| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-11-18 | |||
| 15:26:29 | mriedem | yeah | |
| 15:36:34 | sean-k-mooney | bauzas: could you take a look at https://review.opendev.org/#/q/status:open+project:openstack/nova+branch:master+topic:bug/1763766 ideally im aiming to get this merged before m1 | |
| 15:37:20 | sean-k-mooney | stephenfin: similar for https://review.opendev.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/vm-scoped-sriov-numa-affinity if i can get that merged before m1 it would make peopel happy so if you could review it would be nice before i start asking others | |
| 15:59:02 | bauzas | sean-k-mooney: ack, will look | |
| 16:10:13 | openstackgerrit | Mark Goddard proposed openstack/nova master: [ironic] Fix node rebalance race issues https://review.opendev.org/694802 | |
| 16:14:21 | mriedem | mgoddard: hi https://review.opendev.org/#/c/684849/ | |
| 16:14:38 | kashyap | Can someone update the openSUSE stuff here? -- https://wiki.openstack.org/wiki/LibvirtDistroSupportMatrix | |
| 16:14:46 | kashyap | (I've updated for Debian, Ubuntu, RHEL and Fedora.) | |
| 16:14:58 | kashyap | Who's the contact point for Oracle Linux again? | |
| 16:15:46 | mgoddard | mriedem: thanks, I'll check it out | |
| 16:18:31 | artom | kashyap, probably Satan | |
| 16:19:09 | artom | Dammit >< | |
| 16:19:10 | kashyap | artom: Yeah, c'mon | |
| 16:22:07 | mgoddard | mriedem: it could be because I'm using Rocky, but I don't see any errors in RT following rebalance | |
| 16:24:27 | mriedem | on either node? | |
| 16:24:33 | mriedem | s/node/compute service host/ | |
| 16:25:25 | mgoddard | yeah | |
| 16:27:20 | openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Add TODOs for remaining nova-network functional tests https://review.opendev.org/684345 | |
| 16:27:21 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'os-security-group-default-rules' REST API https://review.opendev.org/686807 | |
| 16:27:21 | openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unused '*_default_rules' security group DB APIs https://review.opendev.org/686808 | |
| 16:27:22 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove (most) '/os-networks' REST APIs https://review.opendev.org/686809 | |
| 16:27:22 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove '/os-tenant-networks' REST API https://review.opendev.org/686810 | |
| 16:27:23 | openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove 'USE_NEUTRON' from functional tests https://review.opendev.org/686811 | |
| 16:27:23 | openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove 'networks' quota https://review.opendev.org/686812 | |
| 16:27:24 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove nova-manage network, floating commands https://review.opendev.org/686813 | |
| 16:27:24 | openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove associate, disassociate network APIs https://review.opendev.org/686814 | |
| 16:27:25 | openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove 'nova-dhcpbridge' binary https://review.opendev.org/686815 | |
| 16:27:25 | openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove 'nova-network' binary https://review.opendev.org/686816 | |
| 16:27:26 | openstackgerrit | Stephen Finucane proposed openstack/nova master: docs: Blast most references to nova-network https://review.opendev.org/686817 | |
| 16:27:26 | openstackgerrit | Stephen Finucane proposed openstack/nova master: WIP https://review.opendev.org/686818 | |
| 16:28:19 | mriedem | mgoddard: your commit message says you have sql errors in the logs, likely due to duplicate entry errors? | |
| 16:29:08 | mgoddard | mriedem: I never really got to the bottom of that one. http://paste.openstack.org/show/786272/ | |
| 16:29:15 | mriedem | do you have this patch? https://review.opendev.org/#/c/676514/ | |
| 16:29:26 | mriedem | what version of rocky do you have? | |
| 16:29:49 | mgoddard | yes I have that patch | |
| 16:30:07 | mgoddard | I have 18.2.0, plus a few backports (including that one) | |
| 16:33:10 | mgoddard | mriedem: I have a few things to look at here. I could try my reproducer with your patch, and try your func test with my patch | |
| 16:35:41 | mriedem | yeah it would be good to know if they aren't duplicate bugs but something else | |
| 16:40:49 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Handle NoSuchOptError in nova-manage gracefully https://review.opendev.org/694806 | |
| 16:53:28 | mgoddard | mriedem: my patch passes the func test, with two modifications | |
| 16:53:36 | mgoddard | 1. I didn't hit "Skipping removal of allocations for deleted instances" | |
| 16:54:18 | mgoddard | 2. Part of my change prevents the orphan deletion happening in that case, so I rolled that part back for testing | |
| 16:54:29 | mgoddard | https://review.opendev.org/#/c/694802/1/nova/db/sqlalchemy/api.py | |
| 16:54:56 | mgoddard | ^ that prevents host2 from deleting a compute node after host1 has taken ownership of it | |
| 16:55:14 | mgoddard | which should prevent us from hitting this issue | |
| 16:55:29 | mgoddard | the rest of the change is to help clean up if we have hit it previously | |
| 16:59:24 | mriedem | mgoddard: ok i guess it's up to king solomon then to figure out which patch to go with | |
| 16:59:48 | mriedem | or just -1 my change if you want | |
