15:58:41 <primeministerp> #startmeeting hyper-v
15:58:41 <openstack> Meeting started Tue Dec 11 15:58:41 2012 UTC.  The chair is primeministerp. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:58:42 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:58:44 <openstack> The meeting name has been set to 'hyper_v'
15:58:53 <primeministerp> hmm, my clock is 2 min off
15:59:04 <primeministerp> pnavarro: how's it going
15:59:30 <primeministerp> going to wait a couple more minutes for everyone to join
15:59:33 <pnavarro> Hi primeministerp !
15:59:47 <primeministerp> pnavarro: how's it going
16:00:00 <primeministerp> pnavarro: any progress on the cinder commands not responding
16:00:02 <pnavarro> i'm doing well, and you?
16:00:14 <primeministerp> pnavarro: a bit tired, we got a dog this weekend
16:00:26 <pnavarro> a dog ! oufff
16:00:38 <pnavarro> a new baby
16:00:40 <primeministerp> pnavarro: and I'm coming off a lot of hours last week w/  the cloudbase team
16:00:56 <primeministerp> pnavarro:  yes, lucy the daschund
16:01:24 <primeministerp> pnavarro: so I saw that alex is going to create an installer for the cinder bits
16:01:32 <pnavarro> about cinder, I couldn't take a look this evening
16:01:38 <primeministerp> pnavarro: ok
16:01:57 <primeministerp> pnavarro: do we at least know if it's a bug?
16:02:47 <pnavarro> well, It looks like a bug, but I'm not sure
16:04:05 <primeministerp> no alex yet
16:04:45 <primeministerp> pnavarro: i just pinged tavi
16:04:57 <primeministerp> pnavarro: we made some progress last week
16:05:07 <primeministerp> we refactored all the puppet bits I started
16:05:20 <primeministerp> ociuhandu: tavi!
16:05:30 <ociuhandu> hello all
16:05:31 <primeministerp> ociuhandu: is alex going to join as well
16:05:52 <ociuhandu> yes, he'll be joining in a couple of minutes
16:07:30 <alexpilotti> hi guys
16:07:39 <pnavarro> hi alexpilotti
16:07:43 <alexpilotti> sorry for being late!
16:08:10 <primeministerp> hey no worries
16:08:15 <primeministerp> so
16:08:21 <primeministerp> i guess we'll offically begin
16:08:21 <alexpilotti> my Colloquy was disconnected, can u plz tell me what's teh current topic? :-)
16:08:31 <primeministerp> #topic cinder
16:08:37 <alexpilotti> cool
16:08:40 <primeministerp> pnavarro: and i were discussing the bug
16:08:52 <primeministerp> alexpilotti: from the email
16:08:53 <alexpilotti> what bug? :-)
16:09:02 <primeministerp> alexpilotti: i wanted to confirm it's fit's a bug or not
16:09:20 <primeministerp> alexpilotti: the non functioning cinder commands
16:09:30 * primeministerp looks at the email
16:09:45 <primeministerp> alexpilotti: that you found
16:09:59 <alexpilotti> sure
16:10:22 <alexpilotti> it was a question. How to remove a cinder-volume service
16:10:26 <alexpilotti> the case is simple:
16:10:28 <primeministerp> alexpilotti: removing existing volumes
16:10:29 <alexpilotti> you run devstack
16:10:57 <primeministerp> alexpilotti: is it definately a bug?
16:11:03 <alexpilotti> ok, I'm missing something :-)
16:11:10 <primeministerp> alexpilotti: the email you sent
16:11:22 <alexpilotti> the email I sent today?
16:11:27 <primeministerp> On the controller I didn't find a way to remove an existing cinder-volume service except editing the db.
