| *** bknudson has joined #openstack-sdks | 00:48 | |
| *** etoews has quit IRC | 00:57 | |
| *** jamielennox is now known as jamielennox|away | 01:16 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 01:17 | |
| *** pm90_ has quit IRC | 01:20 | |
| *** jamielennox|away is now known as jamielennox | 01:31 | |
| *** jamielennox is now known as jamielennox|away | 01:41 | |
| *** pm90_ has joined #openstack-sdks | 01:43 | |
| openstackgerrit | Brian Curtin proposed stackforge/python-openstacksdk: Allow Resource attributes to be updated https://review.openstack.org/156448 | 02:07 |
|---|---|---|
| *** jamielennox|away is now known as jamielennox | 02:09 | |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Revert "Skip functional tests until they are fixed" https://review.openstack.org/156449 | 02:11 |
| *** terrylhowe has joined #openstack-sdks | 02:18 | |
| openstackgerrit | Brian Curtin proposed stackforge/python-openstacksdk: Use case insensitive dict for Resource attributes https://review.openstack.org/156135 | 02:33 |
| *** bknudson has quit IRC | 02:37 | |
| openstackgerrit | Brian Curtin proposed stackforge/python-openstacksdk: Allow Resource attributes to be updated https://review.openstack.org/156448 | 03:09 |
| *** pm90_ has quit IRC | 03:26 | |
| *** pm90_ has joined #openstack-sdks | 03:31 | |
| *** pm90_ has quit IRC | 03:50 | |
| *** pm90_ has joined #openstack-sdks | 05:21 | |
| *** pm90_ has quit IRC | 05:24 | |
| *** etoews has joined #openstack-sdks | 05:25 | |
| *** etoews has quit IRC | 05:29 | |
| *** terrylhowe has quit IRC | 05:33 | |
| *** pm90_ has joined #openstack-sdks | 05:38 | |
| *** pm90_ has quit IRC | 05:58 | |
| openstackgerrit | Brian Curtin proposed stackforge/python-openstacksdk: Mutate Resource via mapping instead of dict https://review.openstack.org/156485 | 06:46 |
| *** etoews has joined #openstack-sdks | 06:57 | |
| openstackgerrit | Brian Curtin proposed stackforge/python-openstacksdk: Adjust Container override methods https://review.openstack.org/156487 | 06:58 |
| *** etoews has quit IRC | 07:02 | |
| openstackgerrit | Colleen Murphy proposed openstack/python-openstackclient: Add project-id parameter to network create https://review.openstack.org/156492 | 07:24 |
| *** stevemar has quit IRC | 07:38 | |
| *** fifieldt has joined #openstack-sdks | 08:09 | |
| *** chlong has quit IRC | 08:11 | |
| *** etoews has joined #openstack-sdks | 08:20 | |
| *** etoews has quit IRC | 08:24 | |
| *** etoews has joined #openstack-sdks | 08:39 | |
| *** etoews has quit IRC | 08:43 | |
| *** fifieldt has quit IRC | 08:56 | |
| *** jamie_h_ has joined #openstack-sdks | 08:58 | |
| *** jamie_h_ is now known as jamie_h | 09:00 | |
| *** jamie_h has left #openstack-sdks | 09:02 | |
| *** etoews has joined #openstack-sdks | 09:10 | |
| *** etoews has quit IRC | 09:14 | |
| *** etoews has joined #openstack-sdks | 11:06 | |
| *** etoews has quit IRC | 11:11 | |
| *** etoews has joined #openstack-sdks | 11:25 | |
| *** etoews has quit IRC | 11:29 | |
| *** etoews has joined #openstack-sdks | 12:25 | |
| *** etoews has quit IRC | 12:30 | |
| *** etoews has joined #openstack-sdks | 13:19 | |
| *** HenryG has quit IRC | 13:49 | |
| *** HenryG has joined #openstack-sdks | 13:52 | |
| *** pm90_ has joined #openstack-sdks | 14:00 | |
| *** ayoung has joined #openstack-sdks | 14:12 | |
| *** terrylhowe has joined #openstack-sdks | 14:18 | |
| *** bknudson has joined #openstack-sdks | 14:20 | |
| *** pm90_ has quit IRC | 14:29 | |
| *** pm90_ has joined #openstack-sdks | 14:29 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:54 | |
| openstackgerrit | Brian Curtin proposed stackforge/python-openstacksdk: Adjust Container override methods https://review.openstack.org/156487 | 15:26 |
| *** stevemar has joined #openstack-sdks | 15:28 | |
| openstackgerrit | Brian Curtin proposed stackforge/python-openstacksdk: Use case insensitive dict for Resource attributes https://review.openstack.org/156135 | 15:36 |
| *** pm90_ has quit IRC | 16:02 | |
| *** etoews has quit IRC | 16:03 | |
| *** etoews has joined #openstack-sdks | 16:04 | |
| briancurtin | terrylhowe: what if we do Resource.list(..., paginated=True) and then just bounce out of the more_data list after one iteration when it's false? | 16:31 |
