| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-05-10 | |||
| 20:55:50 | arvindn05 | nice...thanks for the quick turnaround efried | |
| 20:57:22 | oomichi | mriedem: I didn't see http410 case and api-wg guideline doesn't mention about http410 case, so I was curious about that | |
| 20:58:06 | mriedem | oomichi: it's what we used when we removed os-cloudpipe and os-certificates, and it makes more sense then 404 | |
| 20:58:14 | mriedem | *than | |
| 20:58:33 | mriedem | 404 is what we used for capping / deprecating with microversions | |
| 20:58:48 | mriedem | technically we also use 404 when a route is not found | |
| 20:59:48 | oomichi | mriedem: yeah, 404 is my base thinking. Nice to have conversation widely | |
| 21:00:27 | mriedem | oomichi: feel free to bring it up in the ML thread | |
| 21:00:41 | mriedem | but at this point it would be weird to use 410 for those other removed apis but 404 for fping | |
| 21:00:45 | mriedem | IMO | |
| 21:01:23 | oomichi | OK, let me do that | |
| 21:07:19 | mriedem | sean-k-mooney: this is for you https://bugs.launchpad.net/nova/+bug/1770378 | |
| 21:07:19 | openstack | Launchpad bug 1770378 in OpenStack Compute (nova) "windows can‘t start when use vhostuser interface" [Undecided,New] | |
| 21:08:57 | mriedem | sean-k-mooney: i noticed that we have 0 docs in nova or neutron on using vhostuser for networking | |
| 21:08:59 | mriedem | is that somewhere else? | |
| 21:09:19 | mriedem | oh hmm here it is https://docs.openstack.org/neutron/latest/contributor/internals/ovs_vhostuser.html | |
| 21:09:48 | sean-k-mooney | mriedem: stephen added docs to the networking guide | |
| 21:10:06 | mriedem | right but https://docs.openstack.org/neutron/latest/search.html?q=vhostuser | |
| 21:10:48 | mriedem | oh b/c it's vhost-user | |
| 21:12:09 | sean-k-mooney | mriedem: this is where the docs used to be https://docs.openstack.org/neutron/pike/admin/config-ovs-dpdk.html | |
| 21:13:14 | sean-k-mooney | anyway regarding the bug you should be able to run windows guest on vhost-user ports however i do remember issues in the past due to virtio diriver versions | |
| 21:14:31 | sean-k-mooney | mriedem: this might be the fix for the issue http://dpdk.org/dev/patchwork/patch/15541/ | |
| 21:41:16 | mnaser | mriedem: https://review.openstack.org/#/c/565611/ ok for me to merge this or wait till https://review.openstack.org/#/q/Idea86d6bb648b1e6fef8813dbe569724ce81a750 is in? | |
| 21:41:38 | sean-k-mooney | mriedem: i have responded to https://bugs.launchpad.net/nova/+bug/1770378 and set it to incomplete until they respond with the version of ovs and dpdk they are using | |
| 21:41:38 | openstack | Launchpad bug 1770378 in OpenStack Compute (nova) "windows can‘t start when use vhostuser interface" [Undecided,Incomplete] | |
| 21:48:08 | openstackgerrit | Merged openstack/nova master: Move create_tap_dev to the new utility module. https://review.openstack.org/557835 | |
| 21:50:02 | mriedem | takashin: this is the event https://www.openstack.org/summit/vancouver-2018/summit-schedule/events/21749/missing-features-in-openstack-for-public-clouds | |
| 21:50:35 | takashin | mriedem: Thanks. I will attend it. | |
| 21:51:15 | mriedem | https://etherpad.openstack.org/p/YVR-publiccloud-wg-brainstorming | |
| 21:51:20 | mriedem | that's the only etherpad i'm aware of right now | |
| 21:53:06 | mriedem | sean-k-mooney: thanks for commenting on that vhostuser bug | |
| 21:53:16 | mriedem | mnaser: that's good to go now | |
| 21:53:36 | mnaser | mriedem: +w :) | |
| 22:03:12 | mriedem | melwitt: i saw this in triage today https://bugs.launchpad.net/nova/+bug/1770348 | |
| 22:03:12 | openstack | Launchpad bug 1770348 in OpenStack Compute (nova) "Unable to access console via dashboard" [Undecided,Incomplete] | |
| 22:03:13 | mriedem | mnaser: thanks | |
| 22:06:34 | openstackgerrit | Hongbin Lu proposed openstack/nova-specs master: Choose default network on ambiguity https://review.openstack.org/520247 | |
| 22:16:29 | melwitt | mriedem: looking | |
| 22:16:47 | openstackgerrit | Merged openstack/nova stable/queens: Migrate tempest-dsvm-multinode-live-migration job in-tree https://review.openstack.org/567612 | |
| 22:16:52 | openstackgerrit | Merged openstack/nova stable/ocata: Migrate tempest-dsvm-multinode-live-migration job in-tree https://review.openstack.org/567619 | |
| 22:21:49 | openstackgerrit | Merged openstack/nova stable/pike: log stale allocations as WARNING instead of DEBUG https://review.openstack.org/567271 | |
| 22:25:46 | openstackgerrit | Hongbin Lu proposed openstack/nova-specs master: Choose default network on ambiguity https://review.openstack.org/520247 | |
| 22:28:05 | melwitt | mriedem: that would have been my first question too, whether they're using devstack or a real deployment. what's probably wrong is they need to pass the cell nova.conf to the console proxy service instead of the old global nova.conf | |
| 22:29:42 | melwitt | I'm thinking maybe I need to make the release note better to call that out | |
| 22:31:11 | mrf | Hi | |
| 22:31:27 | mrf | does images_rbd_pool at nova.conf support multiple pools? | |
| 22:31:31 | openstackgerrit | Merged openstack/nova master: Make get_instance_objects_sorted() be smart about cells https://review.openstack.org/509003 | |