16:11:29 <primeministerp> yes
16:11:42 <alexpilotti> yep
16:11:48 <alexpilotti> so the case is simple:
16:11:53 <alexpilotti> you have e.g. devstack
16:12:00 <primeministerp> i just want clarity on what the problem is
16:12:02 <alexpilotti> you stop the cinder-volume service
16:12:22 <alexpilotti> you add  cinder-volume running on Windows Storage
16:12:43 <alexpilotti> on the controller you'll still see both services
16:12:54 <alexpilotti> so if you run: cinder volume create 1
16:12:54 <primeministerp> ahh gotcha
16:13:03 <alexpilotti> it will try to use either one
16:13:07 <primeministerp> it will try to create on the wrong vol server
16:13:09 <primeministerp> i get it
16:13:10 <primeministerp> now
16:13:17 <primeministerp> i thought it was on the target end
16:13:23 <primeministerp> something in pnavarro bits
16:13:30 <primeministerp> that wasn't functioning
16:13:30 <alexpilotti> no, nothing on our side
16:13:33 <primeministerp> o
16:13:35 <primeministerp> great
16:13:41 <primeministerp> pnavarro: you're off the hook
16:13:42 <primeministerp> ;)
16:13:49 <pnavarro> thanks
16:13:52 <primeministerp> haha
16:13:54 <alexpilotti> pnavarro: do you have any clue? I'm going to ask the cinder guys otherwise
16:14:00 <primeministerp> alexpilotti: any timeline on the installer?
16:14:04 <alexpilotti> lol
16:14:14 <primeministerp> alexpilotti:sorry
16:14:14 <pnavarro> sorry alex, no idea what's the problem...
16:14:19 <alexpilotti> ah yep. I'm going to release the installer today
16:14:26 <primeministerp> alexpilotti: awesome
16:14:32 <alexpilotti> for cinder-volume only, that's what we need
16:14:33 <primeministerp> pnavarro: no problem on our end
16:14:50 <primeministerp> pnavarro: just didn't automatically remove the linux service
16:14:52 <alexpilotti> IMO we can leave out cinder-api and cinder-scheduler
16:14:56 <primeministerp> pnavarro: if i understand correctly
16:15:01 <primeministerp> alexpilotti: why?
16:15:36 <alexpilotti> because I had some issues on running them with the latest Grizzly bits on Windows
16:15:49 <primeministerp> alexpilotti: *nod*
16:16:15 <alexpilotti> and because those services run usually on a controller, not on each volume node
16:16:33 <alexpilotti> let's say that "cinder-volume" is for Cinder what "nova-compute" is for Nova
16:16:42 <primeministerp> i get it
16:16:53 <primeministerp> thx
16:16:55 <primeministerp> ok
16:17:01 <primeministerp> guess we'll move on
16:17:04 <alexpilotti> of course we want to have all the rest on WIndows as well ;-)
16:17:13 <alexpilotti> sure
16:17:13 <primeministerp> ;)
16:17:21 <primeministerp> #topic quantum
16:17:27 <alexpilotti> ahem
16:17:33 <primeministerp> ?
16:17:36 <alexpilotti> next topic? :-)
16:17:42 <primeministerp> quantum?
16:17:52 <primeministerp> was there something i missed?
16:18:04 <alexpilotti> kidding, due to the CI stuff I was not able to do more than looking at pnavarro's bits
16:18:13 <alexpilotti> pnavarro: do you have anything new?
16:18:30 <primeministerp> alexpilotti: i was asking mostly for pnavarro's input
16:18:35 <primeministerp> pnavarro: you're on the hook again
16:18:38 <primeministerp> ;)
16:18:45 <alexpilotti> lol
16:18:49 <pnavarro> alexpilotti: agent is running ok, i'm still testing agent- plugin communication
16:19:09 <alexpilotti> pnavarro: great!
16:19:13 <primeministerp> pnavarro: great
16:19:24 <pnavarro> so, I'd appreciate any help
16:19:40 <alexpilotti> pnavarro: this week, I swear, I'm working on it :-)
16:19:50 <pnavarro> perfect
16:19:50 <primeministerp> and alexpilotti you'll start work again shortly
16:20:15 <primeministerp> alexpilotti: will we be able to account for the nvgre bits
16:20:22 <alexpilotti> pnavarro: on the agent side, what hyper-V specific features did you test already?
