Wednesday, 2018-03-07

*** k0da has quit IRC00:43
*** AlexeyAbashkin has joined #openstack-powervm01:49
*** AlexeyAbashkin has quit IRC01:53
*** edmondsw has quit IRC02:21
*** edmondsw has joined #openstack-powervm02:22
*** edmondsw has quit IRC02:26
*** edmondsw has joined #openstack-powervm05:09
*** edmondsw has quit IRC05:14
*** edmondsw has joined #openstack-powervm06:58
*** edmondsw has quit IRC07:03
*** AlexeyAbashkin has joined #openstack-powervm07:58
*** k0da has joined #openstack-powervm08:34
*** edmondsw has joined #openstack-powervm08:46
*** edmondsw has quit IRC08:50
*** edmondsw has joined #openstack-powervm10:34
*** edmondsw has quit IRC10:38
*** edmondsw_ has joined #openstack-powervm13:25
*** apearson has joined #openstack-powervm13:39
*** esberglu has joined #openstack-powervm13:48
*** esberglu has quit IRC14:14
openstackgerritSridhar Venkat proposed openstack/nova-powervm master: Support maximum capacity for SR-IOV VNIC
*** esberglu has joined #openstack-powervm14:29
esbergluedmondsw_: See 6371 to restore CI please15:13
*** edmondsw_ is now known as edmondsw15:13
edmondswesberglu ack15:13
edmondswesberglu commented15:19
edmondswmaybe just trim that line down to *only* the bug link15:19
esbergluedmondsw: Still a TODO for pike and queens, I'm proposing backports15:21
esbergluI can update the comment to say that but I think it's clear from the logic15:21
edmondswesberglu you think they will accept backports?15:23
edmondswI think pike is almost EOL15:24
esbergluedmondsw: I don't see why thy wouldn't backport, it doesn't change the default behavior at all15:24
esbergluedmondsw: I've only proposed the queens backport for now I'll let you know what they say15:24
edmondswesberglu alright, changed to +215:25
*** apearson has quit IRC15:45
openstackgerritSridhar Venkat proposed openstack/nova-powervm master: Support maximum capacity for SR-IOV VNIC
*** apearson has joined #openstack-powervm15:54
edmondswesberglu why did you abandon the queens backport saying it was fixed in master?16:22
edmondswoh, nm, that wasn't the backport16:23
esbergluedmondsw: That's a queens backport for the master patch I proposed16:23
edmondswesberglu zuul has normally voted in about 10 minutes on but with the latest PS it's been over 30 and no vote... know of anything going on?16:27
esbergluedmondsw: The zuul queue probably just got backed up. The vote is about to come in it just passed all tests16:28
esberglu^ You can check the status of zuul runs there16:28
edmondswesberglu is that working for you?16:32
esbergluedmondsw: Yep16:32
edmondswesberglu isn't for me :(16:36
esbergluwhat browser?16:37
edmondswchrome or firefox16:37
edmondswthe page loads, but is very slow, and I don't see anything with powervm on there16:37
edmondswsearching doesn't seem to find anything16:37
edmondsw(but takes forever... not sure it's even really working)16:37
esbergluedmondsw: Yeah its slow and buggy sometimes. No powervm because your run finished16:38
edmondswI closed some apps to make sure it wasn't a mem issue on my side, but didn't help16:38
edmondswhey, look at that... finished 2 minutes after I pinged you16:39
*** apearson has quit IRC16:40
*** apearson has joined #openstack-powervm16:44
*** k0da has quit IRC16:53
*** AlexeyAbashkin has quit IRC18:06
edmondswesberglu I never got back to you yesterday afternoon... was there anything we needed to discuss that you were going to bring up if we had a mtg?18:25
esbergluedmondsw: I was just gonna go through the status on my IT changes18:26
edmondswI'm trying to get to those but something keeps popping up18:26
edmondswanything to note that's not in the reviews?18:27
esbergluvSCSI is good for review, the only new thing there is adding extend_volume (which doesn't actually do anything)18:27
esbergluNetwork hotplug is still waiting on live test18:27
esbergluSnapshot is ready for review18:27
esbergluI'm still messing around with the localdisk changes18:28
edmondswoh, you went ahead with extend volume in the vSCSI patch? I thought we'd probably do that as a followon18:28
edmondswdo we have anything testing extend volume in the CI?18:30
edmondswesberglu I presume no, in which case we should look at adding that18:30
esbergluedmondsw: Not testing vscsi in CI18:31
edmondswoh, true18:32
edmondswnevermind then18:32
edmondswesberglu I did start looking at snapshot but need to get back to that18:33
esbergluedmondsw: Assuming you saw that PowerStackers should get approved on friday?18:35
edmondswyep :)18:35
esberglutjakobs is probably getting eager for volume refactor reviews18:36
esbergluI keep starting it but haven't had time to get all the way through18:37
edmondswyeah, same18:37
esbergluI haven't really progressed much on the Pike Undercloud. The glance uwsgi patches for pike were abandoned18:38
