16:02:32 #startmeeting #hyper-v 16:02:33 Meeting started Tue Aug 13 16:02:32 2013 UTC and is due to finish in 60 minutes. The chair is primeministerp. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:02:34 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:02:36 The meeting name has been set to '_hyper_v' 16:02:39 Hello 16:02:41 hi everyone 16:02:49 hola 16:02:49 hi 16:02:52 #topic blueprint review 16:02:54 Hi ROb! 16:03:00 Hi all 16:03:19 #link https://blueprints.launchpad.net/openstack?searchtext=hyper-v 16:03:32 thanks for link - I was just gonna ask... 16:03:33 Hey Iben :) 16:03:57 so everyone I'd like to start by doing a review of the proposed blueprints and seeing where we currently stand 16:04:11 alexpilotti: ready? 16:04:25 yep 16:04:30 pnavarro: do you also have blueprints for cinder? 16:04:41 hi guys 16:05:05 #link https://blueprints.launchpad.net/cinder/+spec/windows-storage-driver-extended 16:05:29 great 16:05:38 let's start w/ nova bits first 16:05:45 cool 16:05:56 so we have a bunch of bps out for review 16:06:06 alexpilotti: let's start w/ wmiv2 16:06:19 #link https://review.openstack.org/#/c/38160/ 16:06:27 alexpilotti: thank you 16:06:30 #link https://review.openstack.org/#/c/39195/ 16:06:56 and #link https://review.openstack.org/#/c/39194/ 16:07:05 it was a big one, so I had to split in in 3 16:07:43 so basically those are waiting for core reviewers to look at them 16:07:43 what's not so good, is that those patches are standing there since a long time waiting for review 16:07:51 yep 16:07:53 yeah 16:07:53 only the first one (last link) got in 16:07:53 i see 16:08:06 that's a problem 16:08:18 the others have some +1, but no additional review love :-) 16:08:29 alexpilotti, do you think I should wait for these ones to be aproved before starting the ephemeral one? 16:08:34 pnavarro: no 16:08:42 pnavarro: no 16:08:55 we have to put them on cascade, which is a PINTA 16:08:58 pnavarro: work off of those patches if possible 16:09:11 we have already another 2 out there, which depend on the WMI ones 16:09:13 last patch change was 3 days ago? 16:09:25 or is that just last comment? 16:09:37 iben: which are you looking at 16:10:00 it says updated aug 9th 16:10:00 iben: yeah, Pedro put a +1 16:10:55 this one: #link https://review.openstack.org/#/c/39195/ 16:11:02 was uploaded on July 30th 16:11:21 only a couple of small nits changed in the meantime 16:11:46 i see - when it says updated that's due to the build status and other comments. 16:11:50 this one: https://review.openstack.org/#/c/38160/ 16:11:53 Aug 1st 16:11:54 dansmith: ping 16:12:09 primeministerp: ack 16:12:14 hey dan 16:12:26 question about the queue on reviews 16:12:39 we have some pretty critical bits waiting in the pipeline for some time now 16:12:56 dansmith: is the general delay due to just backlog and volume? 16:12:58 dansmith: hi! 16:13:11 primeministerp: yes, and priorities 16:13:49 dansmith: as you know we have a bunch of patches out there since some time, even > 10 days 16:14:14 dansmith: is there any way to get some help with adding some priority to those mentioned above 16:14:21 dansmith: do think we can have a rough idea about when you guys will have time to review? 