| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-03-16 | |||
| 09:48:48 | openstackgerrit | Surya Seetharaman proposed openstack/nova stable/queens: Update the nova-manage db archive_deleted_rows description https://review.openstack.org/553733 | |
| 09:49:18 | openstackgerrit | jichenjc proposed openstack/nova master: Move test_allocation_candidates.py https://review.openstack.org/553149 | |
| 09:49:19 | openstackgerrit | jichenjc proposed openstack/nova master: Move test_resource_provider.py https://review.openstack.org/553150 | |
| 09:49:19 | openstackgerrit | jichenjc proposed openstack/nova master: Move test_resource_class_cache.py https://review.openstack.org/553153 | |
| 09:50:43 | stephenfin | lyarwood: Yeah, we're not parsing the config option help text properly so only some of the formatting is being picked up | |
| 09:51:42 | stephenfin | Biggest issue I see to resolving this is all the broken config option descriptions that likely exist at this point. Guess I'll have to try handle any warnings that might be raised | |
| 10:01:01 | lyarwood | stephenfin: urgh, so that doesn't work, still seeing smartquotes | |
| 10:10:06 | openstackgerrit | Surya Seetharaman proposed openstack/nova stable/queens: Make archive_deleted_rows handle a missing CONF.api_database.connection https://review.openstack.org/553742 | |
| 10:12:24 | stephenfin | lyarwood: You need to do that in 'docutils.conf' | |
| 10:12:31 | stephenfin | lyarwood: See https://github.com/rtfd/readthedocs.org/issues/2940 | |
| 10:17:55 | lyarwood | stephenfin: huh | |
| 10:18:18 | lyarwood | stephenfin: FWIW I was using http://www.sphinx-doc.org/en/stable/config.html#confval-smartquotes | |
| 10:18:54 | stephenfin | lyarwood: "New in version 1.6.6" | |
| 10:19:34 | stephenfin | https://github.com/openstack/requirements/blob/master/global-requirements.txt#L421 | |
| 10:19:52 | lyarwood | *facepalm* | |
| 10:19:54 | stephenfin | I wasn't aware of that option but it seems we can't use it anyway | |
| 10:19:58 | stephenfin | :/ | |
| 10:20:14 | lyarwood | I might just leave it in the change anyway for when we do get above 1.6.6 | |
| 10:20:57 | stephenfin | lyarwood: That...could be a while. Some projects are seeing issues with it and I can't figure out why my "fix" won't pass CI https://review.openstack.org/#/c/552100/ | |
| 10:21:33 | stephenfin | lyarwood: If you're backporting, maybe use the older 'html' prefixed option, assuming we're using <1.6 on those branches? | |
| 10:23:00 | lyarwood | stephenfin: yup good point | |
| 10:24:39 | lyarwood | stephenfin: `while setting up extension ['?', 'sphinx.ext.todo']: node class 'todolist' is already registered, its visitors will be overridden` | |
| 10:24:47 | lyarwood | stephenfin: ^ tox -e docs - known issue? | |
| 10:25:09 | stephenfin | lyarwood: I've just spotted that too. That was me (commit 936404b90a954609f0c94582df0e2fd700948cea) | |
| 10:25:21 | stephenfin | Seems that feature (which I wrote) is broken in Sphinx | |
| 10:25:29 | stephenfin | my turn to *facepalm* | |
| 10:25:43 | stephenfin | I'm working on a fix as we speak. Gimme 5 | |
| 10:26:35 | stephenfin | lyarwood: To fix that, just revert said commit | |
| 10:27:28 | lyarwood | stephenfin: it's fine, it had already built the config.html page I care about | |
| 10:31:53 | openstackgerrit | Lee Yarwood proposed openstack/nova master: docs: Disable smartquotes https://review.openstack.org/553730 | |
| 10:33:08 | stephenfin | So oslo.config is doing what I'd expect it to do. Very odd... | |
| 10:33:32 | stephenfin | Namely, emitting the raw lines as its sees them and letting the docutils rST parser do its work | |
