| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2020-03-12 | |||
| 16:34:01 | mordred | kk | |
| 16:34:12 | dtantsur | in my brain SDK and CLI at least have stakes in the API SIG work | |
| 16:34:21 | dtantsur | email++ | |
| 16:34:38 | dtantsur | now that we've ensured that mordred is not going to run away screaming | |
| 16:34:40 | elmiko | agreed dtantsur, and we have been welcome here for awhile so it seems like a natural merge to me | |
| 16:34:44 | elmiko | LOL | |
| 16:34:46 | dtantsur | yup | |
| 16:35:03 | mordred | basically - this channel is all one big (or maybe not _big_) happy family | |
| 16:35:09 | elmiko | \o/ | |
| 16:35:28 | mordred | elmiko: just be warned, this is going to mean you're going to become core on sdk and osc as a result | |
| 16:35:30 | dtantsur | smaller than it should be, but very happy and welcoming | |
| 16:35:48 | dtantsur | here goes The Evin Plan (tm) | |
| 16:36:00 | elmiko | mordred: ack, thanks for the warn^H^H^H^H heads up | |
| 16:37:10 | amotoki | slaweq: mordred: as slaweq mentioned neutron does not support micro-versioning. I explored a way to microversion in neutron, but per PTG discussion perhaps in Denver we agreed that API extension versioning would be simple and reasonable as microversion in neutron may potentially lead to regression in current neutron behaviors. | |
| 16:37:50 | amotoki | I hope I can restore the work on API extension versioning but the horizon PTL and less contributors in horizon steal my time..... :( | |
| 16:38:08 | mordred | amotoki: I understand these issues | |
| 16:58:32 | elmiko | dtantsur, mordred, just sent an email. hopefully it helps get the ball rolling =) | |
| 16:58:55 | dtantsur | thanks! | |
| 17:04:52 | noonedeadpunk | Am I right, that https://docs.openstack.org/api-ref/identity/v3/?expanded=list-projects-detail,list-projects-for-user-detail#list-projects-for-user is just not present in openstackclient? | |
| 17:06:52 | noonedeadpunk | ah, ok, it's used with flag --user | |
| 17:37:50 | noonedeadpunk | them most sad part with glance memer is that glanceclient itself works:( http://paste.openstack.org/show/790619/ | |
| 17:39:50 | noonedeadpunk | and things are so because of the way how find_resource is designed.... | |
| 18:36:00 | openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/python-openstackclient master: Fix excpected exception for find_identity_resource https://review.opendev.org/712764 | |
| 20:14:08 | openstackgerrit | Merged openstack/openstacksdk stable/train: Fix microversion negotiation in some bare metal node call https://review.opendev.org/712536 | |
| 20:15:59 | mordred | noonedeadpunk: I think we can fix that | |
| 21:06:04 | umbSublime | Hey I'm trying to see if maybe the osc-lib or sdk has something equivalent to novaclient.utils.flatten_dict or keystone.common.utils.flatten_dict I can't find a method with name flatten_dict, but maybe the name was just refactored | |
| 21:15:18 | openstackgerrit | Merged openstack/openstacksdk master: Rationalize examples and functional extra config loading https://review.opendev.org/712538 | |
| 21:29:31 | umbSublime | I know it's only like 10L of code, but it's quite useful, but maybe it would have it's place in cliff instead of osc/sdk | |
| 22:28:38 | mordred | umbSublime: sorry - don't know of anything like it | |
| 22:28:44 | mordred | noonedeadpunk: oh - wow. that's ... | |
| 22:28:46 | mordred | what? | |
| 22:29:08 | mordred | noonedeadpunk: your fix is showing a slightly deeper issue | |
| 22:42:27 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Don't look up project by id if given id https://review.opendev.org/712810 | |
| 22:42:54 | mordred | noonedeadpunk, mnaser: ^^ try that | |
| 22:43:26 | mordred | the other thing is a rabbit hole - and one that we need to go down - but it'll take me a couple of seconds to even write down what is broken there and what needs to be fixed | |
| 22:43:46 | mordred | the good news is that I dont think fixing it is going to be super HARD - it just needs to be done in a few places all at once | |