16:14:24 alexpilotti: yep, lots of people do.. core reviewers are doing a whole lot of things, reviewing tons of changes and, of course, working on blueprints of their own 16:14:41 dansmith: we fully understand 16:14:42 are these bugs? 16:14:52 dansmith: I know that you are swamped, so I don't bug you too much :-) 16:15:00 dansmith: nope, bps 16:15:06 oh, the namespace thing 16:15:34 yep, splitter in 3 patches as you requested, etc etc 16:15:38 I'm not keen on rewarding this behavior, but I'll try to take another pass through those today 16:15:40 *splitted 16:15:50 dansmith: thank you 16:16:00 dansmith: tx! 16:16:31 dansmith: if there's a better method of "behavior" you prefer going forward we'll make note 16:17:06 dansmith: we appreciate the help everyone has given to this point 16:17:24 primeministerp: well, I think getting pinged for reviews is getting old for a lot of folks, but you have been fairly patient, so .. :) 16:17:46 dansmith: you're very gracious sir 16:17:48 ok 16:17:55 thanks again dansmith 16:17:58 yep 16:18:13 alexpilotti: can we move on to the next? 16:18:17 sure 16:18:28 VHDX 16:18:37 #Link https://review.openstack.org/#/c/40076/ 16:18:51 same story 16:19:06 alexpilotti: yep, no need to beat it in 16:19:16 sure 16:19:23 alexpilotti: we've got dansmith attention, we'll take it as such 16:19:35 I'm not complaining, just pointing it out :-) 16:19:45 alexpilotti: yep 16:19:49 and also: 16:19:52 #link https://review.openstack.org/#/c/38791/ 16:19:54 so - all these blueprints are supposed to show delivery status of "needs code review" for Dan Smith to look at it? 16:20:09 iben for anybody 16:20:16 iben: not necessarily dan 16:20:23 sure 16:20:40 about the satus - wmi-v2 shows "good progress" 16:20:43 iben: but core reviewers are the ones which are needed to get it merged 16:20:52 does that mean it's not ready for code review yet? 16:21:02 are you sure it's not the Neutron one? 16:21:06 yes 16:21:13 I' 16:21:18 yep, I still have to send that one in 16:21:22 I'm looking here: 16:21:25 #link - https://blueprints.launchpad.net/openstack?searchtext=hyper-v 16:21:37 iben: those are the active blueprints 16:21:44 iben: not all has been submitted yet 16:21:46 this is the nova one: #link https://blueprints.launchpad.net/nova/+spec/hyper-v-wmi-v2 16:22:11 iben: we're trying to go through the list to see what still needs to be completed for our H3 milestone 16:22:27 so behind each blueprint are one or more changes that need to be approved. 16:22:31 got it! THanks! 16:22:46 ok let's continue 16:23:07 obvioulsy same for dynamic memory 16:23:14 yep 16:23:18 rdp console 16:23:35 which by the way got committed by Robert, one of the new Cloudbase guys 16:23:49 alexpilotti: great, is here in the meeting? 16:23:54 nope 16:23:55 er is he 16:23:56 o 16:24:18 ociuhandu: is he still @ the office? 16:24:27 it's 7.24 PM here 16:24:32 alexpilotti: i understand 16:24:41 alexpilotti: yes 16:24:51 going to make sure he's in 16:24:52 probably just out at lunch now... 16:24:57 so in terms of the remaining that are labled "not started" 16:25:32 some of them have to be move to Havana, as the review process is slowing everything down 16:25:39 alexpilotti: ok 16:25:56 and everything needs to be done by Sept 4th 16:26:17 of those remaining are we capible of delivering for that deadline? 16:26:27 lol 16:26:39 fair enough 16:26:42 ;) 16:26:54 alexpilotti: then let's talk which don't make it 16:27:05 it really depends on the review process 16:27:29 for example now I'm on the RDP one 16:27:33 alexpilotti: ok, then let' postpone this discussion until next week to see how the rest of the week pans out 16:27:49 with the first one blocked due to an API that still has to be contributed :-( 16:28:07 alexpilotti: hopefully we'll make some progress through the review pipeline 16:28:28 pnavarro: ok let's talk cinder 16:28:47 ok 16:29:03 I saw there was progress over the weekend 16:29:04 fortunately my blueprint is High priority 16:29:07 sorry, I got pinged about the same topic on the Nova list :-) 16:29:31 pnavarro: how far along are you now? 16:29:34 pnavarro: except the ephemeral storage one :-) 