| *** takashin has joined #openstack-placement | 00:10 | |
| *** tetsuro has joined #openstack-placement | 00:18 | |
| *** Sundar has quit IRC | 00:56 | |
| *** mgagne has quit IRC | 02:15 | |
| *** mgagne has joined #openstack-placement | 02:15 | |
| *** tetsuro has quit IRC | 03:58 | |
| *** tetsuro has joined #openstack-placement | 05:18 | |
| *** e0ne has joined #openstack-placement | 06:23 | |
| *** alex_xu has quit IRC | 06:23 | |
| *** alex_xu has joined #openstack-placement | 06:29 | |
| *** e0ne has quit IRC | 06:43 | |
| *** e0ne has joined #openstack-placement | 06:47 | |
| openstackgerrit | Boxiang Zhu proposed openstack/nova-specs master: Scheduler filters evaluated even forced host https://review.openstack.org/645458 | 06:51 |
|---|---|---|
| *** e0ne has quit IRC | 06:59 | |
| *** e0ne has joined #openstack-placement | 07:05 | |
| openstackgerrit | Tetsuro Nakamura proposed openstack/placement master: Prepare for negative member_of queryparam https://review.openstack.org/641391 | 07:42 |
| openstackgerrit | Tetsuro Nakamura proposed openstack/placement master: Prepare for negative member queryparam 2 https://review.openstack.org/641392 | 07:42 |
| openstackgerrit | Tetsuro Nakamura proposed openstack/placement master: Negative member_of query with microversion 1.32 https://review.openstack.org/641393 | 07:42 |
| *** e0ne has quit IRC | 07:45 | |
| *** e0ne has joined #openstack-placement | 07:48 | |
| *** e0ne has quit IRC | 07:51 | |
| *** e0ne has joined #openstack-placement | 07:52 | |
| *** e0ne has quit IRC | 07:52 | |
| *** e0ne has joined #openstack-placement | 07:54 | |
| openstackgerrit | Tetsuro Nakamura proposed openstack/placement master: Spec: Support filtering by forbidden aggregates https://review.openstack.org/647704 | 08:05 |
| *** takashin has left #openstack-placement | 08:06 | |
| *** e0ne has quit IRC | 08:17 | |
| *** helenafm has joined #openstack-placement | 08:48 | |
| *** ttsiouts has joined #openstack-placement | 09:00 | |
| openstackgerrit | Abhishek Kekane proposed openstack/nova-specs master: Support multiple backend of Glance https://review.openstack.org/641210 | 09:01 |
| *** ttsiouts has quit IRC | 09:10 | |
| *** tetsuro has quit IRC | 09:12 | |
| *** ttsiouts has joined #openstack-placement | 09:13 | |
| *** tssurya has joined #openstack-placement | 09:18 | |
| openstackgerrit | Surya Seetharaman proposed openstack/nova-specs master: Support adding the reason behind a server lock https://review.openstack.org/638629 | 09:48 |
| openstackgerrit | Surya Seetharaman proposed openstack/nova-specs master: Support adding the reason behind a server lock https://review.openstack.org/638629 | 10:05 |
| *** e0ne has joined #openstack-placement | 10:12 | |
| *** e0ne has quit IRC | 10:34 | |
| *** e0ne has joined #openstack-placement | 10:42 | |
| *** cdent has joined #openstack-placement | 11:08 | |
| *** e0ne has quit IRC | 11:42 | |
| *** e0ne has joined #openstack-placement | 11:54 | |
| *** e0ne has quit IRC | 11:56 | |
| *** e0ne has joined #openstack-placement | 12:03 | |
| *** e0ne has quit IRC | 12:07 | |
| *** e0ne has joined #openstack-placement | 12:12 | |
| *** e0ne has quit IRC | 12:37 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Fill in the New Features section of contributing.rst https://review.openstack.org/648113 | 12:39 |
| *** cdent has quit IRC | 12:40 | |
| *** e0ne has joined #openstack-placement | 13:36 | |
| *** cdent has joined #openstack-placement | 14:19 | |
| openstackgerrit | sean mooney proposed openstack/nova-specs master: add spec for image metadata prefiltering https://review.openstack.org/647578 | 14:27 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Fill in the New Features section of contributing.rst https://review.openstack.org/648113 | 14:46 |
| dansmith | cdent: edleafe efried: I thought that there were some "map things to traits" functions in os-traits, but I'm not seeing them | 14:48 |
| dansmith | is that appropriate to add there? | 14:48 |
| efried | dansmith: There's a normalize_name method | 14:49 |
| cdent | dansmith: I'm not sure I know what you mean? | 14:49 |
| efried | that generates a valid CUSTOM_* name from an input string. | 14:49 |
| dansmith | efried: that's for just generic stringification right? | 14:49 |
| dansmith | yeah | 14:49 |