| 10:57:28 | openstackgerrit | Stephen Finucane proposed openstack/nova master: conf: Correct documentation for '[pci] passthrough_whitelist' https://review.openstack.org/552874 | |
| 10:57:28 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Follow the new PTI for document build https://review.openstack.org/553751 | |
| 10:57:35 | stephenfin | lyarwood: ^ | |
| 11:02:40 | lyarwood | stephenfin: still end up with smartquotes and unformatted text in config.html | |
| 11:02:50 | lyarwood | stephenfin: but the build passes now ;) | |
| 11:31:19 | stephenfin | lyarwood: Yeah, that's just to fix the broken local build. Did your patch work? | |
| 11:32:01 | stephenfin | My reading of that was that the old option was removed in 1.6.0 and then that alias was added in 1.6.6. For 1.6.0 and 1.6.5, the docutils config file was the only way | |
| 11:32:11 | stephenfin | But that's a guess | |
| 11:59:22 | lyarwood | stephenfin: it did locally, just waiting on check job running, still queued at the moment. | |
| 12:13:18 | openstackgerrit | Yikun Jiang (Kero) proposed openstack/nova-specs master: Complex (Anti)-Affinity Policies https://review.openstack.org/546925 | |
| 12:24:23 | lyarwood | stephenfin: http://logs.openstack.org/30/553730/2/check/build-openstack-sphinx-docs/28fd09c/html/configuration/config.html#pci - appears to be working | |
| 12:39:15 | openstackgerrit | Yikun Jiang (Kero) proposed openstack/nova master: WIP: complex policy https://review.openstack.org/553776 | |
| 12:45:31 | figleaf | ugh - tempest failed again. Is it worth rechecking yet again? | |
| 12:51:29 | tssurya | figleaf: guess not :( | |
| 12:51:48 | Spazmotic | heh | |
| 13:02:15 | Roamer` | a couple of years ago I proposed https://review.openstack.org/#/c/137830/ and then Things Happened(tm) and we stopped working on the StorPool Nova image backend. Looks like we'll write it after all (even if we have to maintain it out-of-tree); if I un-abandon that spec and propose it for inclusion, will somebody be very much against it? | |
| 13:10:30 | fried_rice | artom_: Thank you very much sir. | |
| 13:10:46 | fried_rice | alex_xu_: Thanks! | |
| 13:11:29 | fried_rice | figleaf: I win! | |
| 13:12:17 | figleaf | fried_rice: You probably sabotaged tempest so that you could win :) | |
| 13:12:17 | fried_rice | figleaf: To save some heartache, consider rebasing on top of https://review.openstack.org/#/c/548934/ (which is already in the gate) instead of master. | |
| 13:12:37 | fried_rice | figleaf: Oh, I totally did. I didn't know it was gonna *work* though. | |
| 13:13:11 | figleaf | fried_rice: that patch has a tempest failure, too | |
| 13:13:20 | figleaf | you forgot to un-bork tempest | |
| 13:16:59 | fried_rice | those volumes... | |
| 13:22:08 | openstackgerrit | Merged openstack/nova master: New-style _set_inventory_for_provider https://review.openstack.org/537648 | |
| 13:22:25 | mriedem | https://review.openstack.org/553786 | |
| 13:25:32 | figleaf | mriedem: <fingers crossed> | |
| 13:32:42 | mnaser | how do we feel about having 'ComputeCapabilitiesFilter' but instead parses image properties rather than instance type | |
| 13:33:27 | mnaser | (reason is, i have X customer which needs Y feature, i dont want to start maintaining instance types for every single one, but rather give the ability for the customer to add propreties) | |
| 13:36:11 | leakypipes | mnaser: https://review.openstack.org/#/c/541507/ will make your day. | |
| 13:37:20 | mnaser | leakypipes: i love how the "example" is exactly what we're looking for .. trait:HW_CPU_X86_AVX2=required | |
| 13:38:03 | mnaser | (like, the exact instruction set too) | |
| 13:38:43 | leakypipes | mnaser: bingo. | |
