15:02:52 #startmeeting oVirt weekly sync meeting 15:02:52 Meeting started Wed Feb 1 15:02:52 2012 UTC. The chair is rbergeron. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:02:52 Useful Commands: #action #agreed #help #info #idea #link #topic. 15:03:06 #meetingname oVirt weekly sync meeting 15:03:06 The meeting name has been set to 'ovirt_weekly_sync_meeting' 15:03:18 #topic Agenda for today 15:03:24 #chair oschreib cctrieloff quaid 15:03:24 Current chairs: cctrieloff oschreib quaid rbergeron 15:03:47 oschreib: am I correct in thinking that the major topic for today is go/no go/release status? 15:03:53 indeed 15:03:57 * sgordon is here 15:04:00 * lpeer is here 15:04:06 * oschreib here 15:04:07 * mburns here 15:04:11 If anyone else has anything, speak up, but we should probably get down to business. 15:04:11 we have a internal meeting here today that I need to go to., will take my machine with, but may drop in/ out. 15:04:24 #topic Release Status 15:04:35 oschreib: take it away, where are we at? 15:04:45 We had the go/no go meeting 15:04:46 #chair mburns lpeer sgordon 15:04:46 Current chairs: cctrieloff lpeer mburns oschreib quaid rbergeron sgordon 15:04:53 quaid could also update, but for open day in beijing, still working location for confirmation for date. 15:05:03 * jimjag is here 15:05:08 and mgoldboi and ykaul raised several issues 15:05:20 some of them were not part of the release crteria 15:05:20 * Guest747 here too 15:05:45 We had one clean VDSM blocker 15:05:53 #link http://ovirt.org/wiki/Releases/First_Release_Blockers 15:06:00 we should discuess it's status 15:07:05 oschreib: got a patch from vdsm team but we were blocked on testing it because of https://bugzilla.redhat.com/show_bug.cgi?id=786174 15:07:21 * rharper is here 15:07:22 oschreib: vdsm isn't starting because of it 15:07:40 danken: ping 15:07:45 oschreib - there were multiple major issues in my email., the iptables one, the logo/brand one 15:08:20 acathrow: any BZ on the iptables and logo? 15:08:34 lets try to understand what are the blockers 15:09:18 mburns: any updates from your side first? 15:09:23 do we want to go through the bz's one by one? 15:09:33 I believe we do. 15:09:50 oschreib: sure 15:09:51 oschreib: 2 issues for node 15:09:59 #link https://bugzilla.redhat.com/show_bug.cgi?id=785728 15:10:23 this one we can't reproduce anywhere, but we suspect that it's dmraid related 15:10:40 but is not a blocker 15:10:51 #link https://bugzilla.redhat.com/show_bug.cgi?id=782663 15:11:01 this is engine but directly impacts node 15:11:17 fix is posted, but not merged yet afaict 15:11:28 lpeer: any comment? 15:11:39 mburns: uhm, I had problems with ovirt-node and dmraid, too. So I stopped using dmraid 15:12:17 xTs_w: yes, dmraid doesn't work right (and i don't remember all the details for why) 15:12:27 sounds to me that BZ#https://bugzilla.redhat.com/show_bug.cgi?id=782663 is clearly a blocker. 15:12:27 any other opinion? 15:12:27 we had workarounds for disabling it, but those appear to not be working 15:12:55 we'll add a release note for dmraid, but it's not a blocker 15:13:00 oschreib: sec loking 15:13:02 oschreib: ack, 782663 is a blocker 15:13:18 lets hear lpeer thoughts 15:13:49 mburns: so nothing on UEFI, right? 15:14:09 #agreed https://bugzilla.redhat.com/show_bug.cgi?id=782663 is a blocker 15:14:17 oschreib: testing with uefi all worked (with 1 additional small patch) 15:14:20 back 15:14:36 mburns: good to know. 15:14:37 oschreib: new build will be done when we have final vdsm 15:15:17 we will talk about vdsm in after engine 15:15:33 oschreib: it looks like a blocker to me 15:15:44 lpeer: ok, thanks. 