| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2022-08-23 | |||
| 13:26:13 | jm1 | wow, first X server crash for... a decade? omg.. | |
| 13:28:09 | gtema | lol | |
| 14:04:17 | opendevreview | Areg Grigoryan proposed openstack/openstacksdk master: Bugfix for paginating for complex paths https://review.opendev.org/c/openstack/openstacksdk/+/853656 | |
| 14:19:55 | opendevreview | Dmitry Tantsur proposed openstack/openstacksdk master: Allow passing explicit microversions to Resource methods https://review.opendev.org/c/openstack/openstacksdk/+/849394 | |
| 14:19:58 | dtantsur | rebase party begins ^^ | |
| 14:21:57 | opendevreview | Dmitry Tantsur proposed openstack/openstacksdk master: baremetal: rework node creation to be closer to the backend https://review.opendev.org/c/openstack/openstacksdk/+/849505 | |
| 14:22:19 | opendevreview | Dmitry Tantsur proposed openstack/openstacksdk master: Actually use openstacksdk from source in the Bifrost job https://review.opendev.org/c/openstack/openstacksdk/+/849564 | |
| 14:22:40 | opendevreview | Dmitry Tantsur proposed openstack/openstacksdk master: [WIP] Migrate register_machine to use the proxy layer https://review.opendev.org/c/openstack/openstacksdk/+/849402 | |
| 14:53:26 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack master: Refactored baremetal_node and baremetal_node_info modules https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854113 | |
| 15:33:22 | opendevreview | Will Szumski proposed openstack/ansible-collections-openstack master: Allow networks to be updated https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854170 | |
| 17:04:45 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack master: Refactored baremetal_node and baremetal_node_info modules https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854113 | |
| 17:45:30 | matbu | frickler: hey, according to our discussion yesterday I'm about to revert this patch in master: https://review.opendev.org/c/openstack/osc-lib/+/841307 | |
| 17:45:59 | matbu | frickler: and see if we need a proper fix in train only or not .. i dont think it's something really critical anyway so.. | |
| 19:46:49 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack master: Refactored baremetal_node and baremetal_node_info modules https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854113 | |
| 22:54:27 | opendevreview | Rafael Castillo proposed openstack/ansible-collections-openstack master: Updates volume for 2.0.0 https://review.opendev.org/c/openstack/ansible-collections-openstack/+/853749 | |
| #openstack-sdks - 2022-08-24 | |||
| 01:53:49 | opendevreview | Rafael Castillo proposed openstack/openstacksdk master: Use /volumes/detail endpoint in find_volume proxy method https://review.opendev.org/c/openstack/openstacksdk/+/854293 | |
| 01:54:14 | opendevreview | Rafael Castillo proposed openstack/ansible-collections-openstack master: Updates volume for 2.0.0 https://review.opendev.org/c/openstack/ansible-collections-openstack/+/853749 | |
| 05:28:32 | opendevreview | yangjianfeng proposed openstack/python-openstackclient master: Add router ndp proxy commands https://review.opendev.org/c/openstack/python-openstackclient/+/749038 | |
| 05:37:58 | opendevreview | Merged openstack/openstacksdk master: Use /volumes/detail endpoint in find_volume proxy method https://review.opendev.org/c/openstack/openstacksdk/+/854293 | |
| 06:22:18 | opendevreview | yangjianfeng proposed openstack/python-openstackclient master: Add router ndp proxy commands https://review.opendev.org/c/openstack/python-openstackclient/+/749038 | |
| 06:37:53 | opendevreview | yangjianfeng proposed openstack/python-openstackclient master: Add router ndp proxy commands https://review.opendev.org/c/openstack/python-openstackclient/+/749038 | |
| 09:25:17 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack master: Refactored baremetal_node and baremetal_node_info modules https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854113 | |
| 09:30:26 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack master: Refactored baremetal_node_info module https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854357 | |
