| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-06-14 | |||
| 10:26:12 | openstackgerrit | Merged openstack/nova master: remove get_nw_info_for_instance from compute.utils https://review.openstack.org/451387 | |
| 10:40:58 | cdent | bauzas, alex_xu : some relatively small placement-related fixes: https://review.openstack.org/#/c/468928/ https://review.openstack.org/#/c/469047/ https://review.openstack.org/#/c/469048/ https://review.openstack.org/#/c/471067/ please and thank you | |
| 10:48:48 | openstackgerrit | Merged openstack/nova master: Use plain routes list for os-server-groups endpoint instead of stevedore https://review.openstack.org/462169 | |
| 10:52:08 | alex_xu | cdent: added to my list, will try to reach them | |
| 10:52:08 | openstackgerrit | Sean Dague proposed openstack/python-novaclient master: Updated from global requirements https://review.openstack.org/472919 | |
| 10:52:41 | cdent | thanks alex_xu | |
| 10:52:49 | alex_xu | cdent: np | |
| 11:14:52 | openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova master: live-migration test hook for serial console https://review.openstack.org/347471 | |
| 11:32:17 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Make all timestamps formats equal https://review.openstack.org/421981 | |
| 11:49:12 | vladiskuz | Hey, guys! When I tried use pagination in nova I've faced with problem. When I use '--marker' nova doesn't return instances in error state. Is it a bug or not? | |
| 11:49:13 | vladiskuz | http://paste.openstack.org/show/612533/ | |
| 11:59:33 | sdague | lyarwood: is this still in progress - https://bugs.launchpad.net/nova/+bug/1639293 ? | |
| 11:59:33 | openstack | Launchpad bug 1639293 in OpenStack Compute (nova) "Cinder encrypted vol connection info include full nova internal class name " [High,In progress] - Assigned to Lee Yarwood (lyarwood) | |
| 12:00:32 | openstackgerrit | Monty Taylor proposed openstack/python-novaclient master: Stop posting to os-volumes_boot https://review.openstack.org/473833 | |
| 12:02:54 | lyarwood | sdague: yeah still in progress, we can't remove the full class path method until Queens thanks to Tempest | |
| 12:03:24 | lyarwood | https://review.openstack.org/#/q/status:open+branch:master+topic:deprecate_legacy_encryptor_provider_names | |
| 12:03:59 | sdague | lyarwood: is there going to still be another nova patch there? | |
| 12:04:17 | lyarwood | sdague: nope, we can remove it from the bug now. | |
| 12:29:12 | openstackgerrit | Andrey Volkov proposed openstack/nova master: [placement] Add api-ref for traits https://review.openstack.org/474186 | |
| 12:47:23 | alex_xu | nova api meeting is in 13 mins at #openstack-meeting-4 | |
| 12:49:54 | openstackgerrit | Merged openstack/nova master: api-ref: fix misleading description in PUT /os-services/disable https://review.openstack.org/473997 | |
| 12:54:28 | sfinucan | moshele: I could, but surely we should wait until it's in the wild? | |
| 12:54:36 | sfinucan | ...and in a release | |
| 12:57:05 | moshele | sfinucan: ok | |
| 13:05:15 | Kevin_Zheng | ping gibi | |
| 13:05:24 | Kevin_Zheng | ping | |
| 13:11:28 | openstackgerrit | Matt Riedemann proposed openstack/python-novaclient master: Clean up ShellTest unit tests https://review.openstack.org/473910 | |
| 13:13:56 | Kevin_Zheng | ping mriedem | |
| 13:15:25 | mriedem | Kevin_Zheng: pong | |
| 13:15:33 | mriedem | sdague: want to put https://review.openstack.org/#/c/474029/ in and then i'll cut a stable/ocata release this afternoon | |
| 13:18:10 | Kevin_Zheng | mriedem: saw you guys mentioned me in notification meeting last night, I will have a patch up tomorrow, and I have a question came up today from our product team | |
| 13:18:45 | Kevin_Zheng | Should instance notification payload including STATUS? | |
| 13:19:04 | mriedem | it has vm_state doesn't it? | |
| 13:19:32 | Kevin_Zheng | As it appears in instance details response and we have filter for this field | |
| 13:19:58 | Kevin_Zheng | Yeah, but we currently allow filter by status | |
| 13:20:00 | mriedem | looking at https://docs.openstack.org/developer/nova/notifications.html | |
| 13:20:02 | Kevin_Zheng | In client | |
| 13:20:06 | mriedem | i see | |
| 13:20:07 | mriedem | "state":"active", "task_state":"pausing", "power_state":"running", | |
| 13:20:15 | mriedem | for instance.pause.start | |
| 13:21:05 | Kevin_Zheng | O we have state | |
| 13:21:33 | mriedem | Kevin_Zheng: yes i think that's the vm_state | |
| 13:21:52 | mriedem | i assume you're asking for something like the status in the rest api response for GET /servers/{server_id} which uses https://github.com/openstack/nova/blob/master/nova/api/openstack/common.py#L113 | |
| 13:22:08 | Kevin_Zheng | Yeah | |
| 13:22:23 | mriedem | i.e. https://developer.openstack.org/api-guide/compute/server_concepts.html | |
| 13:23:11 | mriedem | honestly i don't really know why there was ever a distinction in the rest api | |
| 13:23:25 | mriedem | i think the status field in the response predated the vm/task states being in the rest api response | |
