| *** tetsuro has joined #openstack-placement | 00:17 | |
| *** takashin has left #openstack-placement | 03:01 | |
| *** tetsuro has quit IRC | 06:08 | |
| *** e0ne has joined #openstack-placement | 06:24 | |
| *** e0ne has quit IRC | 06:27 | |
| openstackgerrit | Merged openstack/placement master: Move Inventory and InventoryList to own file https://review.openstack.org/642111 | 07:48 |
|---|---|---|
| openstackgerrit | Merged openstack/placement master: Remove InventoryList class https://review.openstack.org/642112 | 07:48 |
| *** tssurya has joined #openstack-placement | 08:15 | |
| *** mcgiggler has joined #openstack-placement | 08:24 | |
| *** e0ne has joined #openstack-placement | 08:35 | |
| openstackgerrit | Merged openstack/placement master: Remove ResourceProviderList class https://review.openstack.org/642138 | 10:02 |
| openstackgerrit | Merged openstack/placement master: Move set_traits tests back to test_resource_provider https://review.openstack.org/643313 | 10:02 |
| *** ttsiouts has joined #openstack-placement | 10:07 | |
| *** ttsiouts has quit IRC | 10:22 | |
| *** ttsiouts has joined #openstack-placement | 10:23 | |
| *** ttsiouts_ has joined #openstack-placement | 10:23 | |
| *** ttsiouts has quit IRC | 10:27 | |
| *** cdent has joined #openstack-placement | 10:38 | |
| *** mriedem has joined #openstack-placement | 10:53 | |
| *** e0ne has quit IRC | 11:03 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Move TestAllocation into test_allocation https://review.openstack.org/641524 | 11:15 |
| *** cdent has quit IRC | 11:31 | |
| *** cdent has joined #openstack-placement | 11:41 | |
| openstackgerrit | Stephen Finucane proposed openstack/placement master: Group API versions by release https://review.openstack.org/641422 | 11:49 |
| efried | stephenfin: early on todays todo list was to update the predecessor for ^ since the docstheme gizmo is in a release now. | 11:51 |
| efried | ima do that now... | 11:51 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Use ``code`` role in api-ref titles https://review.openstack.org/641404 | 12:01 |
| efried | stephenfin: ^ may want to rebase your'n | 12:01 |
| efried | ah bugger, needs another rebase for merge conflictness. | 12:02 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Use ``code`` role in api-ref titles https://review.openstack.org/641404 | 12:04 |
| efried | stephenfin: ^ | 12:05 |
| openstackgerrit | Stephen Finucane proposed openstack/placement master: Group API versions by release https://review.openstack.org/641422 | 12:05 |
| *** e0ne has joined #openstack-placement | 12:12 | |
| cdent | stephenfin: since you're a bit in the placement zone, these two tidy ups are worth having: https://review.openstack.org/#/c/641524/ https://review.openstack.org/#/c/640555/ | 12:40 |
| *** ttsiouts_ has quit IRC | 12:41 | |
| *** ttsiouts has joined #openstack-placement | 12:42 | |
| *** ttsiouts has quit IRC | 12:46 | |
| *** ttsiouts has joined #openstack-placement | 13:04 | |
| openstackgerrit | Eric Fried proposed openstack/placement master: Use ``code`` role in api-ref titles https://review.openstack.org/641404 | 13:29 |
| *** sean-k-mooney has joined #openstack-placement | 13:30 | |
| openstackgerrit | Eric Fried proposed openstack/placement master: Group API versions by release https://review.openstack.org/641422 | 13:32 |
| efried | that was a weird rebase ^ | 13:36 |
| efried | I had to pull it down and do it manually, but gerrit showed it as a trivial one, and kept not only the +2s but the +W. | 13:36 |
| stephenfin | cdent: Both done | 13:40 |
| cdent | thanks | 13:40 |
| openstackgerrit | Merged openstack/placement master: Remove NAME in placement/deploy.py https://review.openstack.org/642739 | 13:42 |
| openstackgerrit | Merged openstack/placement master: Extract user and endpoint creation install doc https://review.openstack.org/643363 | 13:42 |