| 22:48:10 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Optimize adding project to image slightly more https://review.opendev.org/712813 | |
| #openstack-sdks - 2020-03-13 | |||
| 00:20:38 | openstackgerrit | Merged openstack/openstacksdk stable/train: Consistent normalization of Machine objects in the cloud layer https://review.opendev.org/712295 | |
| 10:08:36 | BakaKuna | Hi, I was trying to change the admin state of a dhcp agent yesterday. Mordred gave me a pointer, but I think I just don't understand the flow yet. Should I be able to change the admin state of a DHCP agent (or any other attribut)? And if so, could anyone tell me how that should be done? | |
| 10:10:49 | BakaKuna | I have looked through the documentation, but I can't seem to find a place where an attribute is changed without using a dedicated method. | |
| 10:29:13 | BakaKuna | Found it, network.update_agent() did the trick. This is amazing! How can I only find out about this whole project now? I have been immersed in OpenStack literature the last 6 months since I started a new job :D | |
| 10:39:59 | openstackgerrit | Alfredo Moralejo proposed openstack/osc-lib master: Replace assertItemsEqual with assertCountEqual https://review.opendev.org/712904 | |
| 10:42:33 | noonedeadpunk | mordred: yeah, I tried to do kinda the same as https://review.opendev.org/#/c/712810/1 but also had local tests failing... | |
| 10:43:10 | noonedeadpunk | so ended up in just fixing exception.... | |
| 10:49:11 | noonedeadpunk | also there are the way more usages _find_identity_resource.... | |
| 10:49:46 | noonedeadpunk | actually I'm really wondering why osc thtows the same error for so different exceptions... | |
| 11:50:42 | openstackgerrit | Mark Chappell proposed openstack/ansible-collections-openstack master: os_nova_host_aggregate: Fixup sanity test failures https://review.opendev.org/712917 | |
| 13:53:35 | amoralej | hi, wrt https://review.opendev.org/#/c/712904/, is osc-lib gate broken? | |
| 13:54:09 | amoralej | ERROR: Package 'neutron' requires a different Python: 2.7.17 not in '>=3.6' | |
| 14:06:22 | mordred | noonedeadpunk: yeah - that's the larger mess I was talking about - the same exception multiple causes thing | |
| 14:06:57 | mordred | BakaKuna: \o/ | |
| 14:09:54 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Don't look up project by id if given id https://review.opendev.org/712810 | |
| 14:10:01 | mordred | noonedeadpunk: try that one | |
| 14:11:18 | mordred | amoralej|lunch: might be - let me look | |
| 14:14:05 | openstackgerrit | Monty Taylor proposed openstack/osc-lib master: Stop using lib-forward-testing https://review.opendev.org/712961 | |
| 14:14:15 | mordred | amoralej|lunch: I think that should do it | |
| 14:16:40 | noonedeadpunk | mordred: smae thing for me | |
| 14:16:44 | noonedeadpunk | *same | |
| 14:18:35 | mordred | noonedeadpunk: it's still not working? | |
| 14:20:12 | noonedeadpunk | nope. I don't have even nothing common to `_is_uuid_like` in dir(openstack.connect()) | |
| 14:20:40 | noonedeadpunk | openstacksdk 0.43.0 | |
| 14:21:42 | mordred | yeah - sorry - bad patch - one sec | |
| 14:29:17 | openstackgerrit | Monty Taylor proposed openstack/python-openstackclient master: Don't look up project by id if given id https://review.opendev.org/712810 | |
| 14:29:29 | mordred | noonedeadpunk: ok. one more try - hopefullyt his time with more coffee | |
| 14:31:57 | noonedeadpunk | yeah, if only I've found _is_uuid_like method... | |
| 14:33:16 | mordred | noonedeadpunk: I think we need to apply that pattern to many of the places in osc where identity resources are given | |
| 14:33:51 | mordred | for _most_ things treating input as "name_or_id" works well - but it's actually quite common that users don't have access rights to look up identity resources (like project) by name unless they are an admin | |
| 14:33:56 | noonedeadpunk | mordred: btw, I'd say that search works as well, it just has completely different URL base (so actually client_manager, which is probably do not exist) | |