16:29:43 the bp is in review 16:29:43 I saw there was patch you submitted 16:29:46 that's also low priority, assigned to you 16:30:12 sorry guys, let's finish the previous topic first, if you don't mind 16:30:14 the thing is I'll be on vacation next week 16:30:21 alexpilotti: sorry 16:30:28 pnavarro: let's hold this for a sec 16:30:32 alexpilotti: continue please 16:30:33 ok 16:30:59 we got out on Nova: AMI, Rescue, 16:31:04 Progress update 16:31:19 and phemeral 16:31:24 *ephemeral 16:31:45 IMO AMI, rescue and progress update are totally secondary 16:31:53 alexpilotti: I agree 16:32:02 ephemeral needs to get in 16:32:25 it depends on VHDX 16:32:37 which in turn depends on WMIV2 16:32:59 alexpilotti: right, which is why i said let's see how the week progresses in terms of the review process 16:33:22 alexpilotti: that is unless you have code ready to submit for those as well? 16:33:33 russellb just pinged me to make sure we decide what to do with the ones which are not started :-) 16:33:51 well, it really depends on the RDP console one 16:34:05 alexpilotti: in terms of review time? 16:34:25 yep, as I expect a bit of ping-pong there, to say so :-) 16:35:07 alexpilotti: so priority then should the rdp console pass quickly, would be ephemeral 16:35:08 so my opinion is to focus 100% on what we need absolutely to get in in this release 16:35:22 ephemeral can be done in parallel, it's an easy one 16:35:30 but it depends on VHDX 16:35:39 right vhdx is waiting in the queue 16:36:15 so, if VHDX gets in w/o substantial changes (I don't expect any, frankly) we can get ephemeral done in a breeze 16:36:41 i would assume the rescue is a push 16:36:42 I just marked it as started 16:37:18 alexpilotti: that last comment was more of question 16:37:19 -1 on rescue - why do we need that? is it a core feature? 16:37:21 rescue is also depending on all of the above. A fairly easy one, but I don't want to overlay too many things 16:37:43 alexpilotti: and progress update? 16:37:44 guys, about ephemeral I won't be able to work on it before 26/08 - > too late? 16:37:46 iben: it's just used to rescue images, KVM, XCP, have it 16:37:58 okay - thanks 16:38:03 pnavarro: on 26/8 it should be in 16:38:17 with the code freeze on the 4th... 16:38:23 pnavarro: we'll have to get it in prior 16:39:01 so, maybe it's better to be reassigned to other person, it's my vacation week 16:39:24 pnavarro: lucky you :-) 16:39:24 pnavarro: enjoy your holiday please 16:40:08 so given that pnavarro is unavailable do we need to push it or can you absorb it? 16:40:22 pnavarro: we'll assign to assign internally 16:40:27 agh 16:40:39 we'll have to assign it internally 16:40:45 page fault :-) 16:40:54 hehe 16:41:16 as you said it's not difficult 16:41:36 it's not, luckily, couple of days should be done 16:41:44 ok 16:41:58 so we all agree 16:41:59 pnavarro: it's way more useful if you can help us on the Cinder bits 16:42:10 as you are already doing :-) 16:42:45 perfect then 16:43:05 ok let's continue because we have a lot more to get to 16:43:17 alexpilotti: of the remaining what is a definate push 16:43:18 cool 16:43:25 hyper-v-agent 16:43:25 that's it IMO 16:43:31 for Nova 16:43:42 hyper-v-ami 16:43:49 yep, hyper-v-agent for ceilometer (WIP) 16:43:58 and what about remotefx 16:44:04 As I was saying IMO AMI is almost useless at this stage 16:44:18 alexpilotti: then let's push that to I 16:44:23 remotefx is basically done, I'm waiting for some hardware to test it 16:44:41 wow remotefx ! 16:45:03 yeah, it's very cool 16:45:18 ok then we are clear on what need to be done 16:45:28 I need to put a couple of DX11 cards on a server and make sure it works :-) 16:46:13 alexpilotti: can we move on to the next topic? 