| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-09-27 | |||
| 12:00:18 | kashyap | efried: Indeed, that's it -- the "kvm_init_vcpu failed" error gives it away (it's _not_ main memory allocation, but a prob w/ NUMA interaction) | |
| 12:00:37 | efried | kashyap: sweet. Can you respond? | |
| 12:00:46 | kashyap | efried: I'll reply there; first to try with kernel 4.19 | |
| 12:00:52 | efried | beaut | |
| 12:00:54 | efried | thank you | |
| 12:01:00 | kashyap | efried: Will do; seems like it _just_ arrived? | |
| 12:01:14 | efried | yeah, <5h ago | |
| 12:01:16 | efried | serendipity | |
| 12:01:37 | kashyap | Ah, missed it; as I was reading only the [nova] tag | |
| 12:02:10 | kashyap | (And this didn't have the tag) | |
| 12:02:10 | efried | Couldn't tell you why I actually read this one. Maybe because it said "numa affinity"? | |
| 12:02:20 | kashyap | Likely | |
| 12:02:44 | kashyap | I don't know if we should make it into yet-another tunable; I'd think we should do the _right_ thing there. Still mulling | |
| 12:03:07 | efried | but someone running into this can resolve by upgrading kernel? | |
| 12:03:08 | kashyap | (Where just use the conservative, 'preferred'. I'll let Sean come back to triple-check on me) | |
| 12:03:23 | kashyap | efried: That's my theory. Let's see if this reporter confirms. | |
| 12:03:31 | efried | is there not a case where you would truly *want* to use strict? | |
| 12:03:48 | kashyap | Also the kernel fix is extremely small (yes, yes, that doesn't say _anything_ about risk) and isoloated; so I'd suggest the Linux distros backport it. | |
| 12:03:52 | efried | seems like something something edge HPC realtime something | |
| 12:04:27 | kashyap | efried: There are cases, but I'll double-check w/ the libvirt dev who meditates on this | |
| 12:05:01 | efried | cool. /me watches from a safe distance | |
| 12:05:04 | kashyap | Heh | |
| 12:06:26 | openstackgerrit | Stephen Finucane proposed openstack/python-novaclient master: Stop silently ignoring invalid 'nova boot --hint' options https://review.opendev.org/684762 | |
| 12:08:36 | kashyap | efried: Completely unrelated - I wonder if all the 'no-reply' emails can be directed to another 'openstack-releases' list | |
| 12:08:51 | kashyap | Otherwise the noise feels just too much. | |
| 12:08:56 | kashyap | Maybe I should ask it on the list | |
| 12:09:22 | kashyap | I know that the point of it is to give a heads-up to devs that a new release came out. | |
| 12:09:29 | efried | kashyap: respond to http://lists.openstack.org/pipermail/openstack-discuss/2019-September/009744.html | |
| 12:09:59 | kashyap | Ah, will do | |
| 12:10:03 | kashyap | Thanks for the pointer | |
| 12:10:50 | efried | yahyoubetcha. I didn't care enough to respond; I'm happy to hit [Delete] a couple dozen times twice a year. | |
| 12:11:54 | kashyap | My gripe is, the mailing list archives turn into a toilet | |
| 12:12:10 | kashyap | It just becomes so polluted that the signal is drowned in that noise. | |
| 12:12:34 | kashyap | And it shoos away those who're reading the list to ignore it. I'll respond on that thread. :D | |
| 12:22:13 | zigo | Is Nova really broken with cryptography 2.6 ? | |
| 12:22:27 | zigo | That's kind of annoying me to require 2.7, because the Debian package isn't realy. | |
| 12:22:33 | zigo | (and I'm not the maintainer of that one...) | |
| 12:45:00 | zigo | bauzas: ^ Do you know ? | |
| 12:46:08 | openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (13) https://review.opendev.org/576020 | |
| 12:46:16 | openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (14) https://review.opendev.org/576027 | |
| 12:46:30 | openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (15) https://review.opendev.org/576031 | |
| 12:46:40 | openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (16) https://review.opendev.org/576299 | |
| 12:46:51 | openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (17) https://review.opendev.org/576344 | |
| 12:47:02 | openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (18) https://review.opendev.org/576673 | |
| 12:47:14 | openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (19) https://review.opendev.org/576676 | |
| 12:47:23 | openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (20) https://review.opendev.org/576689 | |
| 12:47:31 | openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (21) https://review.opendev.org/576709 | |
| 12:47:42 | openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (22) https://review.opendev.org/576712 | |
| 13:04:06 | gibi | stephenfin: regarding https://bugs.launchpad.net/python-novaclient/+bug/1845322 I think the same bug exists in openstackclient as well | |