esberglunow it looks like some version of this18:38
esbergluIs probably where that's gonna land18:39
edmondswthose first 2 links are the same18:39
esbergluRight now I'm cherry picking 545483 from master to pike during the deploy which is kind of gross18:40
esbergluBut hopefully we can eventually get that backported18:40
esbergluIt works though18:40
esbergluStill have no idea whats going on with jenkins dropping connections18:41
esbergluBeen working off of this page18:41
esbergluHoping to hop back into that tomorrow18:42
esbergluThat's everything I had for the meeting18:42
edmondswesberglu I looked at the comments in those uwsgi reviews... looks to me like those backports were abandoned prematurely18:44
edmondswlook at mtreinish's comment on 54548318:45
edmondswand the -2 on 525994 was just until they had a chance to discuss at the PTG18:45
edmondswmight be a good idea to reach out to mtreinish and see where this stands / get those unabandoned18:46
esbergluHadn't seen that comment yet. Yeah I can do that18:49
edmondswesberglu re: the volume refactor, how do you feel about using FCVscsiVolumeAdapter instead of FCPVVscsiVolumeAdapter? both IT and OOT.18:51
edmondswI don't think we'd ever have a FC subclass for VdiskVscsiVolumeAdatper that we'd need to distinguish that from, so it would be like how we say FileIOVolumeAdapter instead of FileIOVdiskVscsiVolumeAdapter18:52
edmondswstill need to specify Vscsi just to distinguish from NPIV18:53
esbergluedmondsw: Yeah that makes sense and is way more readable18:53
edmondswesberglu filename could also shorten from to just fcvscsi.py18:54
*** tjakobs has joined #openstack-powervm18:59
edmondswesberglu I spoke to tjakobs and he's fine with the name change as well19:18
esbergluedmondsw: Okay I'll change it IT19:18
edmondswyep, tx19:19
*** AlexeyAbashkin has joined #openstack-powervm19:33
*** AlexeyAbashkin has quit IRC19:37
*** AlexeyAbashkin has joined #openstack-powervm19:38
*** k0da has joined #openstack-powervm19:42
esbergluedmondsw: How should we handle duplicate conf options between nova-powervm and nova?19:46
esbergluIf we add for example the disk_driver conf option to nova19:47
esbergluWe will get duplicate conf option errors with the OOT driver19:47
esbergluSince disk_driver is also a conf option for nova-powervm19:48
edmondswah, good question19:48
edmondswprobably add it to nova and then change nova-powervm to use the nova one and stop defining another19:49
edmondswif we need the default to be different OOT or something, that might get trickier...19:50
esbergluedmondsw: I think that should be fine. I'm looking to see if it will mess up our devstack plugin stuff in nova-powervm quick19:51
edmondswideally nova_powervm would have used a nova_powervm conf group instead of powervm...19:51
*** tjakobs_ has joined #openstack-powervm19:51
edmondswbut I'm not sure we can change that now19:51
esbergluedmondsw: If you're concerned about the defaults, we could name the IT conf options something slightly different19:52
esbergluOr name the IT group something besides powervm19:52
edmondswI want to name the IT group/options with the names that make the most sense19:52
edmondswif anything, I care less about the names for OOT19:52
edmondswsince I hope OOT eventually goes away19:52
*** tjakobs has quit IRC19:54
*** tjakobs_ has quit IRC20:05
*** k0da has quit IRC20:12
edmondswesberglu threw you a comment on localdisk20:17
edmondswyou might be able to avoid a rebase for a while by just overriding the setting dynamically instead of setting it dynamically20:19
esbergluedmondsw: Wasn't planning on bringing imagecache in that change20:19
edmondswohh... ok20:19
edmondswyeah, I think I remember you saying let's split that out20:20
esbergluBut noted for when I get to that point20:20
*** AlexeyAbashkin has quit IRC20:29
*** AlexeyAbashkin has joined #openstack-powervm20:45
*** AlexeyAbashkin has quit IRC20:50
*** AlexeyAbashkin has joined #openstack-powervm21:04
*** AlexeyAbashkin has quit IRC21:28
*** k0da has joined #openstack-powervm21:54
*** k0da has quit IRC21:59
*** k0da has joined #openstack-powervm22:11
*** openstackstatus has quit IRC22:13
*** openstack has joined #openstack-powervm22:15
*** ChanServ sets mode: +o openstack22:15
*** esberglu has quit IRC22:35
*** edmondsw has quit IRC22:38
*** edmondsw has joined #openstack-powervm22:39
*** openstackstatus has quit IRC22:42
*** openstack has joined #openstack-powervm22:44
*** ChanServ sets mode: +o openstack22:44
*** AlexeyAbashkin has joined #openstack-powervm22:46
*** esberglu has joined #openstack-powervm22:50
*** AlexeyAbashkin has quit IRC22:51
*** esberglu has quit IRC23:09
*** apearson has quit IRC23:14
*** AlexeyAbashkin has joined #openstack-powervm23:45
*** AlexeyAbashkin has quit IRC23:49

Generated by 2.15.3 by Marius Gedminas - find it at!