17:01:58 #startmeeting VMwareAPI 17:01:59 Meeting started Wed Oct 9 17:01:58 2013 UTC and is due to finish in 60 minutes. The chair is tjones. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:01:59 tjones: Error: "startmeeting" is not a valid command. 17:02:00 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:02:03 The meeting name has been set to 'vmwareapi' 17:03:10 alright - i am assuming we suspend BP discussion until after hong kong? 17:03:40 listens for comments 17:04:17 from the thunder of comments - i assume agreement ;-) 17:04:21 tjones: i guess we could discuss them a icehouse is open. 17:04:29 sorry, i am slow today 17:04:44 ok lets spend some time there then 17:04:52 #topic blueprints 17:04:53 tjones: Error: "topic" is not a valid command. 17:05:02 argh 17:05:44 anyone have a favorite BP for discussion? #link https://blueprints.launchpad.net/openstack?searchtext=vmwa 17:05:53 #undo 17:05:54 Removing item from minutes: 17:05:54 tjones: Error: "undo" is not a valid command. 17:06:00 #link https://blueprints.launchpad.net/openstack?searchtext=vmware 17:06:01 tjones: Error: "link" is not a valid command. 17:06:28 im getting all sorts of errors from virtbot on my commands btw…. ignoring it 17:06:29 mmm… let's just move forward without commands. not sure what is going on 17:06:55 I wanted to ask if there is already one about glance supporting vm templates 17:07:33 vuil: i think https://blueprints.launchpad.net/nova/+spec/deploy-vcenter-templates-from-vmware-nova-driver 17:07:37 vuil: was talking to gary about this 17:07:39 there is a vmware specific one 17:07:39 *looking* - i thought there was - here it is https://blueprints.launchpad.net/glance/+spec/hypervisor-templates-as-glance-images 17:07:47 looks like we have a couple 17:08:06 i think that we both worked on the first patch set that was posted by Kiran. It was stopped due to the H feature freeze. 17:08:08 I think we may want to step back and look at the whole images issues more generically. Arvind is working on putting together a PRD around this. 17:08:27 ok good - when do you expect that to be available? 17:08:44 danwent: great. 17:09:09 tjones: initial proposal should be before the summit 17:09:29 so we at least have some idea of what we want to do moving forward 17:09:35 danwent: there are a few BP having to do with image handling 17:09:50 tjones: i am also trying to investigate ways that we can improve our caching. this may be related to what dan has mentioned above. so it may all be under the same umbrella 17:09:55 want to make sure we have a holistic approach that covers things like cache management, leveraging array capabilities, etc. 17:10:19 and multiple v-centers, which the template approach doesn't handle well 17:10:48 ok great. Besides the large area of image handling, are there other BP we need to discuss? 17:11:04 we need to take another look at the ovf support as an option for something like multi-vc 17:11:52 danwent - would arvind be covering that as well? 17:12:21 tjones: yes, arvind will be point person for organizing discussions around all things images 17:12:26 excellent 17:13:04 other BP topics? 17:13:47 ok moving on to bugs (not using commands). Favorite bugs for discussion? 17:14:38 today i found one that i need to open 17:15:13 here's the link to today's report http://paste.openstack.org/show/48148/ 17:15:25 when one restarts the nova compute service the vms on the esx are not matched with this registered in the database 17:15:27 garyk: go ahead and then we can talk about the report 17:15:49 just a restart - not a re-stack? 17:16:21 just a restart. a restack is ok as the db is empty 17:16:47 and this is in havana? 17:16:48 i'll post it later with all of the information 17:16:57 it is in all versions 17:17:01 ugh 17:17:14 i actually discovered it with the backports that i did for stable/grizzly 17:19:23 ok that's a bad one - please pass on the # when you have it. Now going on to our current bugs - of the 15 open - 12 need +1 reviews. Non-core folks, please spend some cycles on review this week 17:19:34 tjones: it would be nice if the link that you posted could have the title of the bug. that would make it easier to understand 17:19:57 sure - i can add that. the 2nd section does show the title as well 17:20:06 that would be great 17:20:19 they are the same bugs ordered differently 17:20:31 regarding https://bugs.launchpad.net/nova/+bug/1227825 - datastore selction. 17:20:34 Launchpad bug 1227825 in openstack-vmwareapi-team "datastore selection bug - fills first disk only" [Critical,In progress] 17:21:52 i think that this was regarded as a feature and that is why is was not part of the release candidate. 17:22:09 i thought russellb was considering it? 17:22:35 i also thought the same. 17:22:51 it has havana-rc-potential tag 17:23:19 thanks for picking it up btw - we have 3 +1 on it 17:23:20 ok, great. 17:23:56 could use a few more eyes but i think it's ready to go if we get a rc opening 17:24:28 any other bugs for discussion? 