| 14:34:04 | mordred | so the pattern _completely_ falls apart for things like project | |
| 14:34:36 | noonedeadpunk | they actually can, with https://keystone:5000/v3/users/USER_UUID/projects?name=vexxhost | |
| 14:34:50 | noonedeadpunk | the same should work for groups, etc | |
| 14:34:57 | mordred | awesome. good to know - we'll fix that too then | |
| 14:35:15 | mordred | noonedeadpunk: I think what I'll do is a pass to replace keystoneclient with sdk (similar to the glanceclient patch) | |
| 14:35:47 | noonedeadpunk | yeah, sdk usage is the way cleaner I think | |
| 14:35:48 | mordred | and as part of that we'll add in _is_uuid_like calls and then hopefully fix the project/groups/etc calls | |
| 14:35:55 | mordred | if they aren't already fixed | |
| 14:36:26 | noonedeadpunk | I think it's more workaround rather than fix | |
| 14:36:36 | mordred | is_uuid_like ? | |
| 14:36:53 | mordred | yeah - it's a workaround - although it does also save a roundtrip | |
| 14:37:07 | noonedeadpunk | as api allows regular users to search in projects they have access to by it's name | |
| 14:37:16 | mordred | ++ | |
| 14:38:03 | mordred | yeah. I think that's the right call to use here | |
| 14:54:05 | mordred | Merged openstack/governance master: Merge SDK and OSC teams https://review.opendev.org/711265 | |
| 14:54:10 | mordred | FYI ^^ | |
| 14:55:30 | openstackgerrit | Alfredo Moralejo proposed openstack/osc-lib master: Replace assertItemsEqual with assertCountEqual https://review.opendev.org/712904 | |
| 14:55:44 | amoralej | thanks mordred | |
| 15:02:12 | smcginnis | https://gph.is/2ax5JZu | |
| 15:15:32 | openstackgerrit | Monty Taylor proposed openstack/ansible-collections-openstack master: Remove MANIFEST.json before building https://review.opendev.org/712970 | |
| 15:33:32 | openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/python-openstackclient master: Optimize adding project to image slightly more https://review.opendev.org/712813 | |
| 15:37:07 | openstackgerrit | Mark Chappell proposed openstack/ansible-collections-openstack master: os_nova_host_aggregate: Add support for not 'purging' missing hosts https://review.opendev.org/712979 | |
| 15:49:58 | openstackgerrit | Mark Chappell proposed openstack/ansible-collections-openstack master: os_nova_host_aggregate: Fixup sanity test failures https://review.opendev.org/712917 | |
| 15:49:59 | openstackgerrit | Mark Chappell proposed openstack/ansible-collections-openstack master: os_nova_host_aggregate: Add support for not 'purging' missing hosts https://review.opendev.org/712979 | |
| 16:20:00 | openstackgerrit | Merged openstack/ansible-collections-openstack master: Add option for modules_default groups https://review.opendev.org/712057 | |
| 16:54:25 | openstackgerrit | Sean McGinnis proposed openstack/os-client-config master: Use unittest.mock instead of third party mock https://review.opendev.org/713029 | |
| 17:14:34 | openstackgerrit | Merged openstack/ansible-collections-openstack master: Remove MANIFEST.json before building https://review.opendev.org/712970 | |
| 17:37:21 | openstackgerrit | Merged openstack/ansible-collections-openstack master: os_nova_host_aggregate: Fixup sanity test failures https://review.opendev.org/712917 | |
| 17:43:07 | openstackgerrit | Merged openstack/ansible-collections-openstack master: os_nova_host_aggregate: Add support for not 'purging' missing hosts https://review.opendev.org/712979 | |
| 18:46:13 | openstackgerrit | Merged openstack/openstacksdk master: baremetal: fail-less mode for wait_for_nodes_provision_state https://review.opendev.org/712637 | |
| 18:51:29 | openstackgerrit | Sean McGinnis proposed openstack/os-client-config master: Uncap jsonschema https://review.opendev.org/713049 | |
| 18:51:29 | openstackgerrit | Sean McGinnis proposed openstack/os-client-config master: Use unittest.mock instead of third party mock https://review.opendev.org/713029 | |