15:16:00 lpeer: can you please talk with rgolan1 to fix it and push it into engine_3.0? 15:16:06 oschreib: the patches seems in place i can have them acked and pushed by tomorrow morning 15:16:14 lpeer: good, thanks 15:16:14 oschreib: will do 15:16:28 #info https://bugzilla.redhat.com/show_bug.cgi?id=782663 should be fixed and merged tomorrow 15:17:22 any other comments about node status for first release? sounds to me like we're fine 15:18:20 * rbergeron has nothing 15:18:26 #info ovirt-node is ready, waiting for last vdsm build. 15:18:31 oschreib: should be all good once we get new vdsm and engine bug fixed 15:18:47 lets talk about engine now 15:18:51 mgoldboi: ? 15:19:36 i brought my candidates the last time - snapshot issues - bugs has been modified - lpeer ready to be pushed? 15:19:48 mgoldboi: we already did 15:19:57 afaik 15:20:08 ok- no news on the engine side 15:20:09 oschreib: indeed 15:20:12 cool 15:20:25 #info engine is ready for first release 15:20:34 ok. 15:20:38 vdsm now 15:20:57 oschreib do we consider bootstrap part of engine? 15:21:06 acathrow: part of vdsm 15:21:14 we have a major problem with https://bugzilla.redhat.com/show_bug.cgi?id=786174 and vdsm 15:21:21 iscsid service fails to start 15:21:40 oschreib: ? 15:21:41 thus vdsm isn't starting 15:21:53 I see https://bugzilla.redhat.com/show_bug.cgi?id=773371 in the relesae blokcers, is it really an issue? 15:21:58 14[[07Releases/First Release Blockers14]]4 !10 02http://ovirt.org/w/index.php?diff=2148&oldid=2147&rcid=2216 5* 03Lpeer 5* (+138) 10/* First Release (3.0) Known blockers */  15:22:30 14[[07Releases/First Release Blockers14]]4 !10 02http://ovirt.org/w/index.php?diff=2149&oldid=2148&rcid=2217 5* 03Lpeer 5* (-138) 10/* First Release (3.0) Known blockers */  15:22:31 danken: any comment on both BZs? 15:22:59 I'm waiting for rvaknin's ack for the easier one 15:22:59 http://gerrit.ovirt.org/#change,1360 15:23:36 mgoldboi: ? 15:24:02 if it reconfigures vdsm on bootstrap, I'm happy 15:24:05 oschreib, danken There will not be an ack until iscsid bug will be fixed 15:24:12 danken: Rami (as Dave) faced https://bugzilla.redhat.com/show_bug.cgi?id=786174 which prevented them to test it 15:24:18 merd 15:24:22 rvaknin: why is this related? 15:24:42 oschreib: avoid this French in public rooms! 15:24:43 danken, because vdsmd fails to load and depends on iscsid 15:25:36 the bug is about vdsm not using ssl keys for spice 15:26:12 danken, I can't load vdsmd, how can I know what it really does??? 15:26:14 could you downgrade iscsid just to push vdsm? 15:26:19 no 15:26:37 danken, vdsmd requies iscsi-initiator-utils 14 15:26:40 why? it used to work with earlier versions. 15:26:42 danken: not in node (at least not without significant effort) 15:27:12 I'm speaking of verification of the patch, so that I feel confident to build vdsm 15:27:12 danken, and the only newer iscsi-initiator-utils doesn't work either 15:27:40 rvaknin: but we did not have this problem a week ago 15:28:18 danken, strange, ahh? can it somehow related to the configuration vdsm do in iscsid service? 15:28:51 no, dallen and olevy saw it with upstream vdsm 15:29:02 danken, ok 15:29:40 if i recall correctly, the bug that Federico found (and caused the require -14) was not 100% 15:29:52 danken, I've installed new Fedora 16 (from pxe server so it's quite clean installation) and the iscsid problem was there 15:29:52 bottom line, iiuc, we can't release vdsm in the current status. 15:30:12 we cannot release oVirt 15:30:13 danken, I don't give it a go 15:30:26 danken, for the patch I mean 15:30:37 oschreib: anyhow we couldn't validate the fix and we have another blocker - iscsid on our hands 15:30:52 danken: unfortunately, oVirt needs working vdsm (and all of its dependencies) 15:30:59 any thought on dealing with broken dependencies that might delay oVirt release? 