| *** e0ne has quit IRC | 13:58 | |
| *** efried is now known as fried_rice | 14:00 | |
| *** e0ne has joined #openstack-placement | 14:06 | |
| openstackgerrit | Merged openstack/placement master: Docs: extract testing info to own sub-page https://review.openstack.org/639628 | 14:15 |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Optionally run a wsgi profiler when asked https://review.openstack.org/643269 | 14:16 |
| *** stephenfin is now known as finucannot | 14:19 | |
| *** sean-k-mooney has quit IRC | 14:20 | |
| *** ttsiouts has quit IRC | 14:21 | |
| *** ttsiouts has joined #openstack-placement | 14:22 | |
| fried_rice | cdent: given any thought to plumbing openstacksdk with placement thingies? | 14:22 |
| cdent | not in any significant way, mostly because in my own brain I don't like SDKs, I'd rather just use requests and hork with json blobs | 14:23 |
| cdent | but I know plenty of people feel otherwise, so would be in support of people making it go | 14:23 |
| fried_rice | cdent: for the effort under discussion in -nova, the first stage would be simply swapping out ksa-adapter.get with connection.adapter.get (or something along those lines) so not critical to make that happen. | 14:25 |
| *** ttsiouts has quit IRC | 14:25 | |
| *** ttsiouts has joined #openstack-placement | 14:25 | |
| fried_rice | but future, connection.placement.allocation_candidates(...) kind of thing etc. | 14:25 |
| * cdent nods | 14:25 | |
| cdent | does the connection.adapter.get step get us anything, or is it mostly a move in the right direction? As in, why not just carrying on using ksa? | 14:26 |
| fried_rice | Yeah, the direction being "use sdk instead of [python-*client, ksa, other]" | 14:30 |
| fried_rice | it's a low-surface-area way to prove the sdk plumbing works. | 14:30 |
| fried_rice | in terms of how the connection loads itself up from conf etc. | 14:30 |
| cdent | cool, yeah, seems good to me. | 14:33 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Note removal of OVO in contrib docs https://review.openstack.org/642545 | 14:36 |
| cdent | if someone would care to re-approve that ^ It required a manual rebase | 14:38 |
| cdent | fried_rice: i'm confused | 14:44 |
| cdent | you're asking me to touch a file that that patch never touched | 14:45 |
| cdent | oh wait I got ya | 14:47 |
| * cdent too many things at once | 14:47 | |
| fried_rice | cdent: That's correct. The patch originally made <edit1>, <edit2>. <edit2> originally lived in... | 14:47 |
| fried_rice | yeah | 14:47 |
| *** sean-k-mooney has joined #openstack-placement | 14:49 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Note removal of OVO in contrib docs https://review.openstack.org/642545 | 14:49 |
| cdent | better? | 14:49 |
| fried_rice | +2, qualified. | 14:51 |
| cdent | I bring this pain upon myself | 14:52 |
| cdent | but for the greater good or something | 14:52 |
| cdent | bbl | 14:58 |
| *** cdent has quit IRC | 14:58 | |
| *** ttsiouts has quit IRC | 15:14 | |
| openstackgerrit | Merged openstack/placement master: Add explicit short-circuit in get_all_by_consumer_id https://review.openstack.org/640555 | 15:18 |
| mriedem | fried_rice: edleafe: question on https://review.openstack.org/642545 | 15:33 |
| *** e0ne has quit IRC | 15:51 | |
| *** cdent has joined #openstack-placement | 15:54 | |
| *** ttsiouts has joined #openstack-placement | 15:55 | |
| *** cdent has quit IRC | 16:24 | |
| *** belmoreira has quit IRC | 16:45 | |
| *** mcgiggler has quit IRC | 17:01 | |
| *** e0ne has joined #openstack-placement | 17:10 | |
| *** e0ne has quit IRC | 17:11 | |
| *** ttsiouts has quit IRC | 17:23 | |
| *** ttsiouts has joined #openstack-placement | 17:24 | |
| *** tssurya has quit IRC | 17:24 | |
| *** ttsiouts has quit IRC | 17:28 | |
