15:07:14 <bkp> #startmeeting oVirt Weekly Sync
15:07:14 <ovirtbot> Meeting started Wed Nov 26 15:07:14 2014 UTC.  The chair is bkp. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:07:14 <ovirtbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:07:31 <bkp> #topic Agenda and roll Call
15:07:31 <bkp> #info infra update
15:07:31 <bkp> #info 3.5.z updates
15:07:31 <bkp> #info 3.6.0 status
15:07:31 <bkp> #info conferences and workshops
15:07:33 <bkp> #info other topics
15:07:52 <bkp> Roll call! Who's here for what?
15:08:13 * lvernia_ is here for everything.
15:08:15 * sbonazzo here for integration
15:08:28 * lvernia is here as well.
15:08:51 <bkp> lvernia: *You* run the meeting, then! :)
15:09:10 <lvernia> Cool!
15:09:18 <lvernia> Who's here for infra??? :)
15:09:33 <bkp> #topic infra update
15:09:55 <bkp> What lvernia said. dcaro, you here?
15:11:13 <dcaro> bkp: hey, yep
15:11:26 <bkp> What's the word from infra?
15:12:00 * amureini is here, but not for infra
15:12:09 <dcaro> bkp: we hade an outage last week, while upgrading vdsm to 3.5 on the hosts, that left us with some hosts with vdsm 3.4 and some with 3.5, will have to plan the upgrade more carefully
15:12:57 <dcaro> there was also an issue with the mock cleanup that rendered the slaves unusable, that is solved, but ther'es  anew issue that I'm investigating that forced us to take them out of the pool
15:13:19 * tal here for storage
15:13:52 <dcaro> we also did some progress on getting vdsm unit tests inside mock running
15:14:26 <danken> dcaro: could you give more info?
15:14:28 <dcaro> and we got the engine and jenkins upgraded (to 3.5 and 1.583)
15:14:44 <danken> dcaro: did you get it past the blocking test?
15:14:54 <dcaro> danken: there's an email thread on it, but right now only 2 tests are failing
15:14:56 <dcaro> danken: yes
15:15:09 <dcaro> danken: right now it0s the partition check test (that is not expecting the esp flag)
15:15:34 <dcaro> danken: and there was another tests about slow connections
15:16:24 <bkp> Is this about the outage, or the mock cleanup issue?
15:16:43 <dcaro> bkp: it's about vdsm unit tests inside mock (also related to mock cleanup issue)
15:18:07 <sbonazzo> dcaro maybe a nested vm would be better for such tests, fabiand can give a hand on that I guess. but we can take this offline
15:18:10 <bkp> dcaro: Anything else?
15:18:39 <dcaro> bkp: not on my side, minor fixes here and there
15:18:46 * fabiand can do so, yes
15:18:57 <fabiand> dcaro, is there an update on the nesting problems in phx?`
15:19:02 <dcaro> sbonazzo: yep, still have to upgrade another host to fc20 to start testing nested
15:19:27 <dcaro> fabiand: ^ I have to update on host so we can test, but dima said that it's working well
15:19:45 <fabiand> dcaro, okay, would be great if we can slowly fix this one by one ..
15:20:09 <fabiand> that's all I wanted to ask.
15:20:23 <sbonazzo> dcaro bkp https://docs.google.com/spreadsheets/d/1E8e88Nb_OcT7BQprVF7ER6R8--T8mzgc_uHo3i3zeeA/edit#gid=0
15:20:43 <sbonazzo> dcaro we're collecting infra issues for infra day hackathon
15:21:27 <bkp> sbonazzo: What's the date on that, again?
15:21:28 <dcaro> yep :)
15:22:23 <sbonazzo> dcaro do we have a final date for it?
15:24:31 <bkp> Let's keep going, and say date to be determined.
15:24:36 <sbonazzo> bkp +1
15:24:45 <bkp> #info infra We had an outage last week, while upgrading vdsm to 3.5 on the hosts, that left us with some hosts with vdsm 3.4 and some with 3.5, will have to plan the upgrade more carefully.
15:24:49 <bkp> #info infra An issue with the mock cleanup that rendered the slaves unusable, that is solved, but there's a new issue that I'm investigating that forced us to take them out of the pool
15:24:53 <bkp> #info infra We also made some progress on getting vdsm unit tests inside mock running. Two tests are failing, including the partition check test. sbonazzo suggested nested VM approach. Will need another host updated to F20 to enable such an approach.
15:24:57 <bkp> #info infra Engine and jenkins upgraded (to 3.5 and 1.583, respectively)
15:24:59 <bkp> #info infra Nesting problems in PHX update:  dcaro has to update on host so we can test, but dima said that it's working well
15:25:02 <bkp> #info infra Issues are being collected for the Infra Day Hackathon (date TBD), here: https://docs.google.com/spreadsheets/d/1E8e88Nb_OcT7BQprVF7ER6R8--T8mzgc_uHo3i3zeeA/edit#gid=0
15:25:05 <bkp> #topic 3.5.z updates
15:25:15 <bkp> sbonazzo: You're up.