15:30:59 rvaknin: could you remove iscsid from the NEEDED_SERVICES? 15:31:30 danken, I'll try this solution 15:31:30 just to see that my patch is not completely bogus? 15:31:59 rvaknin: also needs to comment 15:32:00 /sbin/service iscsid force-start 15:32:08 we all agree that https://bugzilla.redhat.com/show_bug.cgi?id=786174 is a blocker, right? 15:32:14 I'm wondering with this complex bootstrap, if there is a way to negotiate what the node is capable of instead of not registering because one of the needed services isn't available 15:32:17 jboggs: ping 15:32:24 rgolan, pong 15:32:38 if I'm not using any iscsi, couldn't we just mark the node as not having iscsi support if the dependencies on the node aren't met? 15:32:47 rharper: you're asking to make this complex bootstrap even more complex.... 15:32:50 no 15:33:01 I'm asking to use feature negotiation 15:33:16 feature negotiation is a huge RFE :) 15:33:57 that feature wasn't required when we only used rhel where we knew exactly what to expect, but it's going to get more important 15:34:10 true 15:34:57 danken: I'm adding https://bugzilla.redhat.com/show_bug.cgi?id=786174 as a blocker, can you handle it with the relevant teams? 15:34:59 anyway, we are diverging. I see that mchristie is on top of the issue 15:35:37 #info https://bugzilla.redhat.com/show_bug.cgi?id=786174 is a blocker for VDSM 15:36:43 oschreib: please add mchristi@redhat.com to minutes' CC. 15:37:00 I'll 15:37:04 ok. 15:37:12 anything else? 15:37:19 mgoldboi: log-collector maybe? 15:37:27 14[[07Releases/First Release Blockers14]]4 !10 02http://www.ovirt.org/w/index.php?diff=2150&oldid=2149&rcid=2218 5* 03Oschreib 5* (+126) 10/* First Release (3.0) Known blockers */  15:37:28 https://bugzilla.redhat.com/show_bug.cgi?id=786143 - logcollector is broken 15:37:45 are we sure log-collector is a gating item? 15:38:07 which will be very sad on the user support/debug side 15:38:10 ultimately the logs are there 15:38:19 sure log collector bundles them all up nicely 15:38:27 but if push comes to shove i think we could survive.. 15:38:52 rvaknin: any idea why NM still thinks it owns our bridge? could you attach ifcfg of bridge and nic? 15:38:58 if it's an easy fix and we are delaying the version anyhow - i think we should take care of it 15:39:09 since we already have multiple issues, and we will probably delay the release even more, I'll try to handle the log-collector personally 15:39:42 #action oschreib to try fix https://bugzilla.redhat.com/show_bug.cgi?id=786143 15:39:46 anything else? 15:40:02 i think this part of that bug "-if you had a previous run of it - it fails to collect:" is actually intentional 15:40:11 have we hit all the bugs in the list? 15:40:12 but sure the plugin bit should be fixed 15:40:27 sgordon: true 15:40:52 * rbergeron keeps seeing bugs being added 15:40:54 RFE would be for it to generate unique directories for each run, but i digress 15:41:08 danken, http://fpaste.org/JLIA/ 15:41:17 we should also work out new target date for release and communicate / update site etc 15:41:25 ok 15:41:34 we have 3 blockers ATM 15:41:49 + the log-collector one 15:42:01 oschreib did anyone look at the list I sent? 15:42:20 as a reminder as we fix these we need to add to http://www.ovirt.org/wiki/Release_Notes 15:42:39 acathrow: I did, and probably danken did as well. 15:42:39 danken, I'm not sure that if you omit "NM_CONTROLLED=yes" - the default is "no" 15:42:50 acathrow: but if we want to add blockers, I need bugs. 15:43:02 rvaknin, agreed, NM is generally greedy :) 15:44:41 14[[07Features/Design/DetailedHotlugNic14]]4 !N10 02http://www.ovirt.org/w/index.php?oldid=2151&rcid=2219 5* 03Ilvovsky 5* (+4450) 10Created page with " == Hotplug..." 