:54:46 when you paste URLs 15:54:53 ohhh 15:54:54 ok 15:54:56 apevec, nope 15:55:16 i'll add the ones left just to be sure 15:55:16 really, it worked w/ old meetbot 15:55:20 ok 15:55:24 #link https://review.gerrithub.io/271026 15:55:27 twice is nice :) 15:55:31 #link https://review.gerrithub.io/269944 15:55:32 It links them in the transcript. Not sure it does in the minutes. 15:55:38 #link https://review.gerrithub.io/265986 15:55:57 i am also working on rdopkg docs migration to readthdocs 15:56:27 chandankumar, check w/ number80,he did asciidoc to md conversion for packaging docs 15:56:41 number80, you used pandoc? 15:56:43 chandankumar: i can help reviewing that or if you need a hand i can help 15:56:59 imcsk8, sure. 15:57:07 let me know 15:57:10 15:57:15 next topic 15:57:21 #topic Chair for next meeting 15:57:31 jruzicka, you will do next meeting? 15:57:42 it was grabs last time :) 15:57:48 yes 15:57:50 hehehe 15:57:58 #action jruzicka to chair next meeting 15:58:12 2 mins left. 15:58:17 need to work on my mad charing skillz :-p 15:58:25 jruzicka, haha 15:58:32 impressive, we did it right on time! 15:58:33 me too, i'm a lousy chair 15:58:44 * number80 has a call 15:58:45 should we end? 15:58:50 imcsk8, you did it great 15:58:52 yup, right on time :) 15:58:52 imcsk8: thanks for chairing :) 15:58:54 it was lots of topic 15:58:59 thanks! 15:59:06 well, lots of stuff is happening ;) 15:59:09 thank you guys 15:59:10 and that's good. 15:59:10 Thank you. 15:59:11 imcsk8, thanks for chairing. 15:59:16 ending meeting 15:59:16 imcsk8: this was a super busy meeting because or release stuff and you still finished on time 15:59:23 kudos 15:59:30 #endmeeting