| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-07-20 | |||
| 09:52:49 | openstackgerrit | do3meli proposed openstack/nova master: docs: add nova host-evacuate command to evacuate documentation https://review.openstack.org/578040 | |
| 10:40:04 | Shilpa | stephenfin: Hi | |
| 10:40:09 | stephenfin | Shilpa: o/ | |
| 10:40:42 | Shilpa | ' | |
| 10:40:42 | Shilpa | stephenfin: on Nova side, for tox -e pep8, i am facing error 'ERROR: pep8: could not install deps [-r/opt/stack/nova/test-requirements.txt]; v = InvocationError('/opt/stack/nova/.tox/pep8/bin/pip install -chttps://git.openstack.org/cgit/openstack/requirements/plain/upper-constraints.txt -r/opt/stack/nova/test-requirements.txt (see /opt/stack/nova/.tox/pep8/log/pep8-1.log)', 1) | |
| 10:41:10 | Shilpa | stephenfin: i have done 'sudo pip install tox', but no sucess | |
| 10:41:20 | stephenfin | Shilpa: Could you share the full output on paste.openstack.org? | |
| 10:42:25 | Shilpa | stephenfin:http://paste.openstack.org/show/726332/ | |
| 10:43:09 | stephenfin | Shilpa: What distro are you using? You need to install python-dev(el) | |
| 10:43:33 | stephenfin | Shilpa: Sorry - python3-dev(el) | |
| 10:43:45 | stephenfin | ext/_yaml.c:4:20: fatal error: Python.h: No such file or directory | |
| 10:43:45 | stephenfin | x86_64-linux-gnu-gcc -pthread -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -g -fstack-protector-strong -Wformat -Werror=format-security -Wdate-time -D_FORTIFY_SOURCE=2 -fPIC -I/usr/include/python3.5m -I/opt/stack/nova/.tox/pep8/include/python3.5m -c ext/_yaml.c -o build/temp.linux-x86_64-3.5/ext/_yaml.o | |
| 10:43:45 | stephenfin | Per this error message | |
| 10:44:49 | Shilpa | stephenfin:any solution to resolve this? | |
| 10:45:14 | stephenfin | Shilpa: Yeah, install that package. 'sudo apt install python-dev' or the likes | |
| 10:47:19 | Shilpa | stephenfin:thanks, will check and let you know in next 5-10 min | |
| 10:52:23 | openstackgerrit | Chris Dent proposed openstack/nova master: [placement] Add /reshaper handler for POST https://review.openstack.org/576927 | |
| 11:09:18 | Shilpa | stephenfin: Thanks able to resolve the pepe8 issue, ' sudo apt-get install python3.5-dev' | |
| 11:34:07 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Start populating NUMACell.network_metadata field https://review.openstack.org/564441 | |
| 11:34:08 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Add functional tests for numa-aware-vswitches https://review.openstack.org/583288 | |
| 11:34:09 | openstackgerrit | Stephen Finucane proposed openstack/nova master: scheduler: Start utilizing RequestSpec.network_metadata https://review.openstack.org/564452 | |
| 11:34:10 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Assorted cleanups from numa-aware-vswitches series https://review.openstack.org/582651 | |
| 11:57:01 | openstackgerrit | Lei Zhang proposed openstack/nova master: Add method to get cpu traits https://review.openstack.org/560317 | |
| 12:05:16 | openstackgerrit | Andrey Volkov proposed openstack/nova master: Docs: Add Placement to Nova system architecture https://review.openstack.org/584338 | |
| 12:25:39 | openstackgerrit | Andrey Volkov proposed openstack/nova master: Docs: Add more colors to Nova system architecture https://review.openstack.org/584345 | |
| 12:44:08 | mriedem | gibi: can you also check the updated regression test? https://review.openstack.org/#/c/583339/ | |
| 12:59:25 | kashyap | mriedem: On that PowerISA thing, what Lei needs to provide is something like this to baselineCPU(): https://libvirt.org/git/?p=libvirt.git;a=blob_plain;f=tests/cputestdata/ppc64-baseline-same-model.xml | |
| 12:59:34 | kashyap | What he's providing right now is _guest_ CPU XML; hence the error. | |
| 12:59:51 | kashyap | I wrote a comment here with a precise example he can use: https://review.openstack.org/#/c/560317/27 | |
| 13:00:46 | mriedem | gmann: question in https://review.openstack.org/#/c/583538/ | |
| 13:00:57 | mriedem | gmann: and i'm fine if you want to take over the nova slow job stuff, i'm not working on it | |
| 13:01:41 | mriedem | alex_xu: maybe you can answer my question in https://review.openstack.org/#/c/583538/ | |
| 13:02:05 | mriedem | but if we're dropping the extensions, i don't know why we also aren't moving into the view builder code to centralize the response body creation | |