17:26:32 we do have another rc potential bug - it's the grizzly regression https://bugs.launchpad.net/nova/+bug/1193980 17:26:35 Launchpad bug 1193980 in nova "Regression: Cinder Volumes "unable to find iscsi target" for VMware instances" [High,In progress] 17:27:24 that has 5 +1 and is ready for core reviewers 17:27:34 yeah, that is a critical one 17:28:09 garyk - the other one that i have as critical for havana is https://review.openstack.org/#/c/41977/ 17:28:54 i backported that the the rc2 candidae. it was −2'ed as the version is not yet open. at the moment it is not clear when russellb will open this. 17:29:21 looks like it merged! 17:29:33 ah - into master 17:29:55 ok - so those are the 3 rc2 candidate bugs we have. anything else? 17:30:38 ok lets take a look at bug triage then http://goo.gl/pTcDG 17:30:56 i think that https://code.launchpad.net/bugs/1235022 could be a candidate (there is a work around though) 17:30:57 Launchpad bug 1235022 in nova "VMware: errors booting from volume via Horizon" [High,In progress] 17:31:18 bascially if a volume that is not vda is used then it causes all kinds of havoc. 17:31:44 is that what ryan ran into during stress testing? 17:32:08 what's the work around? 17:33:24 tjones: not exactly sure how ryna found it 17:33:54 the work around is to do the attachment via the cli and not horizon. the user should also select vda as the volume and sdb 17:35:14 so ugly but could be documented. Do we think this is a common use case - boot from volume? 17:35:37 danwent: that's probably a question for you 17:36:04 it is something that we should definitley back port. this may not make the rc 17:36:16 yeah, i agree with gary 17:36:24 we should make sure to backport it, but it doesn't seem like a blocker for RC 17:36:38 ok - you have already added backport to it. so i'll leave it alone 17:36:48 tjones: we need to be collecting items like this to include in "release notes" 17:37:06 if we end up having a chance to slip it into an RC, great, but not a huge deal 17:37:11 danwent: do we have a release note tag for bugs 17:37:35 tjones: that's a good question for those in the community. i'm not aware of a standard within openstack 17:38:04 ok - i'll add it to my list of actions (since i can't use the action command) 17:39:08 back to triage - we still have 2 waiting for responses and 1 (speanw 40 vms) that is undecided. http://goo.gl/pTcDG Im guessing that's a high prio 17:39:34 i can't set the prio 17:39:42 i think that this one is under investigation. it is concerning 17:40:18 vuil: i think you are looking at this one? 17:40:21 I just confirmed the last one and assigned to me. We are looking at gathering more instrumentation wrt to NFS locking behavior 17:40:22 tjones: i think taht you need to join thenova bugs group and then you can set. let me try and dig up the link 17:40:53 garyk: found it 17:41:03 tjones: https://launchpad.net/~nova-bugs 17:41:07 https://bugs.launchpad.net/nova/+bug/1230047 I am talking about 17:41:11 Launchpad bug 1230047 in nova "VMware: spawning large amounts of VMs sometimes causes errors" [Undecided,Confirmed] 17:42:08 ok i marked it high and left the milestone alone 17:42:25 ready to talk backports or any other bug discussion? 17:43:08 i think that we are at a very importnat stage where testing should be our top priority at the moment 17:43:38 the more coverage we get the better our understanding of the rc canditae will be 17:44:12 so we talked about having a day where we all try to break the rc candidate. create a simple test plan and we all bang on it. lets talk a little more about how we would go about doing this 17:45:08 (thanks to garyk our backport situation is good and we can skip that discussion) 17:45:14 ideas? 17:45:17 :) 17:45:44 tjones: maybe we can decide on test day 17:46:13 that is, we all hang out on irc and discuss tests, problems etc. 17:46:37 an example for this is a fedora test day. there are a lit of instructures for people and then we can follow those 17:47:08 that works, but i'd like it to be slightly more scripted… oh ok i have not seen that. can you send out a link? 17:47:10 i guess we can take it offline to draw up the outline and hopefully the community will also engage 17:47:24 i am looking. give me a sec 17:47:28 ok lets you and i discuss offline - action ;-) 17:47:54 tjones: here is an example - http://fedoraproject.org/wiki/Test_Day:2012-09-18_OpenStack 17:48:21 we can send something out. but i would like to do this next week - so if we find something catstro we may be able to fix it 17:49:02 ok. sounds good 17:49:05 garyk: this is great! 17:49:21 ok anything else we need to discuss - now we are at "open discussion" 17:49:58 i just want to reiterate and ask non-core folks to spend some cycles in review 17:51:16 *listening* 17:52:11 going once…. 17:52:46 i'll be happy to get to sleep a few min earlier tonight ;) 17:53:10 ok with that we will end. :-) thanks all for joining. "talk" to you next week 17:53:13 #endmeeting