| 22:33:04 | melwitt | mrf: I don't believe so. mnaser could you confirm? ^ | |
| 22:33:15 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Add nova-slow job https://review.openstack.org/567697 | |
| 22:33:30 | mriedem | there's your slow job | |
| 22:33:32 | mnaser | mriedem: it does not unfortunately, it can only use a single rbd image backend | |
| 22:33:36 | mnaser | err | |
| 22:33:39 | mnaser | mrf: ** ^ | |
| 22:33:52 | melwitt | mriedem: awesome thanks | |
| 22:33:59 | melwitt | wznoinsk: ^ | |
| 22:36:35 | mriedem | ok stable backports for this https://review.openstack.org/#/q/Ib10081150e125961cba19cfa821bddfac4614408 are ready | |
| 22:36:38 | mriedem | but probably no one around | |
| 22:39:59 | mriedem | oh tonyb mtreinish ^ :) | |
| 22:42:52 | mrf | mmm i got a trouble with 2 computes nodes , maybe some could help each one got diferent images_rbd_pool | |
| 22:43:04 | mrf | and now i cant migrate instances beetween them | |
| 22:45:13 | tonyb | mriedem: Thanks, I'll review them after this meeting | |
| 23:02:45 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Fix 500 error while passing 4-byte unicode data https://review.openstack.org/407514 | |
| 23:03:04 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Adds view builders for keypairs controller https://review.openstack.org/347289 | |
| 23:06:11 | openstackgerrit | melanie witt proposed openstack/nova master: Avoid warning log when image not exist https://review.openstack.org/567522 | |
| 23:17:25 | mriedem | AHA! | |
| 23:17:50 | melwitt | what | |
| 23:18:15 | mriedem | https://github.com/openstack/nova/blob/master/nova/scheduler/filters/availability_zone_filter.py#L43 | |
| 23:18:18 | mrf | O_O | |
| 23:18:19 | mriedem | we don't have a regression | |
| 23:18:36 | mriedem | if you didn't request an az on server create, the az filter doesn't take the current instance az into account | |
| 23:18:37 | melwitt | well I'll be | |
| 23:19:13 | mriedem | and clutchbeyers already left | |
| 23:19:16 | melwitt | disaster averted | |
| 23:25:50 | mriedem | we still have the up-call and data migration to deal with :) | |
| 23:26:03 | mriedem | we still have a regression since pike, just an easier one to deal with | |
| 23:26:34 | mriedem | will push up this functional test since i bothered writing it, and then i'm off | |
| 23:29:40 | melwitt | :( wah wah | |
| 23:29:45 | melwitt | thanks mriedem | |
| 23:30:07 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Add implicit instance AZ + live migration test https://review.openstack.org/567701 | |
| 23:30:13 | mrf | mmm i got a trouble with 2 computes nodes , maybe some could help each one got diferent images_rbd_pool, looking virshdump it point to rbd pool, can it be manually change the xmlfile to the new pool and them move the object at ceph level? | |
| 23:32:02 | melwitt | mrf: maybe. you could try asking an advance ceph operator like mnaser or ask on the openstack-operators mailing list where people can respond | |
| 23:33:59 | mrf | thank so much | |
| 23:34:14 | melwitt | good luck | |
| 23:37:07 | openstackgerrit | Tsuyoshi Nagata proposed openstack/nova master: nova improvement of maximum attach volumes more than 26 vols https://review.openstack.org/567472 | |
| 23:43:44 | mrf | Why instance migration with blockmigration is not supported in a stopped vm? Cannot 'os-migrateLive' instance af02564c-8579-4226-96b7-99e3b650c06b while it is in vm_state stopped (HTTP 409) (Request-ID: req-aee842d9-593b-43fc-9487-e5eb191bc732) | |
| 23:48:06 | melwitt | mrf: just a guess, but I think the instance has to be running in order to do a live migration. that is, libvirt must need the instance running in order to do it | |
| 23:49:07 | mrf | yes but if i want only migrate storage of instance | |
| 23:49:16 | mrf | state can me started or stopped | |
| 23:49:30 | mrf | similar to other hypervisors like xenserver, hyperv, vmware | |
| 23:49:43 | mrf | you can do a storage migration without know if machine is live or stopped | |
| 23:50:29 | mrf | block-migration must not depend of machine state | |
| 23:51:54 | melwitt | maybe you want regular cold migration then? https://developer.openstack.org/api-ref/compute/#migrate-server-migrate-action | |
| 23:54:49 | mrf | it drops : [root@controller ~(keystone_admin)]# openstack server migrate --block-migration af02564c-8579-4226-96b7-99e3b650c06b | |
| 23:54:49 | mrf | --live must be specified if --block-migration or --disk-overcommit is specified | |
| 23:56:13 | melwitt | I think you don't need to pass --block-migration for a cold migrate | |
| #openstack-nova - 2018-05-11 | |||
| 00:01:30 | mrf | but it fails because destination host does not have the same rbd_images_pool | |
| 00:01:47 | mrf | i think so | |
| 00:01:58 | melwitt | oh, so this is the same problem you were asking about earlier | |
| 00:04:16 | melwitt | live migration being rejected in a stopped vm is related to how live migration maps to the underlying libvirt function, so you'd have to ask someone who knows more about libvirt as to why that's not allowed. I'd ask kashyap about it but he's EU time zone and not around right now | |
| 00:05:19 | mrf | thanks :) | |