16:01:32 <primeministerp> #startmeeting hyper-v 16:01:32 <openstack> Meeting started Tue Nov 27 16:01:32 2012 UTC. The chair is primeministerp. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:33 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:01:36 <openstack> The meeting name has been set to 'hyper_v' 16:01:40 <primeministerp> hi all 16:01:51 <EmilienM_> primeministerp: hi 16:01:56 <primeministerp> EmilienM_: how are things 16:02:11 <EmilienM_> primeministerp: fine, thx 16:02:26 <primeministerp> figure we'd start with a status update from the different folks 16:02:52 <primeministerp> #topic updates 16:03:21 <primeministerp> cloud-init 16:03:43 <primeministerp> alexpilotti: : would like to give an update on the status of the cloud-init code 16:04:01 <alexpilotti> hi guys! 16:04:25 <alexpilotti> after doing cloud-init for Linux, we moved to cloud-init for Windows 16:05:06 <alexpilotti> we worked on public key atuthentication for Windows, but unfortunately there still a blocking issue there, I'm waiting an answer from MS for that. 16:05:17 <primeministerp> alexpilotti: all the code for hyper-v support of linux guests in upstream though correct 16:05:30 <alexpilotti> beside that, we are porting the Python cloud-init to Windows 16:05:38 <primeministerp> alexpilotti: i'll follow up again in a day if i don't hear back from anyone 16:05:48 <alexpilotti> as it's quite modular 16:06:11 <alexpilotti> which means that we can treat wWindows as any other "Linux distro" in cloud-init ;-) 16:06:29 <alexpilotti> the big issue there, is that configdrive 16:06:30 <primeministerp> alexpilotti: we prefer just "distro" 16:06:33 <primeministerp> ;) 16:06:37 <alexpilotti> primeministerp: lol 16:06:41 <josecastroleon> hi all :) 16:06:45 <primeministerp> hey josecastroleon 16:07:13 <primeministerp> pnavarro: would you like to give a status of the cinder work 16:07:17 <pnavarro> sure 16:07:21 <alexpilotti> the "big" issue with ConfigDrive is that it mounts a ISO image as a raw physical disk 16:07:37 <primeministerp> pnavarro: you can wait unitl alexpilotti is done 16:07:43 <primeministerp> pnavarro: I thought he was 16:07:44 <alexpilotti> so, we pull it out with raw Win32 Api calls 16:07:55 <pnavarro> no problem 16:07:58 <primeministerp> alexpilotti: sorry 16:08:11 <alexpilotti> and that's done, I expect teh full cloud-init for Windows to be done this week 16:08:26 <primeministerp> alexpilotti: execellent 16:08:28 <alexpilotti> for authentication, we'll inject the adminpass 16:08:45 <alexpilotti> teh one passed by nova boot, waiting for those answers from MS 16:08:55 <alexpilotti> for the SSH publick key auth in RDP 16:09:22 <alexpilotti> but all the rest, including networking, script execution etc is not an issue 16:09:31 <alexpilotti> that's it :-) 16:09:34 <primeministerp> nice 16:09:36 <primeministerp> thx 16:09:41 <pnavarro> ok, from my side 16:09:43 <pnavarro> nova_to_cinder code is review process #link https://review.openstack.org/#/c/16843/ 16:09:55 <primeministerp> awesome 16:09:59 <pnavarro> thanks to alexpilotti and EmilienM_ for helping to set my live migraiton environment 16:10:02 <primeministerp> pnavarro: does that include the move to the wmiv2? 16:10:08 <pnavarro> yes, primeministerp 16:10:12 <primeministerp> perfecto! 16:10:21 <primeministerp> EmilienM_: thanks for helping out! 16:10:33 <alexpilotti> pnavarro: is the above link updated? 16:10:37 <EmilienM_> primeministerp: yw :) thank's to Pedro too 16:10:42 <pnavarro> the review is blocked until this night I'll have some time to fix some issues 16:10:43 <alexpilotti> pnavarro: I see a pylynt error: http://logs.openstack.org/16843/4/check/gate-nova-pylint/3593/console.html.gz 16:11:15 <alexpilotti> nova/virt/hyperv/volumeutils.py:46: [E1101, VolumeUtils.execute] Module 'eventlet.green.subprocess' has no 'STDOUT' member 16:11:27 <pnavarro> I'll fix that later, today I was a bit busy 16:11:27 <primeministerp> pnavarro: yep i see that too 16:11:34 <primeministerp> pnavarro: np 16:12:02 <pnavarro> alexpilotti: I can't fix that, nova.utils is using the same syntax 16:13:10 <alexpilotti> pnavarro: that's interesting :-) 16:13:24 <primeministerp> i'm missing something? 