16:46:21 so to recap, my non-wish-list is: 16:46:49 hyper-v-ami, hyper-v-agent 16:46:50 AMI, progress update, rescue 16:47:34 you're not including hyper-v-agent? 16:47:56 no, I'd like to get that done :-) 16:48:01 ok 16:48:09 then I'm good with that as well 16:48:20 pnavarro: sound good? 16:48:35 we'll move those 3 to I 16:48:48 it's ok for me 16:48:57 great 16:49:00 let's move on then 16:49:09 #puppet-openstack_hyper_v 16:49:18 #topic puppet-openstack_hyper_v 16:49:21 luis_fdez: ping 16:49:23 uey 16:49:27 luis_fdez: where does your refactor stand 16:49:35 i've been waiting to merge the last request 16:49:53 luis_fdez: per your comments asking not to 16:50:23 you can merge it with the last commit I pushed the last day 16:50:36 the additional improvements are not yet ready 16:50:44 luis_fdez: ok, i'll do that after the meeting 16:50:50 now I'm alswo working on integrating the registry module 16:50:55 for creating the nova-compute service 16:51:03 and call nova compute 16:51:07 luis_fdez: btw I've added a puppetfile w/ the pprerequests 16:51:17 ah ok, I'll rebase it 16:51:36 I want to find the best way of calling the python nova-compute from the windows service 16:51:39 luis_fdez: can you also put an example node config in tests 16:51:59 ok 16:51:59 luis_fdez: I still need to try, and now the format has changed dramatically 16:52:11 luis_fdez: i'll be in a position to test hopefully tomorrow 16:52:40 ok, I think this week I'll be able to work hard on the puppet side 16:52:45 last week I took it off 16:52:56 luis_fdez: no worries, we're making progress 16:53:24 also the vcs module is pending 16:53:28 luis_fdez: i've been dealing w/ all the other puppet bits for the infra here, working my way to the hyper-v nodes 16:53:39 luis_fdez: are you modifying upstream I assume 16:53:56 vcs? I didn't touch it yet 16:54:03 luis_fdez: is that next? 16:54:13 i think so... suggestions? 16:54:27 luis_fdez: well if we're going to do a from source install it's necesary 16:54:39 luis_fdez: and i'd rather just add windows support to the upstream modules 16:54:49 i agree 16:54:55 luis_fdez: this does mean that I'll have to break out additional modules 16:55:12 luis_fdez: and our dependancy list is going to be massive 16:55:42 luis_fdez: however we knew it would eventually be necessary 16:56:07 I'll try to make the current version as stable as possible and then jump to the next approach from source 16:56:17 luis_fdez: I have some comments on that 16:56:25 luis_fdez: under the next topic 16:56:28 ok 16:56:30 :) 16:56:35 we can move if you want 16:56:37 #topic hyper-v ci 16:57:03 so in regards to that piece 16:57:06 from source 16:57:14 as well as other pertaining to ci 16:57:31 there is additional help currently ramping up 16:57:41 yeah - more resources! 16:57:41 to assist with the puppet work 16:57:49 yes 16:58:10 perfect :) 16:58:10 I spent the last week giving them an overview 16:58:40 there will be more to come, as it's going to take time for the others to get ramped up 16:58:55 but eventually you'll see them here and on the other channels as well 16:59:03 ok 16:59:39 our goal is to have everything up and fully funtionaly before the april timeframe 17:00:03 however there will be substancial rampup in the immediate weeks 17:00:42 that's about all I have 17:00:42 help is welcomed 17:00:48 for now 17:01:04 does anyone have anything else to add 17:01:06 so we can end it 17:01:10 we're already running over 17:01:26 just thanks for all the hard work 17:01:31 this is gonna be great! 17:01:37 if not i'm ending the meeting, we can continue on the hyper-v channel 17:01:40 tx! 17:01:42 #endmeeting