| 13:07:42 | openstackgerrit | Corey Bryant proposed openstack/nova master: In Python3.7 async is a keyword [1] https://review.openstack.org/584365 | |
| 13:17:27 | mriedem | kashyap: ok | |
| 13:20:30 | mriedem | gmann: alex_xu: i see several of these already merged https://review.openstack.org/#/q/topic:bp/api-extensions-merge-rocky+(status:open+OR+status:merged) but 2 things: | |
| 13:20:47 | mriedem | 1. rather than cram it all into that single create() method, it would be nice to instead continue the pattern we have in here https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/helpers.py#L55 | |
| 13:21:07 | mriedem | 2. this series isn't moving the view builder code from the extensions into the actual server ViewBuilder, which is something we should do | |
| 13:21:43 | mriedem | maybe the goal for rocky is to just remove the server_create extension method and then in stein we refactor/cleanup the create() controller method to move things into https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/helpers.py#L55 and also move stuff into the view builder? | |
| 13:22:12 | mriedem | personally it's the view builder that bothers me the most, because i have to hunt all over the extensions to find what puts something into the server response body | |
| 13:22:45 | mriedem | and the view builder part is going to become very important for tssurya's change for handling a down cell | |
| 13:27:53 | mriedem | i see in https://review.openstack.org/#/c/583882/ it says "Remaining part in this BP is to merge the response extensions which is last part of this work." - is that the view builder part? if so, then i guess we can just merge what you have in rocky and do the view builder (and move stuff out of create() into helpers) in stein | |
| 13:37:07 | openstackgerrit | Andrey Volkov proposed openstack/nova master: Docs: Add more colors to Nova system architecture https://review.openstack.org/584345 | |
| 14:03:59 | dansmith | finucannot: you see that the second patch in your series is failing on pep8 right? | |
| 14:04:16 | finucannot | superdan: I did not but I'll fix now | |
| 14:19:36 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Start populating NUMACell.network_metadata field https://review.openstack.org/564441 | |
| 14:19:37 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Add functional tests for numa-aware-vswitches https://review.openstack.org/583288 | |
| 14:19:38 | openstackgerrit | Stephen Finucane proposed openstack/nova master: scheduler: Start utilizing RequestSpec.network_metadata https://review.openstack.org/564452 | |
| 14:19:39 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Assorted cleanups from numa-aware-vswitches series https://review.openstack.org/582651 | |
| 14:21:22 | smcginnis | melwitt: Morning. Do you take care of osc-placement releases? Or is that delegated to someone else? | |
| 14:21:45 | melwitt | smcginnis: I usually do, yes | |
| 14:23:05 | melwitt | smcginnis: I had thought osc-placement would go along with client library releases since it's a plugin for a client, does it actually count as a non-client library? | |
| 14:23:33 | smcginnis | melwitt: Ah - right now it is flagged as a library. | |
| 14:23:57 | melwitt | oh :( | |
| 14:24:56 | mriedem | gmann: johnthetubaguy: am i remembering correctly that there was discussion about deprecating all of these dumb legacy server extension policy rules for things like whether or not we should 'config_drive' in the server response? those are all required in the response in our docs and tempest schema validation testing, so we should deprecate those policy rules. | |
| 14:24:57 | smcginnis | melwitt: No worries. When you do the next release for it you can switch over the "type" value to be "client-library" instead of "library". | |
| 14:25:27 | smcginnis | melwitt: Just wanted to make sure we weren't missing something since I saw it had a good handful of commits that have not been released. | |
| 14:29:37 | melwitt | smcginnis: yeah, thank you for pinging me about it. if I'd known it was "non-client library" I'd have proposed a release for it yesterday, for sure | |
| 14:32:19 | openstackgerrit | Lei Zhang proposed openstack/nova master: Add method to get cpu traits https://review.openstack.org/560317 | |
| 14:33:47 | smcginnis | melwitt: Makes sense that it should be a client lib if it's an osc plugin. Just miscategorized I'd say. | |
| 14:40:21 | melwitt | smcginnis: coolness. I'll make sure to update that type value next release (next week) | |