| briancurtin | this way at the lower level, which i think we agree that you have to know some of what you're doing, there's not really anything added there. then it would still be equally hidden from Proxy.list_blah calls | 16:33 |
| briancurtin | **more_data loop | 16:33 |
| *** bknudson has quit IRC | 16:50 | |
| openstackgerrit | Brian Curtin proposed stackforge/python-openstacksdk: Support listing non-paginated response bodies https://review.openstack.org/156664 | 17:04 |
| terrylhowe | briancurtin: I like the search code, it is cool, but I think a lot of people will want code that does minimal I/O | 17:05 |
| terrylhowe | briancurtin: It seems like part of your discomfort is page returns a dict and maybe we should change that so it returns a list of resource. It is kind of inconsistent that it returns dicts | 17:06 |
| briancurtin | terrylhowe: how about https://review.openstack.org/#/c/156664/ instead of my Resource.page change? that allows page to stay as-is and list is able to work for both cases | 17:07 |
| *** etoews has quit IRC | 17:07 | |
| briancurtin | terrylhowe: i like the idea of find, but it returns duplicate when you search by name for most services since they don't use field but instead use fields (or other way around, cant remember off the top of my head). when we know a resource uses those particular arguments to do fine grained searches we should add overrides | 17:08 |
| *** etoews has joined #openstack-sdks | 17:09 | |
| briancurtin | something i did the other day had one resource find-able within a service and another resource in that same service wasn't findable. i agree the one i proposed is IO heavy, but it functions in the lack of a more specialized offering for that service | 17:10 |
| briancurtin | (i cant remember which service this was, might have been when i was finding images and flavors, i think one worked but other didnt) | 17:11 |
| terrylhowe | https://review.openstack.org/#/c/156664/ puts the problem to the user which might be best | 17:11 |
| terrylhowe | that way we don’t have the allow_page thing to manage and allow people to override | 17:12 |
| briancurtin | i think it's probably the best way given the other technical approaches proved too hard (extra requests, etc). if you're working at the resource layer, you already have to know a little more about the API than you do at the proxy layer, which is where this would be covered up | 17:13 |
| terrylhowe | the search code is slick, but I’d also like to see a high perf method as well | 17:16 |
| terrylhowe | particularlly in the object store area, the seach method could be painful. I know I’ve seen an absurd number of objects in a container and it has caused a lot of issues, but should be supported | 17:17 |
| briancurtin | as a base method the current implementation doesn't actually work though - it chooses arguments that not all resources support. i'd love to see specializations but i dont think we can offer a base method that is going to raise duplicate a lot of the time | 17:18 |
| *** pm90_ has joined #openstack-sdks | 17:25 | |
| *** etoews has quit IRC | 17:27 | |
| *** etoews has joined #openstack-sdks | 17:29 | |
| *** pm90_ has quit IRC | 17:33 | |
| *** pm90_ has joined #openstack-sdks | 17:33 | |
| briancurtin | terrylhowe: what we should probably do is just create a resource with the given value as ID and get_by_id (i think you said this somewhere but i cant find it) | 17:34 |
| briancurtin | one call, one response, either it's found or its not. my list impl is wildly inefficient, but we should just try to directly get the value sent in | 17:34 |
| briancurtin | we'd have to make Resource.id not be read-only. i dont know if there are other implications to this, i think we just did it for some safety | 17:35 |
| *** etoews_ has joined #openstack-sdks | 17:36 | |
| *** etoews has quit IRC | 17:36 | |
| briancurtin | if we don't get it back when it's sent as the id_attribute, make a second call with name_attribute set. that's two calls but potentially less data. i dont know the speed difference, but for something like find_images where there are 30 images, two calls vs one list is probably ok | 17:37 |
| briancurtin | i'm going to hack that up and see how it works out | 17:44 |
| *** bknudson has joined #openstack-sdks | 18:01 | |
| *** etoews_ has quit IRC | 18:02 | |