| 09:37:00 | stephenfin | gtema: easy OSC review for a new nova microversion here https://review.opendev.org/c/openstack/python-openstackclient/+/831902 | |
| 09:37:09 | gtema | oki, looking | |
| 09:37:42 | stephenfin | as I paste that, I wonder if we need to bump the novaclient version in requirements.txt? | |
| 09:37:45 | stephenfin | we probably do | |
| 09:38:02 | stephenfin | (pity lower-constraints is no longer a thing for stuff like this) | |
| 10:04:13 | opendevreview | Stephen Finucane proposed openstack/python-openstackclient master: Remove invalid note from requirements files https://review.opendev.org/c/openstack/python-openstackclient/+/854364 | |
| 10:04:13 | opendevreview | Stephen Finucane proposed openstack/python-openstackclient master: Bump python-novaclient version https://review.opendev.org/c/openstack/python-openstackclient/+/854363 | |
| 11:41:26 | stephenfin | frickler: I left a lot of comments on https://review.opendev.org/c/openstack/python-openstackclient/+/749038/ Maybe you can help me understand some of the naming decisions? | |
| 11:52:14 | frickler | stephenfin: already at it ;) | |
| 11:52:23 | stephenfin | thanks! | |
| 12:04:44 | frickler | slaweq: ^^ maybe you have some opinion on this, too. I actually like stephenfin's idea of using "router ndp proxy" as command prefix instead of "network ndp proxy" | |
| 12:37:16 | opendevreview | Will Szumski proposed openstack/ansible-collections-openstack master: Allow networks to be updated https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854170 | |
| 12:40:01 | jm1 | gtema: are you sure that openstacksdk's self._baremetal_client.post() will accept sdk attribute names instead of openstack api attribute names? e.g. schema validation fails because i am passing 'id' https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_908/854113/12/check/bifrost-collections-src/90849ec/job-output.txt | |
| 12:54:50 | opendevreview | Rafael Castillo proposed openstack/ansible-collections-openstack master: Updates volume for 2.0.0 https://review.opendev.org/c/openstack/ansible-collections-openstack/+/853749 | |
| 13:02:49 | opendevreview | ribaudr proposed openstack/openstacksdk master: Microversion 2.91: Support specifying destination host to unshelve https://review.opendev.org/c/openstack/openstacksdk/+/853949 | |
| 13:04:20 | gtema | <jm1> "gtema: are you sure that..." <- nom xxx_client.post accepts server side structure | |
| 13:04:47 | gtema | you will only need this way if there is no corresponding proxy method already existing | |
| 13:05:15 | jm1 | gtema: i was thinking the same but "<gtema> <jm1> "gtema: or will _baremetal_client..." <- it does exactly that" ;) | |
| 13:05:43 | gtema | na ja, it depends on how you formulated your question and how I understood it ;-) | |
| 13:06:58 | jm1 | gtema: XD anyway. this means we need something like dtantsur's https://review.opendev.org/c/openstack/openstacksdk/+/849402/7 | |
| 13:08:27 | dtantsur | I want to allocate time next week to finish it properly and see if I need to address anything else | |
| 13:08:32 | gtema | maybe. If I would correctly understand what and where is wrong. Point of switching cloud layer to rely on proxy layer is to get rid of xx_client.<METHOD> calls. Ansible module "should" also rely on proxy layer itself (as long as possible and makes sense) | |
| 13:13:01 | jm1 | gtema: register_machine will do two things, first create a baremetal node and then create baremetal ports. if anything fails, operations will be rolled back. so register_machine consists of funtionality which i do not want to replicate in aoc. imho it belongs to openstacksdk. | |
| 13:25:08 | gtema | yupp, but for some reason you was asking me whether you can do baremetal_client.post and that confused me on the purpose. All good | |
| 13:27:16 | jm1 | gtema: ah ok sorry. will you fix it anyway? 😳 | |
| 13:27:45 | gtema | fix what? and that is exactly what is confusing: I do not really understand what should be fixed | |
| 13:38:26 | opendevreview | Rafael Castillo proposed openstack/ansible-collections-openstack master: Updates volume for 2.0.0 https://review.opendev.org/c/openstack/ansible-collections-openstack/+/853749 | |