16:13:47 <pnavarro> alexpilotti: yeah.. very interesting 16:14:31 <primeministerp> so what's broken 16:14:35 <primeministerp> something in pylint 16:14:37 <alexpilotti> pnavarro: I'm going to take a look at this 16:14:55 <pnavarro> thanks alexpilotti 16:15:42 <primeministerp> ok 16:15:56 <alexpilotti> pnavarro: add: # pylint: disable=E1101 16:16:11 <pnavarro> ok 16:16:15 <alexpilotti> pnavarro: look here, line 186 https://github.com/openstack/nova/blob/master/nova/utils.py 16:16:50 <alexpilotti> pnavarro: just 2 lines above my patch :-D 16:17:04 <pnavarro> alexpilotti: ok ! I didn't realize about that 16:17:28 <primeministerp> awesome 16:17:31 <primeministerp> realtime even 16:17:33 <primeministerp> ok 16:17:42 <primeministerp> alexpilotti: quantum update? 16:17:49 <primeministerp> or ociuhandu rdp? 16:18:19 <alexpilotti> working on it, confirmed release for G-2. Wil need to sync with Pedro this week :-) 16:18:26 <primeministerp> ok 16:18:32 <primeministerp> sound good 16:18:42 <primeministerp> ociuhandu: want to discuss anything on rdp? 16:18:53 <ociuhandu> primeministerp: i'm now looking into the ways to pass the required parameters to wsgate 16:19:28 <primeministerp> ok 16:19:39 <primeministerp> josecastroleon: anything to add from CERN? 16:19:51 <alexpilotti> luis_fdez: thnks for the CPU patch! https://review.openstack.org/#/c/16796/ 16:19:57 <primeministerp> o yes 16:19:57 <josecastroleon> sure 16:20:02 <josecastroleon> plenty of things 16:20:02 <primeministerp> thanks luis_fdez 16:20:06 <luis_fdez> alexpilotti: :) ... 16:20:09 <pnavarro> CERN guys, will you assist to openstack Paris event next thursday? 16:20:18 <josecastroleon> Jan is going there 16:20:35 <josecastroleon> I saw that AD patch is committed in G-1 16:20:46 <pnavarro> ok 16:21:05 <josecastroleon> we have just found a problem on LDAP when the output is more than 1k 16:21:13 <pnavarro> EmilienM_ will be talking about quantum, let's see if I can convince him to add some slide about Hyper-V 16:21:31 <EmilienM_> pnavarro: ahah 16:21:37 <primeministerp> EmilienM_: you shoud 16:21:38 <alexpilotti> josecastroleon: I love your scalability test environment ;-) 16:21:39 <primeministerp> er should 16:21:43 <josecastroleon> cool 16:21:46 <EmilienM_> ok 16:21:51 <primeministerp> alexpilotti: which one is that? 16:21:54 <ociuhandu> sorry guys, have to run now, will be back online a bit later. Have a great one 16:21:59 <josecastroleon> my fault 16:22:00 <EmilienM_> let's continue the conversation on #openstack-hyper-v after meeting 16:22:00 <primeministerp> ociuhandu: ok thx 16:22:03 <alexpilotti> primeministerp: CERN itself :-D 16:22:07 <EmilienM_> to define the content 16:22:08 <primeministerp> haha 16:22:45 <alexpilotti> pnavarro: when would you like to meet for Quantum? 16:22:58 <primeministerp> josecastroleon: were the ad patches for rold/tenant id mapping? 16:23:09 <josecastroleon> no we have more than 2600 VMs on the test environment 16:23:10 <pnavarro> what about tomorrow night? 16:23:21 <EmilienM_> josecastroleon: wow 16:23:29 <alexpilotti> josecastroleon: cool! :-) 16:23:45 <alexpilotti> josecastroleon: what VM density do you have on Hyper-V? 16:23:52 <pnavarro> alexpilotti: I'll send you a meeting request later 16:24:00 <alexpilotti> pnavarro: ok! 16:24:01 <sagar_nikam> josecastroleon:how many WIndows 2012 16:24:10 <josecastroleon> we have only a few 16:24:14 <josecastroleon> not so many 16:24:40 <primeministerp> josecastroleon: are the hyper-v nodes integrated in w /the rest? 16:24:46 <josecastroleon> linux guys have overtaken us 16:24:47 <alexpilotti> josecastroleon: would you be interested in helping with the tests of Cloud-init for Windows? 