16:20:51 <ociuhandu> sorry guys, i have to run now, will ba back in about 45 minutes
16:21:03 <ociuhandu> primeministerp: i'll ping you when i get back
16:21:03 <primeministerp> ociuhandu: ok, let's touch base re: ci
16:21:09 <primeministerp> ociuhandu: great
16:21:15 <pnavarro> I tested ports updates
16:21:22 <ociuhandu> bye all & yalk to you later
16:22:28 <primeministerp> pnavarro: nice
16:22:30 <alexpilotti> pnavarro: this time, we should think about a FakePlugin since we don't have the rush
16:22:36 <pnavarro> alexpilotti: yeah
16:23:09 <primeministerp> alexpilotti: and by fake plugin?
16:23:20 <primeministerp> for my ignorance
16:23:24 <primeministerp> please
16:23:34 <primeministerp> same as the wmi test interface
16:23:43 <primeministerp> but for quantum only?
16:23:52 <alexpilotti> basically a class that emulates the wmi calls, by creating an asbtarction layer on them
16:24:06 <primeministerp> gotcha
16:24:07 <alexpilotti> so in the code, instead of calling WMI we call this new layer
16:24:28 <alexpilotti> which can be replaced by a fake one during tests
16:24:40 <alexpilotti> we have to do the same for Nova and Cinder of course
16:24:46 <pnavarro> alexpilotti the wmi library creator gave you some hint about that?
16:25:27 <alexpilotti> pnavarro: nope :-(
16:25:27 <alexpilotti> pnavarro: but if we wrap all the calls we are fine
16:25:49 <primeministerp> alexpilotti: aren't you wrapping them 2x then?
16:25:50 <alexpilotti> pnavarro: this IMO will not reach the level of coverage we had with the dynamic mocks, but it's ok
16:26:13 <alexpilotti> primeministerp: yes, no comment
16:26:18 <primeministerp> hahaha
16:26:27 <alexpilotti> indeed we are not really wrapping them twice
16:26:43 <alexpilotti> there while be a lower level fine grained layer
16:26:48 <primeministerp> *nod*
16:26:49 <primeministerp> hehe
16:26:51 <alexpilotti> and a high level (the driver itself)
16:27:02 <primeministerp> fair enough
16:27:25 <alexpilotti> the point is that we should test the low level one as well… that's the funny part
16:27:27 <primeministerp> alexpilotti: maybe we can leverage the new relationship w/ tom for assistance
16:27:39 <primeministerp> w/ getting you information
16:27:47 <primeministerp> resources
16:28:04 <alexpilotti> unfortunately the way in which WMI is done is the problem
16:28:26 <alexpilotti> the amazing flexibility it provides becomes a huge PINTA for testing
16:28:53 <primeministerp> *nod*
16:29:02 <primeministerp> we definately don't have it easy
16:29:08 <pnavarro> yeah, we can't use mocks library
16:29:20 <primeministerp> so why not specifically
16:29:32 <primeministerp> the nature of having to test the network itself?
16:30:23 <alexpilotti> setting up the tests will be relatively easy, as we can take the OVS ones as a reference
16:30:56 <alexpilotti> pnavarro: is your repo updated?
16:31:02 <pnavarro> yeah
16:31:23 <pnavarro> alexpilotti: I've added the binary
16:31:37 <alexpilotti> pnavarro: this it the latest plugin? https://github.com/pnavarro/quantum/tree/master/quantum/plugins/windowsvswitch
16:31:45 <pnavarro> yes
16:31:51 <primeministerp> #link https://github.com/pnavarro/quantum/tree/master/quantum/plugins/windowsvswitch
16:32:03 <pnavarro> If you don't like the name, we can change it
16:32:34 <primeministerp> pnavarro: name doesn't bother me
16:32:49 <alexpilotti> pnavarro: I'd go with hypervswitch, but that's a detail :-)
16:32:55 <pnavarro> oki doki
16:32:59 <primeministerp> pnavarro: he's a bit right there
16:33:15 <primeministerp> alexpilotti: ... and your the purist
16:33:20 <primeministerp> ;)
16:33:26 <primeministerp> ok
16:33:29 <primeministerp> should we move on?