| 14:42:22 | jm1 | gtema: register_machine assumes that kwargs consists of keys which openstack api accepts. For example, you have to pass instance_uuid. you cannot pass keys of openstacksdk's baremetal.*.Node class such as 'id'. | |
| 14:43:20 | jm1 | gtema: it would be great if register_machine would be changed in a way that you do not have to use openstack api keys but openstacksdk's baremetal.*.Node keys | |
| 14:44:08 | jm1 | gtema: all other methods are somehow passing the resource/proxy layer. register_machine is one of those relicts which expose openstack api directly to the user/developer | |
| 14:48:12 | jm1 | gtema: hence it would be great to change register_machine to offer a resource/proxy layer based api to the user instead of exposing openstack api | |
| 15:21:19 | gtema | jm1: copied your messages into my "notes" to not to loose them. Will try to dig deeper around Friday | |
| 15:26:53 | jm1 | gtema: ok great! thanks! | |
| 15:46:31 | opendevreview | Merged openstack/openstacksdk master: Allow Resource.to_dict to allow returning unknown values https://review.opendev.org/c/openstack/openstacksdk/+/843963 | |
| 16:02:36 | opendevreview | Will Szumski proposed openstack/ansible-collections-openstack master: Allow networks to be updated https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854170 | |
| 20:47:43 | opendevreview | Rafael Castillo proposed openstack/ansible-collections-openstack master: Updates volume for 2.0.0 https://review.opendev.org/c/openstack/ansible-collections-openstack/+/853749 | |
| 23:11:29 | opendevreview | Rafael Castillo proposed openstack/openstacksdk master: Support unknown attributes in resource.__getitem__ https://review.opendev.org/c/openstack/openstacksdk/+/854461 | |
| 23:16:24 | opendevreview | Rafael Castillo proposed openstack/openstacksdk master: Support unknown attributes in resource.__getitem__ https://review.opendev.org/c/openstack/openstacksdk/+/854461 | |
| #openstack-sdks - 2022-08-25 | |||
| 00:12:19 | opendevreview | Takashi Kajinami proposed openstack/python-openstackclient master: Fix missing closing bracket https://review.opendev.org/c/openstack/python-openstackclient/+/854462 | |
| 03:39:32 | opendevreview | Ke Niu proposed openstack/shade master: remove unicode prefix from code https://review.opendev.org/c/openstack/shade/+/854510 | |
| 03:40:32 | opendevreview | Ke Niu proposed openstack/shade master: remove unicode prefix from code https://review.opendev.org/c/openstack/shade/+/854510 | |
| 06:42:37 | opendevreview | yangjianfeng proposed openstack/python-openstackclient master: Add router ndp proxy commands https://review.opendev.org/c/openstack/python-openstackclient/+/749038 | |
| 06:51:28 | opendevreview | yangjianfeng proposed openstack/python-openstackclient master: Add router ndp proxy commands https://review.opendev.org/c/openstack/python-openstackclient/+/749038 | |
| 07:38:39 | opendevreview | Merged openstack/openstacksdk master: Support unknown attributes in resource.__getitem__ https://review.opendev.org/c/openstack/openstacksdk/+/854461 | |
| 07:55:25 | jm1 | gtema: actually i think we should implement that jmespath_filters handling inside openstacksdk https://review.opendev.org/c/openstack/openstacksdk/+/853052/comments/11073ef3_681b3030 | |
| 07:56:18 | gtema | what exactly do you mean? with my patch jmespath is supported, but you need to pass it. So I do not deal with creating jmes out of anything else | |
| 07:58:02 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack master: Refactored baremetal_node_action module https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854357 | |
| 08:04:50 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack master: Refactored baremetal_node_action module https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854357 | |
| 08:14:26 | opendevreview | Areg Grigoryan proposed openstack/python-openstackclient master: Added "openstack image namespace list" command https://review.opendev.org/c/openstack/python-openstackclient/+/854557 | |
| 08:49:14 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack stable/1.0.0: Release 1.9.0 version https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854564 | |
| 08:51:47 | jm1 | gtema: cloud layer functions such as search_projects offer a filters parameter which accept(ed in the past) a jmespath expression. imho openstacksdk should continue to support that | |
