| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2020-04-08 | |||
| 07:40:58 | brinzhang | gibi: although we merged that patch, but the config_drive issue also exist, what do you want to plan to do? | |
| 07:42:54 | gibi | brinzhang: there is a Nova API bug regarding config_drive filtering, that I don't want to solve this week due to FF and due to the fact that it would need a bigger API change to make it consistent across the API. Regarding the client side, the client now works as a pass through which is not wrong but stephenfin argued for a bit nicer CLI interface | |
| 07:43:40 | gibi | alex_xu: I will try but no promisses | |
| 07:44:02 | alex_xu | gibi: thanks | |
| 07:45:14 | brinzhang | gibi: ok, I left some comment in bug 1871409, and I looked your chat records, yes, it's a bigger change | |
| 07:45:15 | openstack | bug 1871409 in OpenStack Compute (nova) "cannot filter the list servers to show only those that does not have config drive" [Medium,In progress] https://launchpad.net/bugs/1871409 - Assigned to Balazs Gibizer (balazs-gibizer) | |
| 07:45:40 | brinzhang | gibi: we are decide to completed this in gmann's next cleanup api? | |
| 07:45:53 | gibi | brinzhang: no, we only added it to the list | |
| 07:47:02 | brinzhang | gibi: ok, thanks, got it, if need my help that I can. | |
| 07:47:50 | gibi | brinzhang: thanks. I would not rush on that as config_driver filtering is broken at least in the last 3 years and nobody filed a bug | |
| 07:48:16 | gibi | so we will take the time to figure out how we want to fix it | |
| 07:49:38 | brinzhang_ | gibi: Yes, we rarely use config_drive to filter servers too. | |
| 07:54:33 | brinzhang_ | gibi: the bp/destroy-instance-with-datavolume separate patch updated and the novaclient also update, if you have time today, pls review again, I am sorry to add your work, thanks. | |
| 08:00:13 | gibi | brinzhang_: I have comment about the unit test https://review.opendev.org/#/c/711194 | |
| 08:01:11 | gibi | now I will look at the client patch | |
| 08:01:24 | brinzhang_ | gibi: thanks, will update | |
| 08:05:09 | brinzhang_ | gibi: the test case that I missed the api_version=2.85, and I will add d-o-t to the request body, and set 2.85 in request | |
| 08:09:12 | bauzas | gibi: morning, FWIW I worked last night on the functional test, even if it's still a WIP, we verify whether traits work for two pGPUs https://review.opendev.org/#/c/717975/8/nova/tests/functional/libvirt/test_vgpu.py@214 | |
| 08:09:29 | bauzas | and Zuul says +1 :p | |
| 08:10:30 | bauzas | (WIP because I need to update some tests for the fakelibvirt) | |
| 08:12:35 | gibi | bauzas: morning. ack | |
| 08:16:02 | openstackgerrit | Brin Zhang proposed openstack/nova master: Separate update and swap volume policies https://review.opendev.org/711194 | |
| 08:16:13 | brinzhang_ | gibi: the test update done, thanks | |
| 08:19:22 | gibi | brinzhang_: could you quickly fix this nit? then I will +2 https://review.opendev.org/#/c/711194/18/nova/tests/unit/policies/test_volumes.py@199 | |
| 08:19:40 | bauzas | stephenfin: fwiw, left comments on https://review.opendev.org/#/c/708436/ but I'm mostly +2 | |
| 08:20:34 | gibi | brinzhang_: two nits actually | |
| 08:20:36 | openstackgerrit | Brin Zhang proposed openstack/nova master: Separate update and swap volume policies https://review.opendev.org/711194 | |
| 08:21:46 | gibi | brinzhang_: thanks | |
| 08:22:45 | openstackgerrit | Brin Zhang proposed openstack/nova master: Separate update and swap volume policies https://review.opendev.org/711194 | |
| 08:22:46 | brinzhang_ | gibi: yes, that caused for me, I am sorry to that unexpected errors. | |
| 08:23:03 | bauzas | stephenfin: also I wonder whether we should document the fact that some keys having namespaces would be strictly validated | |
| 08:23:16 | bauzas | stephenfin: inhttps://review.opendev.org/#/c/710037/ | |