16:33:35 <primeministerp> to ci talk?
16:33:47 <alexpilotti> pnavarro: I'm loking at the utils.py, good work!
16:34:11 <pnavarro> thanks alexpilotti !
16:34:30 <primeministerp> #topic CI
16:34:43 <primeministerp> alexpilotti: this will be quick
16:34:52 <primeministerp> since i'll have to catch up w/ tavi later
16:34:54 <alexpilotti> lol
16:35:12 <primeministerp> but I figure we'll coordinate
16:35:18 <primeministerp> then
16:35:25 <alexpilotti> sure, ociuhandu will be back in ca 1hr
16:35:26 <primeministerp> i'm assuming you guys haven't picked up yet
16:35:33 <pnavarro> i'll read the logs later, I have to go
16:35:36 <primeministerp> from where we left off
16:35:40 <primeministerp> pnavarro: thanks for everything
16:35:51 <pnavarro> ciao !
16:35:52 <alexpilotti> I resumed the work this afternoon (aka 2 hours ago)
16:36:18 <primeministerp> alexpilotti: on the scripting?
16:36:22 <alexpilotti> pnavarro: hasta luego!
16:36:46 <alexpilotti> primeministerp: yep, I'm polishing up the scripts for the controller
16:36:54 <primeministerp> alexpilotti: ok
16:37:10 <primeministerp> alexpilotti: i figured out why my post script wasn't running out of the preseed
16:37:19 <primeministerp> alexpilotti: fixed and going test
16:37:24 <alexpilotti> cool
16:37:28 <primeministerp> also hoping to resume work on the windows unatted bit
16:37:28 <primeministerp> s
16:37:42 <primeministerp> next couple of days are going to be tough for me
16:37:45 <primeministerp> due to meetings
16:37:47 <alexpilotti> primeministerp: BTW VPN works fine also from this side of the "pond" :-)
16:37:48 <primeministerp> and now the move
16:37:52 <primeministerp> alexpilotti: great
16:38:03 <primeministerp> alexpilotti: i resurected my asa at home
16:38:12 <primeministerp> I'm going to put a static tunnel up for me
16:38:16 <primeministerp> now that I have to move
16:38:18 <primeministerp> my office
16:38:48 <primeministerp> also ordered sound protective bluetooth headset for working in the dc
16:39:07 <primeministerp> anyway
16:39:23 <primeministerp> o
16:39:27 <primeministerp> mordred: ping
16:39:43 <primeministerp> mordred: whats the best way for us to start the process of integration our hyper-v work
16:40:02 <primeministerp> alexpilotti: i was thinking we can start adding the bits we've created
16:40:21 <primeministerp> alexpilotti: that way we don't dump a whole lot on top
16:40:36 <primeministerp> alexpilotti: and can increment at a minimum the working parts
16:40:45 <primeministerp> alexpilotti: any thoughts
16:40:59 <primeministerp> alexpilotti: also then we'd be working out of git
16:41:45 <primeministerp> well mordred might be busy i'll try to follow up w/ him later
16:42:47 <alexpilotti> primeministerp: Since finishing the "layer 1" stuff is almost done, can we set up a meeting wit mordred this or next week?
16:42:47 <primeministerp> ok
16:43:00 <primeministerp> alexpilotti: that's my thoughts
16:43:05 <primeministerp> alexpilotti: i'll work on that
16:43:06 <alexpilotti> cool
16:43:19 <primeministerp> alexpilotti: and let's plan on having a skype when tavi gets back
16:43:25 <primeministerp> just to touch base
16:43:33 <alexpilotti> coo, I'll be here as well
16:43:38 <primeministerp> ok great
16:43:47 <primeministerp> i've got to head in into the office
16:43:52 <primeministerp> so i'll be offline for a while
16:43:59 <primeministerp> ok closing the meeting
16:44:03 <primeministerp> #endmeeting