| dansmith | cdent: I mean something that converts glance image format names to trait names | 14:49 |
| efried | there's nothing else I'm aware of that does "mapping" | 14:49 |
| efried | oh, yeah, I would expect that to live not in os-traits itself. | 14:49 |
| dansmith | okay I thought there was.. | 14:49 |
| dansmith | ack | 14:49 |
| efried | like the mapping for compute capabilities | 14:49 |
| efried | I could be wrong. | 14:50 |
| edleafe | that's definitely not an os-traits thing | 14:50 |
| cdent | I think we decided that os-traits was going to be almost entirely "just the symbols" | 14:50 |
| dansmith | thought someone said it would be in os-traits when we were discussing this image thing, since the image types are public from glance | 14:50 |
| dansmith | ack, just checking | 14:50 |
| cdent | and thus far that's how it has been | 14:50 |
| *** e0ne has quit IRC | 14:50 | |
| dansmith | ah, here: | 14:51 |
| dansmith | https://github.com/openstack/os-traits/blob/master/os_traits/hw/gpu/cuda.py | 14:51 |
| dansmith | that's enough to have some tests, and probably what I'm thinking of | 14:51 |
| dansmith | but anyway, I don't care where it is, just didn't want to put it in nova if ya'll wanted it next to the traits themselves | 14:51 |
| cdent | huh, that's news to me | 14:52 |
| cdent | dansmith: I'd say write it somewhere it works while making it work, and then when we see it if it looks os-traits stuff, we move it? | 14:52 |
| dansmith | sure.. the only reason to put it here is because it translates public glance api things to public os-traits things, which shouldn't change, but that's not the norm for other non-image things | 14:53 |
| dansmith | so putting it in nova makes plenty of sense I think | 14:53 |
| cdent | my original hope for os-traits and os-resource-classes has been that they wouldn't have any (beyond the symbols) that was only used by one project. So basically if a second project started wanting something, we'd put it it. But that might be too limiting. I don't care a huge deal one away or another. | 14:54 |
| edleafe | would any service other than nova need it? | 14:54 |
| edleafe | sorta jinx | 14:54 |
| openstackgerrit | Dan Smith proposed openstack/os-traits master: Add IMAGE_TYPE_SUPPORT_* traits https://review.openstack.org/648147 | 14:56 |
| dansmith | edleafe: yeah, I could see cinder using it | 14:56 |
| dansmith | so, do we have to merge that and do an os-traits release before I can use them in test nova patches? | 14:57 |
| cdent | depends on whether it is a tox-sibling or not | 14:58 |
| * dansmith stares blankly | 14:58 | |
| cdent | :) | 14:58 |
| * cdent goes looking | 14:59 | |
| cdent | dansmith: for unit tests, no. for functional tests you might be able to make it a tox-slbling by adding to the required-projects section in functional job def. In any case, i'd try a depends-on just to see what happens. I can't remember off the top of my head (that's been swapped out today) | 15:02 |
| *** e0ne has joined #openstack-placement | 15:02 | |
| dansmith | what do other people do in this case? can we just merge that patch before the nova stuff is really ready? | 15:03 |
| cdent | we can (and should) improve this | 15:03 |
| dansmith | should be pretty uncontroversial aside from the prefix, given this is already a public enum | 15:03 |
| cdent | yeah, merging trait changes early has been a common thing | 15:03 |
| dansmith | okay | 15:03 |
| cdent | gibi probably has some experience in this area | 15:03 |
| cdent | dansmith: commented on that patch, it's got one main problem, easy to fix | 15:13 |
| gibi | dansmith, cdent: do you want to have someting how placement is installed in nova functional test env? https://github.com/openstack/nova/blob/771e102654b54d257f437f16be2aeeef51fa0783/tox.ini#L91 | 15:15 |
| gibi | cdent: in the bandwidth case the related resource classes was still defined in nova when they was merged so in that case there was no dependency issue | 15:16 |
| *** e0ne has quit IRC | 15:49 | |
| *** e0ne has joined #openstack-placement | 15:53 | |
| dansmith | cdent: ah, I think I knew that once upon a time, thanks | 15:58 |
| cdent | dansmith: shore | 15:58 |
| openstackgerrit | Dan Smith proposed openstack/os-traits master: Add COMPUTE_IMAGE_TYPE_SUPPORT_* traits https://review.openstack.org/648147 | 15:59 |