| 08:24:26 | brinzhang_ | gibi: maybe you should +2 again, I missed your +2, my network is not very good | |
| 08:24:33 | gibi | brinzhang_: done :) | |
| 08:24:45 | brinzhang_ | gibi: thanks | |
| 08:29:02 | bauzas | stephenfin: also, please help me, I don't get how you autodoc https://6319631a8afc7845c7f6-af2016a5632320f910deb9dcbf495ac6.ssl.cf2.rackcdn.com/710037/13/check/openstack-tox-docs/eb7469b/docs/configuration/extra-specs.html | |
| 08:29:07 | brinzhang_ | gibi: about the base patch of bp/destroy-instance-with-datavolume, I have some nits and lack some test case pre v2.85 need to follow up, and I will do that tomorrow, that I am not forget ^^ | |
| 08:29:46 | gibi | brinzhang_: if there is only nits and test then such change can go in after FF | |
| 08:30:32 | bauzas | stephenfin: I mean, i understood this was done by https://review.opendev.org/#/c/710037/13/doc/ext/extra_specs.py | |
| 08:30:36 | brinzhang_ | gibi: ok, thanks for remind, that's better for me | |
| 08:31:31 | bauzas | stephenfin: actually, nevermind I found it | |
| 08:43:27 | stephenfin | bauzas: cool, let me know if you still have issues with it. I wrote a good guide for the Sphinx doc on this thing https://www.sphinx-doc.org/en/master/development/tutorials/recipe.html | |
| 08:43:46 | stephenfin | though it's a bit more complicated than what we're doing | |
| 08:46:43 | openstackgerrit | Lee Yarwood proposed openstack/nova master: api: Introduce microverion 2.87 allowing boot from volume rescue https://review.opendev.org/701430 | |
| 08:46:44 | openstackgerrit | Lee Yarwood proposed openstack/nova master: compute: Extract _get_bdm_image_metadata into nova.utils https://review.opendev.org/705212 | |
| 08:46:44 | openstackgerrit | Lee Yarwood proposed openstack/nova master: libvirt: Support boot from volume stable device instance rescue https://review.opendev.org/701431 | |
| 08:46:45 | openstackgerrit | Lee Yarwood proposed openstack/nova master: libvirt: Break up get_disk_mapping within blockinfo https://review.opendev.org/714962 | |
| 08:47:06 | lyarwood | stephenfin: ^ I did need the API samples after all for my microversion | |
| 08:47:18 | stephenfin | how come? | |
| 08:47:31 | stephenfin | why didn't I? :-\ | |
| 08:48:25 | bauzas | stephenfin: not really, I don't want to hold this | |
| 08:49:11 | lyarwood | stephenfin: my functional tests in nova/tests/functional/api_sample_tests/test_rescue.py are rerunning the sample tests against the microversion, slightly over the top but meh. | |
| 08:50:25 | gibi | stephenfin: do you feel that we need to do someting with the way 2.83 client support merged during the night? https://review.opendev.org/#/c/713089/ | |
| 08:51:02 | stephenfin | gibi: We don't *have* to, but I'm drafting the patch now to see if it makes sense :) | |
| 08:51:38 | gibi | stephenfin: that patch would be top of the current merged solution or would revert the current solution first? | |
| 08:51:51 | stephenfin | I think on top would be wisest? | |
| 08:52:23 | gibi | stephenfin: yeah, I'm not too affraid of doing a breaking CLI change as nobody consumed the patch we merged during the night yet | |
| 08:52:38 | stephenfin | Yeah, exactly | |
| 08:52:47 | stephenfin | Catch it before it's released | |
| 08:53:05 | gibi | stephenfin: release will happen tomorrow night so we have some time | |
| 08:53:55 | gibi | stephenfin: I'm really glad that you have the time to tinker with that client patch, I feel a bit overwhelmed due to FF | |
| 08:56:00 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add test coverage of existing server policies https://review.opendev.org/718348 | |
| 08:56:15 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add test coverage of existing server policies https://review.opendev.org/718348 | |