| 13:23:45 | mriedem | which came in the OS-EXT-STS extension | |
| 13:23:52 | mriedem | it seems redundant t ome | |
| 13:24:10 | Kevin_Zheng | Hmm | |
| 13:24:32 | Kevin_Zheng | So you prefer not adding it | |
| 13:25:26 | Kevin_Zheng | And perhaps apps should rely on vm state instead of status | |
| 13:25:26 | sdague | mriedem: done | |
| 13:25:27 | Kevin_Zheng | ? | |
| 13:27:06 | mriedem | Kevin_Zheng: i don't have a strong feeling either way, we could add status to the InstancePayload if we wanted, i'd ask gibi too | |
| 13:27:11 | mriedem | but this is why we have versioned notifications now | |
| 13:27:14 | mriedem | so we can change them | |
| 13:27:50 | mriedem | Kevin_Zheng: i would like some more details on the use case, but that could come in a change or ML thread or something else | |
| 13:28:00 | mriedem | sdague: thanks | |
| 13:28:07 | Kevin_Zheng | yeah, I just find it might be useful as we already support filter by it | |
| 13:28:40 | sdague | mriedem: is there a current bug leader? | |
| 13:29:10 | stvnoyes | good morning mriedem. I have the cinder v3 migrate rv ready to go. I've been testing it with jgriffith 's POC changes and can migrate back and forth OK with a bfv instance and instances with attached volumes . I would appreciate your feedback on the changes. thanks. https://review.openstack.org/#/c/463987/ | |
| 13:29:10 | sdague | mriedem: I was staring at the bugs again today and thinking about the way tags get "abused" in github for workflow | |
| 13:29:19 | sdague | and think we could use some of that | |
| 13:29:30 | mriedem | sdague: there is no bug master | |
| 13:29:41 | mriedem | stvnoyes: ok | |
| 13:29:56 | mriedem | sdague: abused? | |
| 13:30:20 | sdague | mriedem: github has no real workflow states | |
| 13:30:25 | sdague | it's either open or clossed | |
| 13:30:50 | sdague | which means the entire incomplete / confirmed / triaged isn't used | |
| 13:30:55 | sdague | as it doesn't exist | |
| 13:31:12 | mriedem | oh right, i remember | |
| 13:31:13 | sdague | and instead projects use tags and bug bots to do workflow with tags | |
| 13:31:15 | mriedem | with GHE | |
| 13:31:25 | mriedem | inprogress, blocked, etc | |
| 13:31:31 | sdague | which, is actually a lot more flexible in a lot of ways | |
| 13:31:41 | sdague | because you can customize workflow to your project | |
| 13:31:51 | pabardina | Hello, I would like to use the "nova trigger-crash-dump", but how can I validate if it works correctly ? The cli doesn't do anything (on mitaka) thanks | |
| 13:32:39 | sdague | I was thinking about things like "needs.sub-team-triage" "needs.os-info" "needs.version-info" | |
| 13:32:42 | mriedem | pabardina: here are the functional tests for trigger crash dump in mitaka if those help https://github.com/openstack/python-novaclient/blob/stable/mitaka/novaclient/tests/functional/v2/test_trigger_crash_dump.py | |
| 13:33:44 | sdague | launchpad doesn't let you have : in tags, but it does let you have "." | |
| 13:34:07 | sdague | a bunch of that could be automatically applied with some text processing | |
| 13:34:19 | mriedem | Kevin_Zheng: when you say filter by status, you mean this right? https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/servers.py#L188-L208 | |
| 13:34:30 | mriedem | we really filter by vm/task state | |
| 13:34:37 | mriedem | the status thing is just an alias for the API user | |
| 13:34:46 | sdague | because a lot bugs just end up with no logical next step | |
| 13:35:06 | Kevin_Zheng | hmm yes | |
| 13:35:38 | mriedem | sdague: automatically applied how? cron job that pulls the bug down and scans the text and applies a tag? | |
| 13:35:54 | mriedem | i'm not even sure if we're supposed to be actively moving to storyboard or not | |
| 13:35:57 | mriedem | i've avoided that conversation | |
| 13:36:09 | sdague | mriedem: yeh, there are mqtt events on bug changes, we don't have to cron it | |
| 13:36:23 | Kevin_Zheng | but for SL for example, if we want to support filter by status, we have to maintain the mapping like in Nova | |
| 13:36:46 | Kevin_Zheng | vm_state + task_state -> status | |
| 13:36:54 | mriedem | Kevin_Zheng: true; when you say support filter by status in searchlight (SL reminds me of softlayer), are you talking about something added to the searchlight REST API? | |
| 13:37:31 | mriedem | sdague: yeah good point | |
| 13:38:41 | efried | johnthetubaguy yt? | |
| 13:38:51 | Kevin_Zheng | no, they are just transforming user input query params to ElasticSearch Query, so if it has this field, it can search | |
| 13:40:18 | sdague | mriedem: anyway, I'm playing with some things to see what might be useful to make the bug backlog not too crazy | |
| 13:41:01 | Kevin_Zheng | it can also transforming status query to vm_state + task_state, but means maintain the mapping :) | |
| 13:41:44 | mriedem | edleafe: dansmith: semi interesting placement related bug https://bugs.launchpad.net/nova/+bug/1697825 | |