| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-03-01 | |||
| 14:28:44 | mriedem | bauzas: johnthetubaguy: ^ fyi on the docs thing | |
| 14:29:10 | johnthetubaguy | mriedem: oh, cool | |
| 14:31:15 | sdague | kevinbenton: so... rolling conversation on #openstack-qa that is related | |
| 14:31:17 | asettle | Hey mriedem sure have. We'd been discussing it before I went to lunch. I'd seen this patch before, but wasn't sure if it ticked all the boxes. | |
| 14:31:21 | asettle | Does this *deal* with everything? | |
| 14:31:56 | mriedem | i haven't read it yet in detail since it's a WIP | |
| 14:32:03 | mriedem | but i also pointed out we have these docs in the nova devref | |
| 14:32:04 | mriedem | https://docs.openstack.org/developer/nova/placement.html https://docs.openstack.org/developer/nova/cells.html | |
| 14:34:08 | asettle | mriedem: having it in nova dev ref is great, but we do need to have the docs in docs.o.o updated as its the user facing documentation. If people can't install from our site, it's a problem. | |
| 14:34:46 | mriedem | yeah i get that, but i also hope people trying to sort this out are reading our devref | |
| 14:35:07 | mriedem | i can poke through the docs change, but i also just got into the office so it's not going to happen right now | |
| 14:36:07 | asettle | No problem mriedem :) I'd like to start a bigger conversation about this too with docs liaisons for the base projects that we document but there does not appear to be a docs liaison for nova. Do you know of who would be able to fill this role? I am hoping to liaise and generate a better way for this content to be updated when necessary (to avoid running into an issue like this again :)) | |
| 14:36:10 | mriedem | i've added a couple of nova devs that helped write the docs that are in nova | |
| 14:36:27 | mriedem | we don't really have a docs liaison for stuff like this no | |
| 14:36:29 | asettle | Cool :) that will be very helpful, thank you | |
| 14:37:01 | asettle | mriedem: okay - well, we have a docs liaison for every project in OpenStack (give or take). It would be more than beneficial to have a nova representative to contact to assist dealing with documentation concerns. Could you help me find someone to fill that position? | |
| 14:37:17 | asettle | Previously mikal and sdague have helped out with the api-ref and api-guides | |
| 14:37:35 | mriedem | i can ask. fwiw, we've had mandatory liaisons in the past which are just a title but don't necessarily mean people are doing anything. | |
| 14:37:42 | mriedem | by default i'm the liaison | |
| 14:38:03 | mriedem | but i'll bring it up in our meeting | |
| 14:38:05 | asettle | mriedem: yeah, so, that's kind of what I'd like to fix :) the 'not doing anything' part ;) | |
| 14:38:16 | mriedem | i've love people to do things too :) | |
| 14:38:20 | mriedem | i can't force it though | |
| 14:38:23 | asettle | If you're okay with it, I'll add you into my email to the liaisons to discuss a good method of communication for the install guides? | |
| 14:38:31 | mriedem | sure | |
| 14:38:32 | asettle | Of course, community ;) | |
| 14:38:40 | asettle | Thanks! mriedem could you pm me your best email? :) | |
| 14:38:45 | mriedem | we could just sign up johnthetubaguy, he signs up for everything else | |
| 14:38:52 | mriedem | asettle: mriedem.os@gmail.com | |
| 14:39:05 | asettle | Hahah plus he's in my timezone ;) | |
| 14:39:14 | johnthetubaguy | mriedem: heh, thats like docs, sounds hard | |
| 14:39:26 | mriedem | wurdz r hrd | |
