14:00:48 <edmondsw> #startmeeting PowerVM Driver Meeting
14:00:49 <openstack> Meeting started Tue Sep 18 14:00:48 2018 UTC and is due to finish in 60 minutes.  The chair is edmondsw. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:50 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:52 <openstack> The meeting name has been set to 'powervm_driver_meeting'
14:00:55 <edmondsw> #link agenda: https://etherpad.openstack.org/p/powervm_driver_meeting_agenda
14:01:54 <edmondsw> efried mujahidali y'all here?
14:02:06 <mujahidali> ^
14:02:20 <efried> ō/
14:02:23 <edmondsw> #topic In-Tree Driver
14:02:44 <edmondsw> I haven't had any time to look at this in a while, and probably won't for a while
14:02:56 <edmondsw> efried anything we should discuss here, particularly anything from the PTG?
14:03:31 <efried> um
14:03:39 <efried> no, I don't think so.
14:03:42 <edmondsw> cool
14:03:44 <edmondsw> moving on
14:03:51 <edmondsw> #topic Out-of-Tree Driver
14:04:11 <edmondsw> I've got a WIP patch up to update our devstack examples: https://review.openstack.org/#/c/601628/
14:04:43 <edmondsw> still working on some things there, but the SEA changes work
14:04:51 <edmondsw> it's an ongoing, as-able type of thing
14:06:11 <edmondsw> mdrabe has also been working on secure boot: https://review.openstack.org/#/c/595877/
14:06:37 <edmondsw> there was a bug in pypowervm 1.1.17 so we had to release 1.1.18 for this to work properly
14:06:54 <edmondsw> I believe the patch has been updated to use 1.1.18 now, but last I checked that wasn't in u-c yet so it wouldn't work
14:07:21 <edmondsw> #action edmondsw check on u-c for pypowervm
14:07:39 <edmondsw> anything else to discuss for OOT?
14:08:38 <edmondsw> mdrabe how are things looking for you to test MSP support, time-wise?
14:09:04 <mdrabe> I can spare a moment
14:09:12 <mdrabe> Do we have a multinode env?
14:09:45 <edmondsw> I have one devstack aio up that you could use, but not a 2nd compute to go with it
14:09:56 <edmondsw> rather, I have a second system partly setup, but not devstacked
14:10:22 <edmondsw> I am tied up in other things atm, but if you're available to work on that I could give you a devstack local.conf that I think should work and you could try it / play with that to get things working
14:11:10 <efried> I can lend my machine if you need a second.
14:11:17 <mdrabe> Maybe, that'll take me some time. Got some other stuff to get to first
14:11:26 <edmondsw> thanks, but it'll need to be the 2nd that I already identified, so they share an SSP
14:11:37 <edmondsw> rather, I already set it up to use the same SSP
14:11:41 <efried> ight
14:12:05 <edmondsw> mdrabe ok, let me know when you can get to it
14:12:13 <edmondsw> #topic Device Passthrough
14:12:17 <edmondsw> efried ^
14:12:52 * efried totally unprepared to report on anything
14:13:13 <efried> Sean Mooney and I are going to work through some of the yaml file format ideas
14:13:36 <efried> Rahul is going to start working on a PoC based on what we've got at the moment.
14:14:00 <efried> Cyborg just might have a spawn-only solution in Stein.
14:14:33 <efried> well, a flavor vehicle, so technically I suppose you could hot attach with resize.
14:14:42 <efried> not sure if that's going to be a supported path.
14:15:01 <efried> closing on nrp is top priority
14:15:27 <efried> That's about all I've got for now, until I get my head organized again. Hopefully more coherent next week.
14:15:30 <efried> Any questions?
14:15:50 <edmondsw> much discussion at the PTG about this stuff?
14:16:04 <efried> yes, from several angles.
14:16:08 <efried> Mainly cyborg
14:16:22 <efried> Let me summarize what that's going to look like for phase 1
14:16:38 <efried> cyborg API/CLI to create a "device profile"
14:16:47 <efried> Stuff a reference to that profile into the flavor
14:16:49 <efried> boot with the flavor
14:17:12 <edmondsw> is that device profile anything like what we've designed to put in our yaml?
14:17:25 <efried> compute communicates with cyborg to turn that reference into an object to pass to virt. Virt does the attach.
14:17:48 <efried> It's analogous to a neutron port.
14:18:12 <efried> So you create it with e.g. a resource class and traits, and it's just hanging out there in the world.
14:18:29 <efried> Then you boot with it and there's a point where it gets "bound" - only then does it become part of an instance+host.
14:18:46 <efried> And then the bound thingy is what gets passed to the virt driver, which does the attach.
14:20:36 <efried> As for the yaml file, we agreed we were going to try to make it a common format that could be read by cyborg and/or nova.
14:20:44 <edmondsw> I assume "boot with it" = "cyborg programming an FPGA"
14:20:48 <efried> no
14:21:04 <efried> phase 1 I believe we're just talking about non-programmable device passthrough.