| 15:01:24 | mriedem | finucannot: +2s on the next 2 in the series for numa stuff with comments that can be addressed in the follow up patch | |
| 15:01:28 | mriedem | especially the functional test one | |
| 15:05:03 | finucannot | mriedem: Cool, I'll stick the fixes in with https://review.openstack.org/#/c/582651/ | |
| 15:05:13 | finucannot | Thanks for tackling those too. Much obliged | |
| 15:10:19 | openstackgerrit | Eric Fried proposed openstack/nova master: Check provider generation and retry on conflict https://review.openstack.org/556669 | |
| 15:29:27 | superdan | finucannot: any chance you could clean up those unrelated bits in the functional test before you go for the week? | |
| 15:29:36 | superdan | I really hate to merge something with _that_ much noise | |
| 15:29:47 | finucannot | superdan: Fair point. Yeah, lemme tackle that now | |
| 15:29:51 | superdan | okay thanks | |
| 15:36:23 | mriedem | finucannot: comments in https://review.openstack.org/#/c/564452/23 | |
| 15:36:30 | mriedem | you're missing all of the move operation stuff | |
| 15:36:40 | mriedem | which is probably a patch that needs to come *before* the API enables this for new servers | |
| 15:42:21 | mriedem | the move stuff probably requires 2 changes | |
| 15:42:35 | mriedem | 1. model network phsynes and tunneled in nova.network.model.Network which goes into the info cache | |
| 15:42:55 | mriedem | 2. the actual stuff in conductor that populates RequestSpec.network_metadata from the instance info cache prior to calling the scheduler to find a target host | |
| 15:56:00 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Add functional tests for numa-aware-vswitches https://review.openstack.org/583288 | |
| 15:56:01 | openstackgerrit | Stephen Finucane proposed openstack/nova master: scheduler: Start utilizing RequestSpec.network_metadata https://review.openstack.org/564452 | |
| 15:56:02 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Assorted cleanups from numa-aware-vswitches series https://review.openstack.org/582651 | |
| 16:00:15 | openstackgerrit | Lee Yarwood proposed openstack/nova master: libvirt: Remove reference to transient domain when detaching devices https://review.openstack.org/584433 | |
| 16:21:49 | openstackgerrit | Corey Bryant proposed openstack/nova master: In Python3.7 async is a keyword [1] https://review.openstack.org/584365 | |
| 16:25:56 | openstackgerrit | Stephen Finucane proposed openstack/nova master: scheduler: Start utilizing RequestSpec.network_metadata https://review.openstack.org/564452 | |
| 16:25:57 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Assorted cleanups from numa-aware-vswitches series https://review.openstack.org/582651 | |
| 16:33:42 | openstackgerrit | Pierre Blanc proposed openstack/nova master: Docs: Add guide to migrate instance with snapshot https://review.openstack.org/584442 | |
| 17:11:51 | tomtom001 | hello, does anyone know if the scheduler_default_filters=AllHostsFilter does anything? Are there any problems using it? | |
| 18:27:07 | mriedem | powerkvm ci is happy on this now https://review.openstack.org/#/c/560317/ | |
| 18:27:27 | mriedem | tomtom001: the AllHostsFilter is a noop | |
| 18:27:30 | mriedem | it accepts all hosts | |
| 18:28:25 | mriedem | superdan: on your bfv disk thing, i got the move / migrate patch up on top of that | |
| 18:28:37 | mriedem | fried_rolls also helped flush out some latent bugs in our disk usage reporting in the RT | |
| 18:29:10 | superdan | mriedem: oh awesome thanks, hadn't seen | |
| 18:29:16 | superdan | mriedem: since you're here.. on the down host thing... | |
| 18:29:35 | superdan | I was thinking -- can't we get away with the online migration without needing a marker record? | |
| 18:30:10 | superdan | we can query for max_limit records, where queued_for_delete is NULL and group_by or filter by cell to make the instance lookups more efficient | |
| 18:30:42 | superdan | we needed the marker for the initial roll-out of those records, because we didn't know where to pick up because we were creating them, | |
| 18:31:03 | superdan | but since we're just healing those that have no value set, we can use the NULL-ness to whittle that down I think | |
| 18:31:07 | mriedem | the migration is just copying how the reqspec migration was done | |
| 18:31:08 | mriedem | which uses a marker | |
| 18:31:28 | superdan | the creating of the reqspec? | |
| 18:31:30 | mriedem | yeah | |