Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-03-13
14:41:00 brtknr in the database?
14:41:09 brtknr probably...
14:41:22 brtknr it exists as an entity somewhere...
14:41:26 brtknr but its not deletable
14:41:45 mriedem can you actually check the nova_api.build_requests table for a record with that instance_uuid?
14:41:52 mriedem because it should be temporary
14:43:51 brtknr mriedem: im checking, one sec
14:44:58 brtknr mysql> select * from build_requests;
14:44:59 brtknr Empty set (0.00 sec)
14:46:49 openstackgerrit Matt Riedemann proposed openstack/nova stable/queens: DNM: testing tempest-slow fix for subnode config https://review.openstack.org/643059
14:47:37 brtknr mriedem: ive just deleted another server and its gone into the same state... still shows up under server list but cannot be deleted...
14:48:42 mriedem brtknr: so it's probably in a cell but the instance_mappings record in the nova_api database for it does not have cell_id set for some reason
14:48:51 mriedem check the nova_api.instance_mappings table for the instance_uuid in question,
14:49:01 mriedem the cell_id column should be set, but i'm guessing it's not in your case
14:49:44 mriedem that would suggest something is failing in your nova-conductor service during scheduling where the instance mapping is not updated to point at the cell in which the instance record actually exists
14:50:01 mriedem which is i'm guessing either cell0 or the 'nova' database (cell1) for you
14:57:20 openstackgerrit Merged openstack/nova stable/pike: [pike-only] Fix resize_instance rpcapi call https://review.openstack.org/603439
14:57:27 openstackgerrit Merged openstack/nova stable/pike: tox: Don't write byte code (maybe) https://review.openstack.org/636919
14:57:36 openstackgerrit Merged openstack/nova stable/pike: PCI: do not force remove allocated devices https://review.openstack.org/635074
14:57:46 openstackgerrit Merged openstack/nova stable/pike: Note the aggregate allocation ratio restriction in scheduler docs https://review.openstack.org/623552
14:58:47 dansmith wow, pike
15:15:10 mriedem dansmith: if you like those, mayhap you want to hit these other rocky +2'ed changes https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/rocky+label:Code-Review=2
15:16:25 dansmith no good deed
15:23:07 openstackgerrit Merged openstack/nova stable/rocky: Handle templated cell mappings in nova-status https://review.openstack.org/642611
15:26:14 dansmith mriedem: I got stopped yesterday with that bigger ironic change at the top there
15:27:06 mriedem just plow right through those concerns
15:29:33 openstackgerrit Balazs Gibizer proposed openstack/nova master: Refactor _heal_allocations_for_instance to make place for port healing https://review.openstack.org/637953
15:29:34 openstackgerrit Balazs Gibizer proposed openstack/nova master: nova-manage: heal port allocations https://review.openstack.org/637955
15:29:34 openstackgerrit Balazs Gibizer proposed openstack/nova master: Refactor _heal_allocations_for_instance (2) https://review.openstack.org/637954
15:29:35 openstackgerrit Balazs Gibizer proposed openstack/nova master: cache neutron ports in heal allocation https://review.openstack.org/638207
15:43:00 openstackgerrit Merged openstack/nova master: Remove duplicate cleanup in functional tests https://review.openstack.org/636996
15:47:38 mriedem gibi: i went over the docs for bw providers https://review.openstack.org/#/c/642064/ some comments inline
15:47:46 mriedem let me know if you want to discuss my comments
15:47:52 gibi mriedem: thanks. looking
15:49:41 gmann melwitt: mriedem nova-next is failing with same reason - http://logs.openstack.org/17/639017/8/check/nova-next/21327c4/logs/screen-n-cpu.txt.gz#_Mar_13_14_31_14_886123
15:51:26 mriedem gmann: i don't think that's the same,
15:51:29 mriedem and i don't think nova-next is multinode
15:52:06 mriedem http://logs.openstack.org/17/639017/8/check/nova-next/21327c4/logs/libvirt/qemu/instance-00000001.txt.gz
15:52:11 gmann mriedem: not same as tempest-slow. i mean cert issue which was occurring early starting from when migrating it to bionic
15:52:13 mriedem -vnc 127.0.0.1:0,tls,x509verify=/etc/pki/libvirt-vnc: Failed to start VNC server: Cannot load certificate '/etc/pki/libvirt-vnc/server-cert.pem' & key '/etc/pki/libvirt-vnc/server-key.pem': Error while reading file.
15:52:21 gmann yeah this one.
15:52:42 gmann melwitt: was trying it with 1024 bit.
15:53:02 mriedem ok i don't know about the vnc / cert stuff
15:53:21 gmann mriedem: sorry for mixing it. it is for https://review.openstack.org/#/c/639017/
15:55:37 cfriesen I think this should be a fairly uncontroversial bugfix (has +2 from efried): https://review.openstack.org/#/c/641480/
15:58:26 mriedem dansmith: you'll never believe it by the $11 lady won
15:58:29 mriedem *but the
15:58:36 mriedem i've been carole'ed
15:58:47 dansmith mriedem: now you have to explain to maya why she's not an idiot after all?