| 13:04:06 | openstack | Launchpad bug 1845322 in python-novaclient "'--hint' argument to 'nova boot ' should be a key-value pair" [Low,In progress] - Assigned to Stephen Finucane (stephenfinucane) | |
| 13:09:47 | gibi | stephenfin: opened a story for it https://storyboard.openstack.org/#!/story/2006628 | |
| 13:18:10 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Add TODO note for mox removal https://review.opendev.org/576758 | |
| 13:22:49 | openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/nova stable/pike: Avoid redundant initialize_connection on source post live migration https://review.opendev.org/683008 | |
| 13:27:34 | efried | stephenfin: Is your nova-net removal series | |
| 13:27:43 | efried | ... "done"? Like, at least proposed all the way up? | |
| 13:27:47 | efried | (I haven't looked) | |
| 13:28:13 | efried | curious whether we could just do away with https://review.opendev.org/#/c/576758/ and/or do the actual removal on top of your series. | |
| 13:30:14 | efried | ...looks like just the test bits are proposed at this point. | |
| 13:45:44 | stephenfin | gibi: Yeah, I pushed patches for that | |
| 13:46:08 | stephenfin | Was trying to figure out why I couldn't file against openstackclient on launchpad. That would be why,heh | |
| 13:46:45 | stephenfin | efried: Yeah, just the functional test piece pushed so far. I'm working on the rest of it slowly | |
| 13:47:06 | stephenfin | Trying to avoid a mega "remove all the things" patch | |
| 13:47:20 | efried | stephenfin: as for cellsv1, good approach. | |
| 13:48:11 | stephenfin | efried: And the ec2 cleanup...which isn't actually merged yet (I thought it was). Would you mind looking at them if/when you have a chance? https://review.opendev.org/#/q/topic:ec2-cleanup+(status:open+OR+status:merged) | |
| 13:48:12 | gibi | stephenfin: cool reading the openstackclient patch now | |
| 13:51:34 | dansmith | mriedem: I think I'm +2 on all the +2able things in stein | |
| 13:51:42 | dansmith | mriedem: (re: your mention from the meeting yesterday) | |
| 13:51:55 | openstackgerrit | melanie witt proposed openstack/nova master: Reduce scope of 'path' query parameter to noVNC consoles https://review.opendev.org/685194 | |
| 13:52:22 | mriedem | dansmith: ack | |
| 13:53:11 | mriedem | lyarwood: if you're about can you hit these? https://review.opendev.org/#/q/project:openstack/nova+branch:stable/stein+status:open+label:Code-Review=2 | |
| 13:55:13 | mriedem | dansmith: can you hit https://review.opendev.org/#/c/680871/ and https://review.opendev.org/#/c/676498/ for rocky? | |
| 13:55:27 | artom | mriedem, you saw ya.wang's review on that NUMA fix? | |
| 13:55:34 | dansmith | I've got the rocky set up right now, are those two more important? | |
| 13:56:07 | mriedem | more important? | |
| 13:56:28 | mriedem | the former is a long-standing issue for baremetal instances, the latter is an lxc thing which is low priority | |
| 13:56:33 | mriedem | lxc+py3 | |
| 13:56:38 | mriedem | artom: not yet | |
| 13:56:41 | dansmith | mriedem: meaning you pointed me at two rocky reviews, just wondering if I can hit those when I get to them in the rocky queue or if I should do those first for some reason | |
| 13:56:42 | mriedem | artom: does it work? | |
| 13:56:48 | artom | mriedem, yeah, tl;dr, he's +1 | |
| 13:56:51 | mriedem | dansmith: oh, they're just ready | |
| 13:56:53 | dansmith | mriedem: or were you just spoon-feeding me? | |
| 13:56:54 | mriedem | meaning already merged in stein | |
| 13:57:07 | mriedem | yes, pureed stable food | |
| 13:57:37 | mriedem | artom: +2 | |
| 13:57:46 | mriedem | yaawang: thanks for testing https://review.opendev.org/#/c/684409/ | |
| 13:57:57 | artom | yaawang, yep, much appreciated :) | |
| 14:05:20 | efried | stephenfin: done. Do you have any paste savvy? | |
| 14:05:35 | efried | dansmith: or you? | |
| 14:05:37 | stephenfin | very little, unfortunately | |
| 14:05:43 | dansmith | not really, no | |
| 14:05:50 | dansmith | I'm always mystified by that stuff | |
| 14:06:12 | efried | damn and blast | |
| 14:07:12 | dansmith | what is the question? | |
| 14:07:16 | efried | google seems clueless as well | |
| 14:07:29 | efried | dansmith: this https://review.opendev.org/#/c/685181/ | |
| 14:08:03 | dansmith | is this a thing we want to do? | |
| 14:08:15 | efried | tldr version discovery endpoints shouldn't require auth, but I don't know how to make /v2.1 use a different paste composite from /v2.1/$everything_else without enumerating $everything_else | |
| 14:08:25 | dansmith | like, should I be able to probe vexxhost to figure out what version it's running with no credentials? | |
| 14:08:31 | efried | absolutely | |