| 13:39:01 | mnaser | i guess i'll have to live with custom flavors till this code gets shipped | |
| 13:39:17 | leakypipes | mnaser: should be done in <2 months | |
| 13:39:34 | leakypipes | right arvindn05? ;P | |
| 13:39:59 | mnaser | well isnt it nice we have 6 month release cycles so i can pick it up in a release in 6 months max | |
| 13:40:06 | mnaser | ok it's friday i'll be nice | |
| 13:42:37 | leakypipes | mnaser: :) | |
| 13:46:08 | openstackgerrit | sahid proposed openstack/nova-specs master: virt: allow instances to be booted with trusted VFs https://review.openstack.org/485522 | |
| 13:51:11 | openstackgerrit | Ed Leafe proposed openstack/nova master: Add 'member_of' param to GET /allocation_candidates https://review.openstack.org/552098 | |
| 13:51:14 | figleaf | fried_rice: ^^ rebased | |
| 13:53:26 | openstackgerrit | Merged openstack/nova-specs master: Fix spec name for "report CPU features as traits" https://review.openstack.org/552159 | |
| 13:53:28 | mriedem | sahid: i see the problem in that linuxbridge multinode job | |
| 13:54:48 | mriedem | posted a new version | |
| 13:55:34 | leakypipes | figleaf: +2 | |
| 13:57:48 | sahid | mriedem: cool thanks, it's that change in "hosts", right? | |
| 13:58:59 | mriedem | yeah | |
| 13:59:13 | mriedem | subtle differences in the ansible playbook for a multinode job | |
| 14:00:14 | sahid | ok, i also working on backporting the change in one of our rhos version. so i hope we could do some internal tests | |
| 14:02:20 | stephenfin | lyarwood: Been looking at this oslo.config bug https://bugs.launchpad.net/nova/+bug/1755783 It seems nested_parse is getting confused with our output for that option http://paste.openstack.org/show/703014/ | |
| 14:02:22 | openstack | Launchpad bug 1755783 in OpenStack Compute (nova) "PCI configuration documentation renders with smart quotes" [Undecided,In progress] - Assigned to Stephen Finucane (stephenfinucane) | |
| 14:02:40 | stephenfin | lyarwood: So it's probably going to need a significant rewrite of that extension to unconfuse it :( | |
| 14:03:02 | stephenfin | (all that stuff in the <paragraph> tag should have been parsed itself but wasn't) | |
| 14:03:42 | dansmith | sahid: I thought you said you already tested this locally back in an earlier version | |
| 14:03:58 | dansmith | sahid: does that mean you haven't actually tested this manually yet? | |
| 14:04:27 | dansmith | specifically "I'm making tests locally and everything seemed to work" | |
| 14:04:46 | mnaser | are live migration job failures common-ish in gate? aka safe to recheck? | |
| 14:06:00 | fried_rice | figleaf: I'm sorry I'm just getting to this, but I've got some comments on that guy. | |
| 14:06:27 | figleaf | fried_rice: go for it | |
| 14:07:21 | sahid | dansmith: that was for a previous version of that patch | |
| 14:08:00 | dansmith | sahid: right, the version that _couldn't_ work...so you haven't tested the latest approach locally? | |
| 14:08:56 | fried_rice | figleaf, leakypipes: The thing so far that we might potentially want to hold up for is that the syntax isn't consistent with member_of in GET /resource_providers. Looks like this discussed earlier in the life of this patch, but not addressed. | |
| 14:09:16 | sahid | dansmith: no sure what you mean by the version that _counldn't work | |
| 14:09:23 | dansmith | heh | |
| 14:17:09 | figleaf | fried_rice: what are the differences that you see? | |
| 14:18:12 | fried_rice | figleaf: GET /resource_providers?member_of=in:<list of UUIDs> -- yours has no 'in:'. Personally I think 'in:' is silly and would like to see it go away. But it's inconsistent. But I don't like it. But it's inconsistent. I'm torn. | |