16:24:52 <josecastroleon> sure 16:24:56 <primeministerp> josecastroleon: hahah 16:24:59 <alexpilotti> josecastroleon: tx! 16:25:57 <alexpilotti> josecastroleon: here's the solution to fight back: run this on their controller: nova list | awk '{if (NR > 3 && $2 != "") {system("nova delete " $2);}}' :-D 16:26:35 <primeministerp> ok 16:26:35 <josecastroleon> ups 16:26:45 <primeministerp> anything else on updates? 16:26:51 <primeministerp> want to move to ci discussion 16:26:58 <alexpilotti> ok 16:27:04 <primeministerp> #topic ci discussion 16:27:10 <primeministerp> mordred: ping 16:27:22 <primeministerp> jeblair: ping 16:28:02 <primeministerp> was hoping to get some involvement from the ci guys 16:28:19 <jeblair> primeministerp: hi, i'm dealing with an operational issue at the moment 16:28:34 <alexpilotti> jeblair: hi! 16:28:37 <primeministerp> jeblair: np was hoping to get some input/insight into the ci bits 16:28:47 <primeministerp> jeblair: we're looking to stand up a replica next week 16:28:52 <primeministerp> jeblair: on real hardware 16:29:00 <primeministerp> jeblair: tied into zuul 16:29:13 <primeministerp> then start integrating in hyper-v nodes 16:29:23 <primeministerp> and commiting that upstream 16:29:57 <primeministerp> jeblair: i was looking at the existing openstack-ci project and it looks like it's been refactored out into different sub projects 16:30:14 <jeblair> primeministerp: great, i think you were planning on commenting as a third-party tester, like: http://ci.openstack.org/third_party.html 16:30:16 <primeministerp> jeblair: I there somewhere where I can find the puppetmaster manifests being used, o 16:30:23 <pnavarro> guys, I have to go 16:30:37 <alexpilotti> pnavarro: bye!! 16:30:46 <primeministerp> jeblair: yes, however I'm going to have to spin up an existing setup first 16:30:51 <mordred> primeministerp: hey 16:30:57 <jeblair> primeministerp: everything is in http://github.com/openstack/openstack-ci-puppet 16:31:11 <primeministerp> ahh 16:31:15 <jeblair> primeministerp: there should be modules in there you can use to manage zuul servers, jenkins, etc... 16:31:15 <primeministerp> that's there it moved too 16:31:20 <primeministerp> ok 16:31:21 <jeblair> primeministerp: (hasn't moved) 16:31:34 <primeministerp> so our plan is to use that and spin up a replica 16:31:39 <mordred> ++ 16:31:47 <primeministerp> hook the replica into the event stream 16:32:00 <jeblair> (i'm going to turn this over to mordred, so i can get back to tracking down the issue we're having, sorry) 16:32:08 <mordred> primeministerp: I'm slammed today, but will be more than happy to spend some quality time with you tomorrow or next week walking through things 16:32:10 <primeministerp> jeblair: that's exactly what we are going to do 16:32:24 <primeministerp> jeblair: i've started manifiests to build all the deployments bits for windows 16:32:35 <primeministerp> jeblair: we doing it from iron 16:32:35 <mordred> neat 16:32:57 <primeministerp> jeblair: however i'll need a windows vm to build all rest 16:33:15 <alexpilotti> mordred: next week me and ociuhandu will be in Cambridge with primeministerp setting up the CI thing 16:33:21 <primeministerp> have to figure how to make it easiest to inegrate windows in 16:33:24 <mordred> devananda, lifeless: ^^ you and primeministerp should talk - he's spinning up bare metal testing for hyper-v/windows 16:33:27 <primeministerp> er integrate 16:33:35 <primeministerp> gah 16:33:35 <mordred> alexpilotti: awesome 16:33:36 <primeministerp> stuck key 16:33:38 <primeministerp> ok 