| openstackgerrit | Merged openstack/placement master: Move TestAllocation into test_allocation https://review.openstack.org/641524 | 17:47 |
| *** e0ne has joined #openstack-placement | 18:53 | |
| finucannot | fried_rice: Considered running the entirety of placement through black? | 19:26 |
| finucannot | (In relation to https://review.openstack.org/641723) | 19:26 |
| finucannot | (https://pypi.org/project/black/) | 19:27 |
| fried_rice | finucannot: you mean in reference to https://review.openstack.org/#/c/641497/ ? | 19:28 |
| finucannot | whoops, yeah | 19:28 |
| fried_rice | finucannot: In all seriousness, for pypowervm we used pep8 *and* sonar, and adding the latter was jeeeest a little bit too much. | 19:28 |
| fried_rice | there were good bits, but overall it was more pain than benefit. | 19:28 |
| fried_rice | so no, I have no intention of getting super militant. But I don't think it's unreasonable to do pretty much all of pep8. Especially since we're already there, and any divergence would be regression. | 19:29 |
| finucannot | Agreed. Adding tools like that to an existing project is rarely, if ever, work the effort | 19:30 |
| *** e0ne has quit IRC | 19:37 | |
| *** openstack has joined #openstack-placement | 19:45 | |
| *** ChanServ sets mode: +o openstack | 19:45 | |
| openstackgerrit | Merged openstack/placement master: Note removal of OVO in contrib docs https://review.openstack.org/642545 | 20:54 |
| openstackgerrit | Merged openstack/placement master: Remove unused table constants from allocation_candiate.py https://review.openstack.org/643314 | 20:54 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Get rid of backslash continuations in code https://review.openstack.org/641497 | 22:04 |
| fried_rice | edleafe: Before I go crazy, like this ^ | 22:04 |
| fried_rice | (by "go crazy" I mean "go through the effort of doing it all like that") | 22:04 |
| fried_rice | edleafe: specifically https://review.openstack.org/#/c/641497/1..2/placement/objects/resource_provider.py | 22:05 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Get rid of backslash continuations in code https://review.openstack.org/641497 | 22:13 |
| fried_rice | edleafe: wasn't that hard, went ahead and did 'em all ^ | 22:13 |
| edleafe | fried_rice: much, much better! | 22:23 |
| fried_rice | woot | 22:23 |
| edleafe | more readable, no? | 22:23 |
| fried_rice | Yes. Things in me grate at the extra assignments, but it is the least of the evils, so \o/ | 22:23 |
| fried_rice | it's also slightly surprising to me, until I examine and really think about it, that it didn't remove any magic doing it this way. | 22:24 |
| edleafe | Maybe add a comment so that some future person who has similar gratings doesn't undo that | 22:24 |
| fried_rice | mmph, everywhere? | 22:25 |
| edleafe | what do you mean by magic? | 22:25 |
| edleafe | At least once, not everywhere | 22:25 |
| fried_rice | Well, sqla is magic to me in general. Like the fact that comparator type things as method args get translated into sql query terms and shit. | 22:25 |
| fried_rice | so I thought that a().b().c() might actually do something more magical than x = a(); x = x.b(); x = x.c() | 22:26 |
| edleafe | yeah, I know what you mean. I personally hate any ORM, and sqla is no exception | 22:26 |
| edleafe | nah, that | 22:26 |
| edleafe | that's basic python | 22:26 |
| fried_rice | right, but like I said, I had to really look at and think about it to convince myself there *couldn't* be sqla magic happening. | 22:27 |
| fried_rice | edleafe: oh, while I have you here | 22:28 |
| fried_rice | I was about to start splitting up resource_provider.py | 22:28 |
| edleafe | at 5:30 on a friday? | 22:29 |
| fried_rice | I still want consumers to be able to say | 22:29 |
| fried_rice | from placement.objects import resource_provider as rp_obj | 22:29 |