15:25:40 <sbonazzo> bkp sent status this morning here: http://lists.ovirt.org/pipermail/users/2014-November/029543.html
15:25:55 <sbonazzo> bkp bottom line: we needed to postpone RC build 1 week due to 4 blockers
15:26:03 <sbonazzo> bkp today they're 3
15:26:10 <sbonazzo> bkp still need an ETA for them
15:27:27 <sbonazzo> bkp there are also 2 regression to be evaluated for 3.5.1
15:28:32 <sbonazzo> bkp next week seems to be still risky for 3.5.1 RC
15:29:19 <bkp> sbonazzo: Noted. Any other issues?
15:29:46 <sbonazzo> bkp: almost all jenkins jobs for 3.5 are now green (except one for node, fabiand team is already on it)
15:30:00 <fabiand> sbonazzo, yep, we need nesting to get them green again ..
15:31:01 <sbonazzo> bkp that may be all for 3.5
15:31:19 <bkp> Got it...
15:31:29 <bkp> #info 3.5.z updates Full status here: http://lists.ovirt.org/pipermail/users/2014-November/029543.html
15:31:32 <bkp> #info 3.5.z updates Three blockers (1162640, 1160846, and 1159839) in place for 3.5.1 RC, no ETA set on fixes.
15:31:35 <bkp> Two regressions (1160846 and 1118349) are tagged for 3.5.1 and need to be evaluated.
15:31:38 <bkp> #info 3.5.z updates RC will need to be postponed one week to accomodate blockers, but that date is even risky.
15:31:41 <bkp> #info 3.5.z updates Almost all jenkins jobs for 3.5 are now green (except one for node, fabiand team is already on it) (fabiand needs nesting to get them green, see infra note above)
15:31:45 <bkp> #topic 3.6 status
15:32:01 <sbonazzo> bkp missed a #info on the regressions
15:32:58 <bkp> My bad.
15:32:59 <sbonazzo> bkp about 3.6, sent a status  recap this morning: http://lists.ovirt.org/pipermail/users/2014-November/029545.html
15:33:01 <bkp> #info 3.5.z updates addendum: Two regressions (1160846 and 1118349) are tagged for 3.5.1 and need to be evaluated.
15:33:24 <sbonazzo> bkp no blockers as expected, bug count increasing
15:34:14 <sbonazzo> bkp there was a repoclosure error on missing mom-0.4.3 but alitke released it today fixing the closure
15:34:35 <sbonazzo> bkp: just need verification and karma on fedora systems
15:35:00 <danken> sbonazzo: adam litke sent an email about it. msivak knows about the issue
15:35:12 <sbonazzo> bkp as per last ovirt sync meeting today is last day for release criteria / process
15:35:25 <sbonazzo> danken: saw it, but it's not an issue anymore
15:35:33 <amureini> bkp, sbonazzo wrt 1118349 that you just mentioned - it's a slight regression, but a very uninteresting usecase. I wouldn't block anything on it.
15:36:03 <sbonazzo> amureini: please retarget to 3.5.2 or 3.6.0 if it won't be in 3.5.1
15:36:36 <amureini> sbonazzo, shoot, thought I did. on it now.
15:37:03 <bkp> sbonazzo: Did anyone add anything to the criteria discussion?
15:37:23 <sbonazzo> bkp about the release process, sven kieske posted a comment on alpha / feature freeze order
15:37:45 <sbonazzo> bkp and I agree with him that a slight rephrase it's needed for clarification
15:38:15 <sbonazzo> bkp "feature acceptance review" should be before alpha and feature freeze at alpha end
15:38:30 <sbonazzo> bkp other than that I've not yet seen any other comments
15:38:58 <sbonazzo> bkp so I guess we can go with it and proposed fixes by Sven
15:39:23 <bkp> sbonazzo: What would be the best way to finalize and announce?
15:40:07 <sbonazzo> bkp: well, probably better to prepare a wiki page with the new process / criteria and then publicize it
15:40:14 <bkp> sbonazzo: And where is that thread again?
15:40:55 <sbonazzo> bkp sven proposed changes are replies to  http://lists.ovirt.org/pipermail/users/2014-November/029545.html
15:41:14 <sbonazzo> bkp original thread is http://lists.ovirt.org/pipermail/devel/2014-September/008695.html
15:41:42 <sbonazzo> lvernia: comments? :-)
15:41:55 <lvernia> sbonazzo: I've said all I had to say... :)
15:42:07 <sbonazzo> lvernia: thanks
15:42:29 <bkp> Okay, sbonazzo want me to set up the wiki with the revised process?