14:21:16 <efried> I mean nova boot --flavor X
14:21:35 <efried> where flavor X contains a reference to the device profile thingy (VAR - Virtual Accelerator Reference - I believe)
14:22:05 <edmondsw> then I have no idea what this means: "Then you boot with it and there's a point where it gets "bound" - only then does it become part of an instance+host."
14:22:45 <edmondsw> if it's not programming, then the thing exists on the host before any of this process has started, nothing is creating it
14:24:02 <efried> openstack var create --resource-class GPU --traits CUSTOM_FOO CUSTOM_BAR
14:24:02 <efried> ==> $var_uuid
14:24:02 <efried> openstack flavor update --flavor X --extra-specs hw:cyborg:var:$var_uuid
14:24:02 <efried> nova boot --flavor X
14:24:31 <efried> or are you talking about inventorying and discovery?
14:25:41 <efried> actually maybe s/var_uuid/profile_name/ -- I don't think these are one-time-use gizmos, I think they're more like flavors themselves. So the "binding" step actually picks a specific device on the host.
14:25:48 <efried> now I can't remember
14:25:52 <efried> But Sundar is going to put up a spec.
14:25:59 <edmondsw> I'll wait and read the spec
14:26:12 <edmondsw> ping me when it's up?
14:26:17 <efried> https://etherpad.openstack.org/p/stein-ptg.cyborg-nova-new
14:27:04 <edmondsw> lot there
14:27:23 <edmondsw> is this being actively looked at, or are comments there a waste of time?
14:27:39 <edmondsw> s'why I prefer commenting in reviews
14:27:58 <edmondsw> plus they have more structure so easier to make sense of them
14:29:29 <efried> I don't know if Sundar is still consuming that to construct his spec.
14:29:46 <efried> In any case, yeah, I'll let you know when I see that spec come through.
14:29:50 <edmondsw> tx
14:29:54 <efried> But in the meantime, if you have questions/comments, I may be able to answer them
14:29:59 <efried> because they may have already been discussed
14:30:09 <efried> we had several hours to get to this point.
14:30:22 <edmondsw> I wasn't able to follow what you meant above, so...
14:30:52 <efried> above where?
14:30:56 <edmondsw> a lot of it
14:31:31 <edmondsw> looking at the etherpad helped me understand 'bound
14:31:48 <efried> guess it helps to be familiar with the neutron flow
14:31:49 <edmondsw> I'll see if I can figure out the rest later if i get some time
14:31:52 <efried> which I *barely* am.
14:32:02 <edmondsw> might, but I don't think that's my issue here
14:32:20 <edmondsw> anyway, moving on
14:32:21 <efried> k, well, let me know if you want to discuss further.
14:32:23 <efried> yeah
14:32:29 <edmondsw> #topic PowerVM CI
14:32:43 <edmondsw> mujahidali how are things?
14:33:01 <edmondsw> yesterday you had identified a couple nodes that were always timing out.. that fixed?
14:33:06 <mujahidali> Jobs running on Neo24 and ne08 are taking too long to complete and eventually failing after timeout and neo 6 was having some problem with pvmctl so restarted neo6 and redeployed neo,6,8,24.
14:33:34 <mujahidali> CI run looks better now.
14:33:40 <edmondsw> great, tx
14:35:09 <edmondsw> that it?
14:35:18 <mujahidali> I have tried installing nodepool,zuul and jenkins  on my test env but things are a little complicated
14:35:33 <edmondsw> that sounds like an understatement :)
14:35:41 <mujahidali> nodepool and zuul are not launching
14:35:49 <edmondsw> did you also install zookeeper?
14:35:55 <mujahidali> no
14:36:01 <edmondsw> I remember esberglu saying that would be needed with the newer versions
14:36:07 <mujahidali> okay
14:36:44 <mujahidali> I first want to try it on dev env then stage and then prod.
14:37:04 <edmondsw> you have a dev env other than the staging env?
14:37:06 <edmondsw> if so, great
14:37:22 <mujahidali> I am just using vms
14:37:44 <mujahidali> to look how installation and config is working.
14:37:44 <edmondsw> k
14:37:50 <edmondsw> sure
14:38:33 <mujahidali> we had a discussion for the blacklist and whitelist for the 700 test case with :efried
14:39:12 <mujahidali> I am not able to conclude from that discussion.
14:39:28 <edmondsw> I think we table that for the time being
14:39:36 <mujahidali> okay
14:39:53 <edmondsw> anything else?
14:39:56 <mujahidali> that's it from me.
14:40:01 <edmondsw> thanks
14:40:07 <mujahidali> :)
14:40:08 <edmondsw> #topic Open Discussion
14:41:02 <edmondsw> nothing from me here
14:41:07 <edmondsw> anyone else?
14:41:19 <mujahidali> no
14:42:02 <edmondsw> alright, thanks everyone
14:42:05 <edmondsw> #endmeeting