| 14:39:38 | johnthetubaguy | yup | |
| 14:40:40 | asettle | woah you guys are telling me | |
| 14:40:51 | asettle | People actually expect me to be eloquent as well | |
| 14:40:52 | asettle | It's outrageous | |
| 14:41:22 | mriedem | i did hear a british guy on a show the other night say "quintessential" at least 5 times | |
| 14:41:50 | artom_ | That's sort of the point of "quit" ;) | |
| 14:42:07 | artom_ | Or "quint" event | |
| 14:42:15 | asettle | mriedem: that's long. Over 2 syllables is out of my jurisdiction. | |
| 14:49:15 | openstackgerrit | Kevin Benton proposed openstack/nova master: Add some metadata logging to root cause ssh failure https://review.openstack.org/439560 | |
| 14:52:11 | jaypipes | kevinbenton: have you gone to sleep yet? | |
| 14:52:26 | jaypipes | kevinbenton: btw, I'm in sunnyvale... are you? | |
| 14:52:30 | kevinbenton | jaypipes: no, but i'm going to right now :) | |
| 14:52:34 | kevinbenton | jaypipes: yes i am! | |
| 14:52:54 | kevinbenton | jaypipes: catch up with you this afternoon and dinner? :) | |
| 14:52:58 | jaypipes | kevinbenton: the habit later? | |
| 14:53:12 | jaypipes | kevinbenton: yup, see ya in the office. | |
| 14:53:15 | kevinbenton | jaypipes: ack | |
| 15:03:30 | tojuvone | Guess we hit the PBR 2.0.0 thingy with pep8 | |
| 15:05:02 | openstackgerrit | Lee Yarwood proposed openstack/nova master: WIP objects: Add attachment_id to BlockDeviceMapping https://review.openstack.org/437665 | |
| 15:05:02 | openstackgerrit | Lee Yarwood proposed openstack/nova master: WIP compute: Move detach logic from manager into driver BDM https://review.openstack.org/439520 | |
| 15:07:20 | openstackgerrit | Timofey Durakov proposed openstack/python-novaclient master: Deperecate cell_name cli arg for migration-list https://review.openstack.org/439141 | |
| 15:12:05 | mriedem | sfinucan: nevermind, we can take your change to deprecate the firewall_driver option and i'll rebase my change on top of it to change the default to the noop driver | |
| 15:12:18 | mriedem | that will work nicely, yes very nice indeed....muwahahah | |
| 15:13:03 | sileht | Hi nova devs, any idea why we get: http://logs.openstack.org/10/439510/4/check/gate-telemetry-dsvm-integration-gnocchi-ubuntu-xenial/07bd989/logs/screen-n-api.txt.gz#_2017-03-01_14_07_46_133 | |
| 15:13:27 | edmondsw | johnthetubaguy could you take a quick look at https://review.openstack.org/#/c/389314 ? Other commenters are holding it up waiting for your feedback | |
| 15:13:41 | mriedem | sileht: did nova-manage db sync run after the cell0 db was mapped? | |
| 15:13:54 | mriedem | i.e. run nova-manage cell_v2 map_cell0 and then run nova-manage db sync | |
| 15:13:56 | sileht | mriedem, we just setup devstack | |
| 15:14:05 | sileht | mriedem, it should do it for us, no ? | |
| 15:14:20 | mriedem | well let's see | |
| 15:14:38 | sileht | mriedem, here: http://logs.openstack.org/10/439510/4/check/gate-telemetry-dsvm-integration-gnocchi-ubuntu-xenial/07bd989/logs/devstacklog.txt.gz#_2017-03-01_14_07_18_983 | |
| 15:14:39 | mriedem | nova_cell0 is created here http://logs.openstack.org/10/439510/4/check/gate-telemetry-dsvm-integration-gnocchi-ubuntu-xenial/07bd989/logs/devstacklog.txt.gz#_2017-03-01_13_57_36_563 | |
| 15:14:52 | mriedem | map_cell0 is called here http://logs.openstack.org/10/439510/4/check/gate-telemetry-dsvm-integration-gnocchi-ubuntu-xenial/07bd989/logs/devstacklog.txt.gz#_2017-03-01_14_07_18_983 | |