| 08:59:33 | openstackgerrit | Stephen Finucane proposed openstack/python-novaclient master: Make 'server list --config-drive' a boolean option https://review.opendev.org/718349 | |
| 08:59:40 | stephenfin | gibi: Yeah, I just want to nip this one in the bud before it gets released and that's not possible | |
| 09:02:22 | johnthetubaguy | what is the REST API bug? | |
| 09:02:32 | bauzas | gibi: stephenfin: hum, just catching-up but I feel this somehow has to be cleared in between us : https://review.opendev.org/#/c/713089/ | |
| 09:03:03 | openstackgerrit | Stephen Finucane proposed openstack/python-novaclient master: Make 'server list --config-drive' a boolean option https://review.opendev.org/718349 | |
| 09:03:04 | bauzas | gibi: stephenfin: FWIW, we should ask not to merge nova api changes unless there are novaclient patches in Depends-ON | |
| 09:03:29 | johnthetubaguy | bauzas: really we should ask for openstack cli right? | |
| 09:03:40 | bauzas | johnthetubaguy: OSC is more problematic in my mind | |
| 09:03:51 | bauzas | but yeah we *should* | |
| 09:04:01 | johnthetubaguy | agreed, I spend hours with customers telling them which commands are totally broken in there | |
| 09:04:09 | stephenfin | johnthetubaguy: 'Instance.config_drive' is a string field but we treat it as sort of a boolean, storing the string 'True' for true and the empty string for False | |
| 09:04:11 | bauzas | at least, a depends-on should be there for a microversion patch | |
| 09:04:32 | stephenfin | johnthetubaguy: only the API rejects '?config_drive=' (the empty string) | |
| 09:04:46 | gibi | bauzas: do you mean the API change depends-on the client patch not the other way around? | |
| 09:05:06 | stephenfin | so it's not possible to say "give me all servers without a config drive" | |
| 09:05:10 | bauzas | gibi: indeed, because of the release cadence | |
| 09:05:22 | johnthetubaguy | stephenfin: and you can't do 'False' I guess? | |
| 09:05:38 | johnthetubaguy | because empty string is falsey | |
| 09:05:54 | stephenfin | nope, because whatever we provide is passed straight through to the DB query on the server side | |
| 09:06:10 | stephenfin | and I guess an exact match lookup done | |
| 09:06:15 | bauzas | gibi: anyway, just a can of worms for now | |
| 09:06:30 | bauzas | and I don't want to go to the rabbit cave | |
| 09:06:45 | gibi | bauzas: could you please add you thinking to the etherpad https://etherpad.openstack.org/p/nova-victoria-ptg around L84 | |
| 09:06:48 | stephenfin | that's my understanding of it, anyway. gibi can correct me if I'm wrong, though he's probably correct in saying we should wait til after feature freeze to work on this bugfix | |
| 09:06:57 | bauzas | but I'm afraid we would run against client changes if we don't do it | |
| 09:07:00 | gibi | stephenfin: you are correct | |
| 09:07:05 | stephenfin | \o/ | |
| 09:07:24 | bauzas | gibi: and the fact that stephenfin takes time to write the client changes instead of the BP owner makes me think that I'm right | |
| 09:07:25 | gibi | stephenfin: I would also add that this bug exists at least since 2017 | |
| 09:08:01 | gibi | bauzas: it is not exactly that. the owner wrote a change, we even merged a fixed version of it, stephenfin just has an improvement idea on the CLI interface | |
| 09:08:11 | bauzas | gibi: writing client changes should be the responsibility of the bp owner, not the nova maintainers IMHO | |
| 09:08:13 | stephenfin | bauzas: tbc, the client changes were done. I'm just tweaking it because I didn't like how it was done | |
| 09:08:24 | bauzas | oh ok, then nevermind | |
| 09:08:29 | stephenfin | but my patch is by no means mandatory | |
| 09:08:48 | bauzas | cool then | |
| 09:08:54 | stephenfin | if anything, I'm at fault because lyarwood drafted the novaclient change for *my* series /o\ | |
| 09:09:22 | bauzas | again, it's just my personal thoughts, not a policy | |