15:58:56 mriedem i think she's sleeping
15:59:12 dansmith heh
16:00:28 mriedem too much colonial penn commercials
16:00:40 mriedem and oxygen tanks getting you down
16:01:16 dansmith is the oatmeal guy still double-dipping on insurance commercials?
16:02:12 dansmith oh I guess colonial penn was someone else
16:02:18 mriedem alex trebek
16:02:20 mriedem RIP
16:02:29 mriedem oatmeal was wilford brimley
16:02:32 mriedem from "the thing" fame
16:03:01 mriedem https://www.youtube.com/watch?v=3qSbYEbtfoY
16:03:05 dansmith yeah but he also did Liberty Mutual insurance, another daytime TV staple
16:03:07 mriedem here is now doing heavy computer work
16:04:52 mriedem note the (1) empty bottle of vodka on the desk and (2) .38 special in the drawer
16:05:21 dansmith I have never seen this movie, but the highly accurate 1980s computer simulation makes me think I'm missing out
16:05:28 mriedem i think that's a 38 special anyway, it's been awhile since i've listened to lynard skynard
16:05:37 mriedem the thing is one of my all time favorites
16:06:10 mriedem the 80s version of course
16:06:10 dansmith I don't recall them providing details of .38 special identification
16:06:41 mriedem i guess i'm thinking of saturday night special
16:06:51 dansmith I believe there was something about technique.. perhaps holding on loosely.. but that wasn't skynard
16:06:56 mriedem are those equivalent hand guns?
16:07:21 mriedem nobody likes the actual band "38 special"
16:07:22 dansmith I believe that skynard was indeed making a reference to a .38 calibre handgun in that song yes
16:07:39 mriedem I WIN!
16:07:53 mriedem cfriesen: is that maybe due to a newer libvirt changing the type of error returned in that case?
16:12:55 cfriesen mriedem: as far as I can tell the libvirt line has been there since like 2012: https://github.com/libvirt/libvirt/blob/master/src/qemu/qemu_monitor_json.c#L392
16:13:23 gibi mriedem: if I understand you correctly then in the bandwidth doc patch then a possible split would be to only describe the API limitation in the API guide and move the rest of the doc to the admin doc of nova
16:13:53 mriedem cfriesen: yeah i saw the notes in the bug report, +W
16:14:10 mriedem gibi: yeah more or less
16:14:43 mriedem gibi: the api guide tends to be about the non-admin API end user experience, and the stuff about allocation candidates and granular group_policy stuff and flavors and virt drivers should probably be somewhere in the admin docs
16:15:12 mriedem gibi: maybe in here https://docs.openstack.org/nova/latest/admin/adv-config.html
16:16:19 gibi mriedem: make sense, let me restructure the patch
16:17:23 openstackgerrit Matt Riedemann proposed openstack/nova master: api-ref: Add description for BDM volume_size https://review.openstack.org/641255
16:30:29 openstackgerrit guang-yee proposed openstack/nova stable/rocky: pass endpoint interface to Ironic client https://review.openstack.org/643098
16:41:34 mriedem dansmith: related to that az discussion we had yesterday, i realize we probably have another regression with resize revert,
16:41:47 mriedem when resizing, conductor will change the instance.az to whatever az the dest host is in,
16:41:58 mriedem but on revert we don't set that back to whatever the source host is in, if it changed
16:42:06 mriedem which technically it could if the server was created w/o a specific az
16:43:50 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Refactor functional base test classes https://review.openstack.org/643101
16:56:19 gmann melwitt: mriedem nova-grenade-live-migration job also failing on bionic due to no ceph hammer for bionic - http://logs.openstack.org/17/639017/8/check/nova-grenade-live-migration/735242b/ara-report/result/8816413c-7053-418f-886b-e238fdea1ec5/
16:57:23 gmann to proceed on bionic migration we can 1. disable tls-proxy in nova-next 2. make nova-grenade-live-migration n-v till frickler finds out the reason if usage of ceph hammer
16:57:43 gmann if ^^ is ok for you, i can update the patch.
16:58:37 gmann i have planned the bionic migration ~ 2PM PT so we need to make nova change before that to avoid the gate block
16:59:36 dansmith mriedem: yeah.. looking at the az stuff more recently has me thinking we've been neglecting that stuff too much in past years
17:04:06 mriedem gmann: 1 and 2 are mutually exclusive problems correct? do we have bugs reported for those?
17:04:31 mriedem gmann: is it not an option to keep these nova-specific jobs as voting and xenial until the issues are sorted out?
17:04:50 mriedem this is kind of a bad time to have to deal with this right before stein RC1
17:07:12 openstackgerrit Balazs Gibizer proposed openstack/nova master: Documentation for bandwidth support https://review.openstack.org/642064
17:08:05 gibi mriedem: fyi ^^
17:08:39 gibi mriedem: don't rush to rereview as I'm leaving for today

Earlier   Later