16:33:43 <alexpilotti> mordred: do you think we could set up a meeting this week? 16:33:52 <devananda> ooh 16:33:54 <mordred> alexpilotti, primeministerp: please. how about thursday some time 16:34:05 <primeministerp> devananda: i'm doing this windows 16:34:10 <mordred> and between now and then you should set up a time to talk to devananda and lifeless 16:34:10 <primeministerp> devananda: er on 16:34:16 <mordred> because it might influence things 16:34:18 <devananda> so i will be on a plane thursday 16:34:18 <alexpilotti> mordred: for me np anytime on Thu 16:34:29 <primeministerp> devananda: tomorrow? 16:34:58 <primeministerp> so 16:35:19 <primeministerp> mordred: we have 11 hp mach 1 nodes 16:35:29 <primeministerp> mordred: and 6 dell blades 16:35:35 <mordred> primeministerp, alexpilotti: we can meet up separate from deva ... 16:35:43 <primeministerp> mordred: yes please 16:35:48 <alexpilotti> mordred: ok 16:35:51 <primeministerp> mordred: want to talk logistics 16:36:03 <primeministerp> mordred: you have time later today or tomorrow? 16:36:05 <mordred> I've got to walk out the door - you pick a time tomorroow or thursday for us to talk (I'm wide open) and I'll be there 16:36:10 <devananda> primeministerp: fwiw, i dont begin to understand windows server. but i can talk about the nova baremetal driver :) 16:36:13 <mordred> today is a TERRIBLE day for me :) 16:36:35 <primeministerp> devananda: you don't need to, i just need to be able to chainload a pxe from pxelinux 16:36:46 <primeministerp> devananda: or understand how to fit it into your plans 16:36:56 <primeministerp> devananda: will you have time tomorrow? 16:37:04 <mordred> chat in a bit... 16:37:24 <devananda> hm. somethin i think chainload is the wrong thing. nobodycam and lifeless both have more opinion and knowledge on that layer than I 16:37:44 <devananda> but yes, i can probably squeeze in some time tmw 16:38:02 <primeministerp> devananda: do you want want to start a thread w/ the appropriate parties? 16:38:13 <primeministerp> devananda: ppouliot@microsoft.com 16:38:22 <devananda> ack 16:38:30 <devananda> this might also give you some context: https://review.openstack.org/#/c/11088/ 16:38:30 <primeministerp> perfect 16:38:37 <josecastroleon> i will come in 5 minutes 16:38:48 <devananda> and https://review.openstack.org/#/c/15830/ 16:39:11 <primeministerp> devananda: are you pulling bare metal instances from glance? 16:39:32 <devananda> primeministerp: yes 16:39:34 <EmilienM_> see you guys, ++ 16:39:38 <primeministerp> devananda: I get it 16:39:42 <alexpilotti> EmilienM_: bye! 16:39:58 <devananda> primeministerp: there is separate "deploy ramdisk" which does the partitioning and writes teh imge to disk 16:40:09 <primeministerp> devananda: yep 16:40:12 <primeministerp> devananda: there would have to be 16:40:14 <devananda> (it's presently a bit more complex but we are working on simplifiyng) 16:40:29 <primeministerp> devananda: i have a lot of ideas around this 16:40:34 <devananda> great 16:40:46 <primeministerp> devananda: ayoung and I have been talking on and off for months 16:41:38 <primeministerp> devananda: i'm open tomorrow so if you want to schedule i'll attend 16:41:54 <primeministerp> and include the individuals from your side to discuss 16:41:57 <primeministerp> devananda: thanks again 16:42:12 <primeministerp> alexpilotti: I'll follow up w/ monty 16:42:24 <alexpilotti> primeministerp: ok 16:42:33 <primeministerp> alexpilotti: also going to start looking at those modules 16:42:58 <alexpilotti> primeministerp: ok, ping me anytime if you need help! 16:43:28 <primeministerp> alexpilotti: i'll circle back w/ monty later today and schedule something for tomorrow as well 16:43:46 <primeministerp> ok if that's i'm ending the meeting 16:43:50 <primeministerp> #endmeeting