| 15:15:18 | mriedem | nova-manage db sync is called here http://logs.openstack.org/10/439510/4/check/gate-telemetry-dsvm-integration-gnocchi-ubuntu-xenial/07bd989/logs/devstacklog.txt.gz#_2017-03-01_13_57_36_581 | |
| 15:15:28 | mriedem | so db sync was called before map_cell0 | |
| 15:15:36 | mriedem | so the cell0 db schema isn't migrated | |
| 15:15:52 | sileht | mriedem, so it's a devstack bug ? | |
| 15:15:58 | mriedem | "WARNING: cell0 mapping not found - not syncing cell0." | |
| 15:16:15 | mriedem | sileht: if it were, i'd think all of the jobs would be broken | |
| 15:16:46 | dims | mriedem : tools/discover_hosts.sh does not run in their case | |
| 15:16:46 | sileht | mriedem, our gate is mostly broken, most of the time we got the error, but sometimes it passes | |
| 15:17:00 | dims | mriedem : so they don't run "nova-manage cell_v2 discover_hosts --verbose" | |
| 15:17:20 | sileht | dims, do we miss something in project-config ? | |
| 15:17:49 | dims | sileht : remember this one - http://git.openstack.org/cgit/openstack-dev/devstack/commit/tools/discover_hosts.sh?id=5c6aa56e11b2dc79adc70154187d3a029e8e0976 | |
| 15:18:17 | mriedem | dims: this isn't a multinode job so i don't think they need discover_hosts as long as simple_cell_setup is getting run | |
| 15:18:21 | sileht | dims, yes, but that was for another job | |
| 15:18:27 | mriedem | which is does here http://logs.openstack.org/10/439510/4/check/gate-telemetry-dsvm-integration-gnocchi-ubuntu-xenial/07bd989/logs/devstacklog.txt.gz#_2017-03-01_14_07_22_457 | |
| 15:18:37 | dansmith | missing cell0 means we're not even getting to discover hosts | |
| 15:18:42 | dims | ah, ignore me then | |
| 15:18:46 | mriedem | sileht: so someone else had a similar sounding issue with using a devstack plugin | |
| 15:18:55 | mriedem | where it was getting called in the wrong order, i think it was shade | |
| 15:19:14 | mriedem | mordred: remember that thing where a devstack plugin was being called from post-stack or test hooks or something? | |
| 15:19:24 | mriedem | so n-api was up before cell0 was synce | |
| 15:19:28 | mriedem | *synced | |
| 15:19:30 | openstackgerrit | Ed Leafe proposed openstack/nova master: placement: ironic inventory reporting https://review.openstack.org/437602 | |
| 15:19:50 | edleafe | jaypipes: ^^ fixed the pep8, and took the liberty of changing the name | |
| 15:20:14 | sileht | mriedem, oh maybe because each time a project that add a new wsgi vhost to apache, it is restarted | |
| 15:20:28 | asettle | mriedem: you are the lucky receiver of the table update; https://wiki.openstack.org/wiki/CrossProjectLiaisons#Documentation (this is where we keep all our liaisons) | |
| 15:21:00 | mordred | mriedem: yah | |
| 15:21:05 | mordred | mriedem: that was a lot of fun | |
| 15:21:31 | mriedem | mordred: remember the details? sounds like gnocchi jobs might be doing the same | |
| 15:21:51 | mriedem | http://logs.openstack.org/10/439510/4/check/gate-telemetry-dsvm-integration-gnocchi-ubuntu-xenial/07bd989/logs/devstack-gate-post_test_hook.txt.gz | |
| 15:22:00 | mordred | yah ... one sec, lemme find our fix patch | |
| 15:22:32 | mordred | mriedem: https://review.openstack.org/#/c/427390/ | |
| 15:22:57 | mordred | so for us "start_nodepool" needs to not happen until the cloud is actually, you know, up and running | |
| 15:23:06 | mordred | we do that in test-config now | |
| 15:23:36 | mordred | sileht: ^^ | |
| 15:23:41 | mriedem | sileht: so the gnocchi test hook might be trying to hit nova-api at the wrong point in the bring up sequence | |