| *** etoews has joined #openstack-sdks | 18:05 | |
| *** bknudson has quit IRC | 18:14 | |
| *** etoews has quit IRC | 18:15 | |
| *** bknudson has joined #openstack-sdks | 18:16 | |
| *** etoews has joined #openstack-sdks | 18:16 | |
| briancurtin | terrylhowe: the first half of this is what we need, second part is probably going to require the full list if we want it to work generally https://gist.github.com/briancurtin/5dd902f504b7bfd98205 - the marker part would be perfect if everything supported that but unfortunately it's not viable | 18:22 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 18:25 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 18:31 | |
| briancurtin | SDK meeting in 25 minutes in #openstack-meeting-3 - https://wiki.openstack.org/wiki/Meetings/PythonOpenStackSDK#Agenda_for_2015-02-16_1900_UTC | 18:35 |
| katco | is there any kind of automated test-suite for testing new sdks? | 18:55 |
| openstackgerrit | Brian Curtin proposed stackforge/python-openstacksdk: Adjust Resource.find to try direct or filtered https://review.openstack.org/156707 | 18:57 |
| briancurtin | katco: as in something that mimics openstack APIs? | 18:59 |
| katco | briancurtin: yeah, something i could throw http requests against | 18:59 |
| briancurtin | terrylhowe: i think https://review.openstack.org/#/c/156707/1/openstack/resource.py might serve what you're looking for. the filters that were hardcoded before become parameters, so you can send the right ones for whichever resource you're searching for. i tried it with finding flavors and filtering on minRam size worked well | 19:00 |
| briancurtin | katco: you might want to look at https://github.com/rackerlabs/mimic - i know the name, don't know what all it does or if it does what you're looking for | 19:01 |
| katco | briancurtin: awesome, tyvm | 19:02 |
| *** openstackgerrit has quit IRC | 19:20 | |
| *** openstackgerrit has joined #openstack-sdks | 19:20 | |
| briancurtin | terrylhowe: hmm...maybe i should make Resource itself be the CaseInsensitiveDict? | 20:31 |
| terrylhowe | maybe | 20:31 |
| terrylhowe | if this is headers only, maybe special handling for that | 20:31 |
| briancurtin | i'll check it out see how that works | 20:31 |
| briancurtin | i thought about teh headers only thing for a bit, and while it's the most common problem, im not certain that it's specific to it. doing separate body/header handling then just adds complexity that we end up unifying anyway, and still need something like a CID | 20:32 |
| briancurtin | if it makes the Resource magic methods too messy perhaps it should stay on its own, but ill check it out | 20:32 |
| terrylhowe | I’m just concerned some service is going to be case sensive and cause headaches | 20:33 |
| terrylhowe | it would be something that would be hard to detect | 20:33 |
| briancurtin | terrylhowe: what we send outbound will have whatever case configure the prop as, so we can be safe there. hard to detect though | 20:33 |
| terrylhowe | okay, so it would only be a potential problem for attributes that aren’t resources maybe | 20:34 |
| briancurtin | i think we can only worry about what we have resources for | 20:35 |
| briancurtin | since we're not able to lookup things that we don't have props for, and we wouldn't be sending attributes we don't have props for in posts/puts | 20:36 |
| openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: add metric proxy and service https://review.openstack.org/155811 | 20:56 |
| openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: add metric proxy and service https://review.openstack.org/155811 | 20:58 |
| openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: add metric proxy and service https://review.openstack.org/155811 | 21:25 |
| *** bknudson has quit IRC | 22:00 | |
| openstackgerrit | Terry Howe proposed stackforge/python-openstacksdk: Revert to KSC auth plugins https://review.openstack.org/156064 | 22:13 |
| openstackgerrit | Colleen Murphy proposed openstack/python-openstackclient: Add project parameter to network create https://review.openstack.org/156492 | 22:20 |
| *** utahcon has quit IRC | 22:36 | |
| *** bknudson has joined #openstack-sdks | 22:48 | |
| *** chlong has joined #openstack-sdks | 23:08 | |
| *** etoews has quit IRC | 23:29 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!