14:00:03 #startmeeting PowerVC Driver Meeting 14:00:04 Meeting started Tue May 15 14:00:03 2018 UTC and is due to finish in 60 minutes. The chair is edmondsw. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:08 The meeting name has been set to 'powervc_driver_meeting' 14:00:13 ō/ 14:00:27 #link https://etherpad.openstack.org/p/powervm_driver_meeting_agenda 14:01:00 esberglu mujahidali mdrabe meeting time 14:01:09 o/ 14:01:33 #topic In-Tree Driver 14:01:47 No news other than we're back in the runway for snapshot and localdisk 14:01:55 any reviews yet? 14:02:09 Nope, exit date for our runway is 5/28 14:02:16 #link https://etherpad.openstack.org/p/powervm-in-tree-todos 14:02:31 we should probably update ^ 14:02:37 I can do that after the meeting 14:03:25 #topic Out-of-Tree Driver 14:03:34 #link https://etherpad.openstack.org/p/powervm-oot-todos 14:04:02 the refactor is definitely on hold for the time being... I should probably note that in the etherpad 14:04:30 @efried do you recall what the get_instance_diagnostics item is about? 14:04:44 ... 14:05:06 o/ 14:05:18 nothing about that looks familiar to me. 14:05:34 k, I'll have to dig in and remind myself what that is 14:05:43 obviously nobody working on it atm 14:05:56 chhagarw or gman-tx how goes iSCSI? 14:06:59 Chhavi is still working on lpm and rr 14:07:00 It doesn't look like all the in-tree drivers have implemented it, in any case. 14:07:04 (diagnostics ^) 14:08:20 efried I think it was something PowerVC wanted but apparently nobody has been working it from there either 14:08:22 I'll check with gautam 14:08:34 ack 14:09:01 gman-tx I saw she abandoned one change in favor of another. I should probably compare and see why she did that 14:09:24 and we have this timeout issue that needs to get solved 14:09:45 gman-tx you're pushing the NovaLink guys to fix on their side as I understood things, is that correct? 14:09:50 edmondsw: there is a review out there for rr and lpm 14:10:16 chhagarw assuming you mean https://review.openstack.org/#/c/567575/ I will look today 14:10:56 do we think that will that be the last thing we need for iscsi? 14:11:19 https://review.openstack.org/#/c/567575/ 14:11:21 yes this one 14:11:51 for nova-powervm this would be the last feature we are planning to get in 14:12:01 for iSCSI 14:12:52 ok great 14:13:11 efried I never did talk to you about MSP after last week's meeting... 14:13:32 let's do that today 14:13:34 ight 14:13:44 anything else for OOT? 14:14:22 #topic: Device Passthrough 14:15:21 efried any updates? 14:15:46 nova side of granular is ready and has had a couple rounds of reviews. Should go soon. 14:16:01 nrp-in-alloc-cands is getting reviews, but it's going to be pretty tight whether it makes it in this week. 14:16:14 Otherwise, nothing new. 14:16:25 are those in a runway? 14:16:34 no, but they're not really needing to be. 14:17:32 k 14:17:36 #topic PowerVM CI 14:17:48 esberglu mujahidali status? 14:18:31 I am working with Eric on scenario testing 14:19:20 mujahidali was able to get all of the OOT stable branches working 14:19:53 And queens IT was missing that nova patch for the str instance UUID 14:20:29 But making progress there 14:20:37 yeah, I applied the patch and started again. 14:20:40 esberglu so scenario testing has been merged only for master at this point? 14:20:44 edmondsw: Yes 14:21:11 if we've got it working for all the OOT stables, is there something I should be reviewing to get that merged? 14:21:39 or are we going to do those together with IT? 14:21:52 edmondsw: Nothing up yet. I think we were going to wait for queens IT 14:22:00 And pike IT won't have scenario 14:22:25 ok 14:22:47 CI status looks pretty good http://ci-watch.tintri.com/project?project=nova 14:23:04 I did noticed failures on https://review.openstack.org/#/c/565487 that you might want to look at efried 14:23:24 anything else for CI? 14:23:25 Yeah CI has been rolling along nicely. 14:23:40 I disable that vopt name by instance UUID change 14:24:12 There are a few tempest failures still open, but all are hitting at very low rates 14:24:24 More details are on the tempest failure etherpad 14:24:37 That would be a good thing for mujahidali to start getting involved in 14:24:42 +1 14:24:53 sure 14:25:12 mujahidali: Let me know if you have question about any of those and I can fill you in 14:25:23 Last item is vSCSI CI 14:25:48 I can't get active RMC connections in the tempest test instances 14:26:10 even once we switched to using the right image? 14:26:12 Previously I was using a blank image, but even with the correct image no RMC 14:26:28 odd 14:26:33 The problem is that the test instances don't have connectivity out 14:26:50 So you can ssh from tempest AIO vm into the tempest test instance 14:27:06 But the tempest test instance can't ping anything 14:27:35 edmondsw: Could you help me take a look at how I set up those networks? 14:27:41 esberglu yes 14:27:43 I may be doing something wrong there 14:28:18 I've started putting together a set of vSCSI tests on neo39 which is a vSCSI stack 14:28:35 This is NOT part of the CI env, but should give a nice starting point for when RMC is working 14:29:05 That's all for CI 14:29:05 +1 14:29:30 esberglu I assume we've not made any more progress on multinode because of these other priorities, correct? 14:30:01 edmondsw: Correct, I haven't looked at multinode since scenario and vSCSI became work items 14:30:32 * edmondsw nods 14:30:43 oh, I forgot to link the etherpad 14:30:47 #link https://etherpad.openstack.org/p/powervm_ci_todos 14:30:51 but we went through everything I think 14:30:57 #topic Open Discussion 14:31:20 note that the summit is next week 14:31:28 PIE development work is starting this week, I'm now splitting my time 14:31:29 edmondsw: Those CI failures are different IT and OOT and seem spurious, unrelated to the patch. In tree we're getting "Unable to insert record" on PUT /rest/api/uom/Tier/87d59125-9472-3f1b-8c17-f429e2a3d535/LogicalUnit during a test of multibyte chars in the instance name. We've sanitized the boot disk name to boot_pvm4_tempest___1_0bd52c14 so that's not a problem. 14:31:45 The other one is trying to disconnect vopt while the media is still "loaded". 14:31:59 efried ok cool, sounds like a recheck 14:32:04 yuh 14:32:12 esberglu ack 14:32:14 but may as well wait 14:32:24 for the pep8 fix, sure 14:32:25 cause pep needs to be fixed, so another ps is on the way eventually. 14:32:49 efried: https://etherpad.openstack.org/p/powervm_tempest_failures both are on there, recheck should go through 14:32:57 rgr 14:33:47 I will cancel next week's meeting because of the summit 14:34:35 I think that's all then, folks. Thanks! 14:34:38 #endmeeting