| fried_rice | but I want to create sub-things like | 22:29 |
| fried_rice | from placement.objects.resource_provider import trait | 22:29 |
| fried_rice | ... aggregate | 22:29 |
| fried_rice | ... inventory | 22:29 |
| fried_rice | which are by name and function "trait stuff you do on a resource provider" etc. | 22:30 |
| edleafe | are those classes or functions? | 22:30 |
| fried_rice | modules | 22:30 |
| fried_rice | which would contain methods relating to "trait stuff you do on a resource provider" etc. | 22:31 |
| fried_rice | the obvious way is to make a dir called placement/objects/resource_provider/ and move the base resource provider stuff (like ResourceProvider and its listy methods) into __init__.py in that dir. Then the other modules could be trait.py, aggregate.py, inventory.py within that dir. | 22:31 |
| fried_rice | But some people don't like code in __init__.py | 22:31 |
| edleafe | then why wouldn't they be in placement.objects.trait etc? | 22:31 |
| fried_rice | because placement.objects.trait contains things that are about the Trait object | 22:31 |
| edleafe | Oh geez, that's hella confusing | 22:31 |
| fried_rice | placement.objects.resource_provider.trait is stuff you do on resource providers that relates to traits. | 22:31 |
| edleafe | It's much clearer to have rp_obj.do_trait_thing() | 22:32 |
| fried_rice | okay; resource_provider.py is 2KLOC - do you think that's small enough, or do you have another suggestion for splitting it, or...? | 22:32 |
| edleafe | The name should indicate that it's trait/inventory/agg related | 22:32 |
| fried_rice | I guess I could make rp_trait.py, rp_inventory.py, ... | 22:33 |
| edleafe | I don't have a problem with 2KLOC, as long as those lines all pertain to resource providers | 22:33 |
| edleafe | If I'm looking for a method that deals with RPs, I expect to find it in resource_provider.py | 22:34 |
| fried_rice | unless it deals with traits | 22:34 |
| fried_rice | then you might look for it in traits.py | 22:34 |
| edleafe | If it's acting on/modifying a trait, then I'd look in traits.py | 22:35 |
| edleafe | If it's acting on/modifying a RP, I wouldn't | 22:35 |
| fried_rice | mhm, because you're familiar with the structure | 22:35 |
| edleafe | not necesarily | 22:36 |
| fried_rice | I'm saying, if I'm looking for something that deals with a resource provider's inventories, and I saw a module called placement.objects.resource_provider.inventories, that's where I'd prolly go look. | 22:36 |
| edleafe | but if I saw placement.objects.resource_provider.py and placement.objects.inventories, I would look at the former first, since the thing I'm changing is an RP | 22:38 |
| fried_rice | unless you assumed that the raison d'être of inventories was to deal with things related to inventories | 22:38 |
| fried_rice | either way, you'd find it in a couple hops, so it's probably not the end of the world. | 22:39 |
| fried_rice | anyway, if there's not an obvious desire to split it up, or not substantial enough to be worth whatever semantic confusion it might or might not entail, I can let it drop. It was just something cdent and I were talking about getting done to further the making of smaller files. | 22:39 |
| edleafe | fried_rice: I'm about to call it a day. Let's discuss this at the meeting on monday | 22:39 |
| fried_rice | I talked him out of doing a major refactor of resource_provider.py (putting things in alpha order etc) because I was going to do this. | 22:39 |
| fried_rice | I think I'm still not in favor of the alpha reordering, because the current order has things divided logically along lines of inv/traits/aggs | 22:40 |
| fried_rice | rgr, I'll add an #open | 22:40 |
| fried_rice | o/ edleafe | 22:40 |
| edleafe | kewl. have a great weekend | 22:40 |
| *** mriedem has quit IRC | 23:13 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!