| 17:00:25 | mriedem | either way i think the functional test needs to stay b/c unit tests aren't really sufficient for this kind of multi-service and db interaction | |
| 17:01:06 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Handle NoSuchOptError in nova-manage gracefully https://review.opendev.org/694806 | |
| 17:02:32 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Print help if nova-manage subcommand is not specified https://review.opendev.org/694806 | |
| 17:41:20 | openstackgerrit | Matt Riedemann proposed openstack/nova stable/stein: Use admin neutron client to see if instance has qos ports https://review.opendev.org/694668 | |
| 17:44:02 | artom | sean-k-mooney, left you things on https://review.opendev.org/#/c/687957/5 | |
| 17:50:37 | sean-k-mooney | just a couple :) | |
| 17:51:33 | sean-k-mooney | ill adress them tomorrow. over all are you ok with the patch | |
| 17:56:36 | artom | Yeah, once we've taken image difference out of the picture, I don't see issues with not running the NUMATopologyFilter on rebuilds | |
| 17:56:40 | artom | (Famous last words) | |
| 17:57:04 | artom | Might be good to get jparker to look at it | |
| 17:57:24 | artom | He came up with interesting test cases for NUMA LM, he might find a way to break it :) | |
| 17:57:36 | artom | Not necessarily the code itself, just the idea of it | |
| 17:58:31 | sean-k-mooney | well the idea is if we assert no numa toplogy constraits change in the api we dont need to use the filter to assert that again since rebuilds cant change host | |
| 17:58:41 | sean-k-mooney | so the existing claim should be valid | |
| 17:59:00 | artom | Yep :) I'm just paranoid | |
| 17:59:23 | sean-k-mooney | you could break the rebuild in other ways | |
| 17:59:31 | artom | Right, that's out of scope | |
| 17:59:36 | sean-k-mooney | e.g. adding a trait that is not present on the host | |
| 17:59:55 | openstackgerrit | Kashyap Chamarthy proposed openstack/nova master: [RFC] Pick NEXT_MIN libvirt/QEMU versions for "V" release https://review.opendev.org/694821 | |
| 17:59:59 | sean-k-mooney | ya im not trying to adress that although i hope we already do | |
| 18:00:05 | artom | ... do we check placement for rebuild? | |
| 18:00:36 | sean-k-mooney | i hope so but im not sure. its on my todo list to look at in general | |
| 18:01:01 | artom | Your ocean-spanning todo list :D | |
| 18:01:28 | sean-k-mooney | yes well when i fix things i keep findind other thngs that need to be fixed | |
| 18:02:24 | artom | https://xkcd.com/1739/ | |
| 18:02:24 | sean-k-mooney | like i found out that apparently we allow you to attach port that violate numa toplogy and you will only find out if you rebuild or otherwise have to go back to the schduler | |
| 18:02:41 | artom | Oh are those your comments on the second patch in that stack? | |
| 18:02:55 | sean-k-mooney | they are on a different patch one sec | |
| 18:03:58 | sean-k-mooney | artom: https://review.opendev.org/#/c/674072/7/nova/tests/unit/virt/test_hardware.py@4240 | |
| 18:04:09 | sean-k-mooney | no wrong one | |
| 18:04:18 | sean-k-mooney | also found that issue... | |
| 18:05:35 | sean-k-mooney | https://review.opendev.org/#/c/689861/3/nova/tests/functional/libvirt/test_numa_servers.py@1106 | |
| 18:06:42 | sean-k-mooney | i need to make that less wordy but ya that existing test was wrong | |
| 18:10:55 | artom | Or make it into its own patch? I haven't dug in yet, is what related to NUMA rebuilds? | |
| 18:11:02 | artom | (That last bit) | |
| 18:12:05 | sean-k-mooney | fixing the attachment api to reject interace if it would volate the numa aware vstichs feature is a spereate thing | |
| 18:12:25 | sean-k-mooney | at present it does not check and then a rebuild or resize to the same host woudl fail | |
| 18:13:48 | sean-k-mooney | i need to file a seperate bug and address it seperately | |
| 18:17:37 | artom | sean-k-mooney, I mean just putting the comments in a separate patch :) | |
| 18:17:43 | artom | Anyways, lemme review that properly | |
| 18:18:08 | sean-k-mooney | i needed to chagne the test so i added the comment to explain why | |
| 18:18:18 | artom | Ack | |
| 18:18:24 | artom | As I said, I hadn't dug in yet | |
| 18:18:55 | sean-k-mooney | i proably should change teh NOTE to a FIXME | |
| 18:21:59 | openstackgerrit | Archit Modi proposed openstack/nova stable/pike: doc: fix and clarify --block-device usage in user docs https://review.opendev.org/694450 | |
| 18:50:53 | mriedem | this is a good pike-only fix for something that was regressed with a backport https://review.opendev.org/#/q/topic:bug/1852207+(status:open+OR+status:merged) | |
| 19:01:36 | artom | sean-k-mooney, grah, forgot the backporting thing - might have left comments that make no sense in that context | |
| 19:24:42 | openstackgerrit | Matt Riedemann proposed openstack/nova master: FUP to Ie1a0cbd82a617dbcc15729647218ac3e9cd0e5a9 https://review.opendev.org/694787 | |
| 19:24:42 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Drop neutron-grenade-multinode job https://review.opendev.org/694789 | |
| 19:44:09 | rafaelweingartne | Hey guys, we are found on Nova that might be a bug | |
| 19:44:46 | rafaelweingartne | when a user is updating something in the VM (e.g. its name), this information is not updated in the KVM XML that is used to describe the VM in the compute host | |
| 19:44:52 | rafaelweingartne | is that expected? Or, is it a bug? | |