15:45:06 danken: do we have a bz for the empty line mis-parsing in the ifcfg? 15:45:52 yes - https://bugzilla.redhat.com/show_bug.cgi?id=782348 15:45:59 * quaid wonders where that feature page template is coming from; looks like fedoraproject.org/wiki artifacts in there. 15:46:00 mgoldboi ^^ 15:47:15 oschreib: so you mentioned 1 of 3 blockers 15:47:26 14[[07Features/Design/DetailedHotlugNic14]]4 !10 02http://www.ovirt.org/w/index.php?diff=2152&oldid=2151&rcid=2220 5* 03Ovedo 5* (-32) 10 15:47:39 is it clear right now that we are currently 'no-go' for today? 15:47:48 sure 15:47:57 we mentioned all of them.... 15:47:59 even more important, can we pick a new target from this? 15:48:25 * quaid thought "+ the log-collector one" might have been the start of a list, but sees it was adding to a list? 15:48:43 1. iscsid, 2. engine- error in node registrations, 3. vdsm reconfigure thingy 15:49:32 quaid: the bottom line was that if the log-collector was the only gating item, I'm not sure we would delay the release. since we will delay it, it should be fixed. 15:50:46 danken: any special reason not to push https://bugzilla.redhat.com/show_bug.cgi?id=782348 as well to the build? 15:51:14 acathrow: quaid; yes we used it as a reference and changed it a little 15:51:46 lpeer: ok, thx; I'll add that to the wiki-gardening task list :) 15:52:05 I may drop again -- we still need a new target date.. 15:52:11 indeed 15:52:50 can we set one right now? 15:52:53 I suggest 09/02 (and do a quick overview in next week's meeting) 15:53:10 itamar,Hello, i am looking for any info on the logs i sent yesterday for the host being 'non responsive' problem 15:53:15 will you have a go/no-go meeting earlier in the week? 15:53:25 oschreib: we need to get an ETA on the iscsid bug (unless we want to pass on vdsm) 15:53:30 or will 1 day be enough to fix remaining items that come up in next week's meeting? 15:54:28 mgoldboi: I think one week should be enough for them, although we need their commitment. 15:54:54 quaid: we can do a seperate go/no go on the 07/02 15:57:17 mgoldboi: I do not mind backporting it into the ovirt build 15:57:34 danken: backporting what? 15:58:29 oschreib: https://bugzilla.redhat.com/show_bug.cgi?id=782348 15:58:37 oschreib: handling an emply lin ein ifcgf 15:58:41 oh, ok 15:58:48 ok, we're about out of meeting time today 15:59:10 did we have anything more on release status? 15:59:16 so what's the conclusions? 16:00:19 :( 16:00:44 oschreib: think the status is no-go 16:00:57 oschreib: and we'll sync again on tuesday 16:01:15 no target date? 16:01:16 with new target of 9-feb 16:01:44 ok, so I'll send the mail to users/board/announce. 16:03:31 anything more for today? 16:04:02 outside of the release discussion? 16:04:11 nothing we need to discuss 16:04:23 I'm looking for workshop space in Beijing 16:04:41 #info retargeting oVirt's first release to Feb 9th 16:04:51 #info go/no go meeting on Feb 7th 16:06:18 rvaknin: dropping the NM_CONTROLLED should be fine - it is for all the other networks, isn't it? 16:06:38 #agreed oVirt first release now 9 Feb 16:06:49 #agreed go/no-go meeting for first release 7 Feb 16:07:00 * quaid just using that convention to make the decision stand-out in the meeting notes 16:07:06 ok, since we're over the hour 16:07:09 good to know 16:07:10 and have nothing more 16:07:16 I'll close in 10 seconds 16:07:19 * quaid counts to himself 16:07:35 #endmeeting