| 09:05:53 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack master: Refactored baremetal_node_action module https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854357 | |
| 09:06:24 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack master: Refactored baremetal_port and baremetal_port_info modules https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854569 | |
| 09:19:28 | opendevreview | Jakob Meng proposed openstack/ansible-collections-openstack master: Refactored baremetal_port and baremetal_port_info modules https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854569 | |
| 10:06:32 | gtema | <jm1> "gtema: cloud layer functions..." <- ok, for cases from cloud layer this is definitely going to live further. It just that implementation will be moved to the list function. Once we agree on the change in question and merge it I will go and update cloud layer functions to unify the logic | |
| 10:25:31 | jm1 | gtema: ack | |
| 11:09:45 | opendevreview | yangjianfeng proposed openstack/python-openstackclient master: Add router ndp proxy commands https://review.opendev.org/c/openstack/python-openstackclient/+/749038 | |
| 11:14:16 | opendevreview | yangjianfeng proposed openstack/python-openstackclient master: Add router ndp proxy commands https://review.opendev.org/c/openstack/python-openstackclient/+/749038 | |
| 11:54:00 | opendevreview | Areg Grigoryan proposed openstack/openstacksdk master: added api requirements for new "openstack image namespace list command" https://review.opendev.org/c/openstack/openstacksdk/+/854612 | |
| 13:16:33 | opendevreview | Merged openstack/ansible-collections-openstack stable/1.0.0: Release 1.9.0 version https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854564 | |
| 14:54:47 | opendevreview | Shnaidman Sagi (Sergey) proposed openstack/ansible-collections-openstack stable/1.0.0: Fix release job https://review.opendev.org/c/openstack/ansible-collections-openstack/+/854634 | |
| 16:09:25 | opendevreview | Takashi Kajinami proposed openstack/python-openstackclient master: Fix missing closing brackets in metavar https://review.opendev.org/c/openstack/python-openstackclient/+/854462 | |
| 16:42:12 | frickler | slaweq: gtema: ralonsoh: via a merge conflict I found https://review.opendev.org/c/openstack/python-openstackclient/+/424733 and it does have a certain point. there are a lot of field that show "None" as value instead of an empty string, only some like description are really empty. my fix would likely go in the other direction though, like make an unset ipv4_address_scope or | |
| 16:42:18 | frickler | provider:segmentation_id field show an empty value instead of "None". what do you think? | |
| 17:18:34 | ralonsoh | frickler, for flat networks doesn't matter | |
| 17:18:49 | ralonsoh | maybe is better an empty string, rather than None | |
| 17:19:27 | ralonsoh | what other parameters could have this None parameter? | |
| 17:19:46 | ralonsoh | hold on, no, it is better to show "None" | |
| 17:19:58 | ralonsoh | empty string is a valid value, None is no value at all | |
| 17:20:08 | ralonsoh | and I've seen a problem related to the port binding because of this | |
| 17:20:25 | ralonsoh | converting a non-value to an empty string could hide this | |
| 17:46:56 | gtema | frickler: unset doesn't work for sure, since user has attribute access. None vs empty value is so terrible to differentiate once server doesn't return the field at all (/serves/details vs /servers). And even more troublesome to differentiate empty value (as None) from real empty value | |
| 17:47:33 | gtema | We have none in the interface and we need to live with it | |
| 20:23:36 | frickler | so you're telling me that the above patch is valid and I should rebase it and also apply it for the dns_domain field, o.k. | |
| #openstack-sdks - 2022-08-26 | |||
| 00:15:57 | opendevreview | yangjianfeng proposed openstack/python-openstackclient master: Add router ndp proxy commands https://review.opendev.org/c/openstack/python-openstackclient/+/749038 | |
| 06:47:55 | frickler | stephenfin: slaweq: gtema: ^^ I think this should be fine now | |
| 10:48:22 | slaweq | frickler +W | |