15:42:49 <sbonazzo> bkp it would be nice
15:43:18 <bkp> Any other 3.6 issues?
15:43:59 <sbonazzo> bkp: still gathering 3.6 features on http://goo.gl/9X3G49
15:44:35 <lvernia> Still?
15:44:44 <lvernia> I may have missed the e-mail about it!
15:44:45 <lvernia> :)
15:45:00 <lvernia> Ahh, I see.
15:45:09 <lvernia> At the end of today's e-mail?
15:45:14 <bkp> Just so I know, when is the closure on that gathering?
15:45:22 <sbonazzo> lvernia: well, it was also in last week email
15:45:28 <lvernia> Ack, sorry.
15:45:31 <sbonazzo> lvernia: not sure about 2 weeks ago
15:45:33 <sbonazzo> :-)
15:45:38 <lvernia> I guess I was all wrapped up in the schedule discussion.
15:46:02 <bkp> In my defense, I have slept since then.
15:46:23 <bkp> That it for 3.6? Anyone else?
15:46:25 <lvernia> :)
15:46:41 <lvernia> So just throw everything on there? Assuming it might be a 10-month cycle?
15:46:44 <lvernia> Or something of the sort?
15:46:53 <sbonazzo> bkp: no decision yet on deadline for feature submission
15:47:05 <jmoskovc> rgolan: still there?
15:48:02 <jmoskovc> rgolan: unping solved it ;)
15:48:24 <bkp> lvernia: I'd say write everything down (and I will always say that) and then if we stick to a six-month cycle, then we can parse out the features from the mega-list.
15:48:42 <sbonazzo> lvernia: I guess it's better to throw everything on there while we reach an agreement on if it's better to split the cycle in 2 or have a long one
15:48:47 <bkp> Better to have more than less.
15:48:55 <lvernia> Ack.
15:49:37 <bkp> K, not hearing any other teams chime in. I think it's too soon to do the team rundown, so if there are no other 3.6 topics?
15:50:35 <sbonazzo> bkp that's all on my side
15:51:13 <bkp> #info 3.6 status Full status here: http://lists.ovirt.org/pipermail/users/2014-November/029545.html
15:51:16 <bkp> #info 3.6 status No blockers (as expected). Bug count on the rise.
15:51:19 <bkp> #info 3.6 status An earlier repoclosure error on missing mom-0.4.3 was repaired and released today by alitke (just needed verification and karma on Fedora systems)
15:51:22 <bkp> #info 3.6 status Today is last day for release criteria/process discussion. skieske made a rephrase suggestion, about "feature acceptance review", which was accepted.
15:51:25 <bkp> #info 3.6 status Updated release criteria/process at http://lists.ovirt.org/pipermail/users/2014-November/028875.html will be posted on wiki and then announced soon.
15:51:28 <bkp> #action bkp to set up new release process page on wiki
15:51:30 <bkp> #info 3.6 status Still gathering 3.6 features at http://goo.gl/9X3G49, no feature deadline set yet.
15:51:33 <bkp> #info 3.6 status Length of release cycle still not confirmed, but gathering all features anyway to have them in one place.
15:51:39 <bkp> #topic conferences and workshops
15:51:47 <J_Man> hello everyone!  Quick Question - is it still the case that hosted-engine can only run on NFS or ISCSI?  Can't use libgfapi to host it?
15:52:01 <bkp> Not much from me, 2014 winding down.
15:53:13 <bkp> #info Conferences Notification on FOSDEM stands is on Dec. 15. Still time to get your devroom submittals in for the virt and IaaS devrooms.
15:53:16 <bkp> #info Conferences I am in discussions with FOSSAsia (http://fossasia.org/) to see about an oVirt workshop at that event in mid-March. This is a Singapore venue, so if anyone from east Asia is interested in speaking/attending, let bkp know.
15:53:20 <bkp> #info Conferences If anyone wants to submit a quick report on DevOpsDays TLV for the Community blog, that would be welcome!
15:53:22 <sbonazzo> J_Man: meeting in progress, it will probably be in 3.6
15:53:24 <bkp> Any other conference news?
15:54:28 <bkp> Very well, moving along...
15:54:31 <bkp> #topic other topics
15:55:02 <bkp> #info Other A reminder that the U.S. will be offline tomorrow and Friday for Thanksgiving holiday.
15:55:05 <bkp> #info Other bkp will be recording demo videos of oVirt beginning after the holidays. If anyone wants to help/suggest topics, ping me.
15:55:23 <bkp> Any other meta-issues to discuss?
15:56:48 <bkp> Going once?
15:57:17 <bkp> Twice?
15:57:58 <bkp> That's it for this week. Thanks to all! See you online and remember: If Plan A fails, you always have another 25 letters to go!
15:58:02 <bkp> #endmeeting