| *** efried is now known as efried_rollin | 16:02 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Fill in the Writing Code section of contributing.rst https://review.openstack.org/648176 | 16:18 |
| *** helenafm has quit IRC | 16:28 | |
| *** cdent has quit IRC | 16:34 | |
| *** ttsiouts has quit IRC | 16:34 | |
| *** ttsiouts has joined #openstack-placement | 16:35 | |
| *** tssurya has quit IRC | 16:39 | |
| *** ttsiouts has quit IRC | 16:40 | |
| *** e0ne has quit IRC | 17:01 | |
| *** cdent has joined #openstack-placement | 17:17 | |
| cdent | edleafe, efried_rollin : am I missing something on tetsuro's forbidden agg spec? seems to be a logic flaw, or maybe just the spec not fully updated | 18:02 |
| *** e0ne has joined #openstack-placement | 18:04 | |
| edleafe | cdent: one is specific to the in: operator, which means any single match will satisfy. in:!agg3<anything else> makes <anything else> redundant | 18:05 |
| edleafe | Line 176 is saying that all conditions must be true | 18:06 |
| edleafe | Line 87 is simply about proper usage of in: | 18:06 |
| cdent | both refer to in:, 87 says we don't support ! in in, generally, 176 says we do | 18:07 |
| cdent | so there's either a typo, or a wrong | 18:08 |
| *** e0ne has quit IRC | 18:11 | |
| edleafe | oh, sorry, I was looking at 183, which doesn't use in: | 18:14 |
| edleafe | OK, yeha, 176 is not right | 18:14 |
| * edleafe heads out to his thrice-weekly torture session | 18:15 | |
| cdent | enjoy | 18:16 |
| *** cdent has quit IRC | 18:36 | |
| *** efried_rollin is now known as efried | 19:45 | |
| openstackgerrit | Merged openstack/placement master: s/rc_cache.ensure_rc_cache/rc_cache.ensure/ https://review.openstack.org/647946 | 20:09 |
| openstackgerrit | Ed Leafe proposed openstack/placement master: Spec: Support filtering by forbidden aggregates https://review.openstack.org/647704 | 20:21 |
| edleafe | cdent: efried: ^^ Updated to address cdent's concern. | 20:22 |
| efried | ack | 20:22 |
| efried | edleafe: one more update if you please. | 21:01 |
| efried | as modified in PS4, those two added cases are really the same thing. | 21:01 |
| edleafe | efried: I almost deleted the second example, as it was pretty much the same as the first. But I can add your version if you think it makes things clearer | 21:02 |
| efried | definitely | 21:03 |
| edleafe | But doesn't in: require matching all the aggs? | 21:03 |
| edleafe | in: implies 'and', and without in: is 'or' | 21:04 |
| edleafe | In your example, the first member of would match RPs in agg1 and agg2. The second would eliminate all that are in agg2, which means nothing would be returned | 21:06 |
| edleafe | So I'm not sure I see any improvement | 21:06 |
| edleafe | ok, I'm drunk | 21:07 |
| edleafe | Just re-read http://specs.openstack.org/openstack/api-sig/guidelines/pagination_filter_sort.html#filtering | 21:07 |
| edleafe | 'in:' is like python 'in' | 21:08 |
| efried | other way around | 21:10 |
| efried | yeah, in: is OR; without in: is AND | 21:11 |
| edleafe | Line 183 is wrong then. Should be &member_of=<agg1>&member_of=!<agg1> | 21:16 |
| openstackgerrit | Merged openstack/placement master: Remove fake resource class from fake_ensure_cache https://review.openstack.org/647948 | 21:17 |
| openstackgerrit | Ed Leafe proposed openstack/placement master: Spec: Support filtering by forbidden aggregates https://review.openstack.org/647704 | 21:30 |
| edleafe | efried: ^^ added those changes | 21:31 |
| openstackgerrit | Dan Smith proposed openstack/os-traits master: Add COMPUTE_IMAGE_TYPE_SUPPORT_* traits https://review.openstack.org/648147 | 21:39 |
| openstackgerrit | Ed Leafe proposed openstack/placement master: Spec: Support filtering by forbidden aggregates https://review.openstack.org/647704 | 21:48 |
| edleafe | efried: ^^ Doh! forgot to fix the typo | 21:48 |
| efried | was just leaving that comment. | 21:48 |
| efried | +2 | 21:48 |
| efried | We'll let Chris merge that and then I'll take the -2 off the bottom patch. | 21:49 |
| edleafe | kewl | 21:49 |
| *** gibi has quit IRC | 23:03 | |
| *** gibi has joined #openstack-placement | 23:22 | |
| *** takashin has joined #openstack-placement | 23:48 | |
| *** tetsuro has joined #openstack-placement | 23:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!