Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2020-12-15
19:13:42 jawad_axd Hi folks! Error 43, windows 10 with gpu passthrough. Added img_hide_hypervisor_id=True to image property. Is there anything I need to do to load nvidia drivers in windows instance?
19:14:10 jawad_axd *anything else
19:48:50 openstackgerrit Ghanshyam proposed openstack/nova master: Reuse code from oslo lib for JSON policy migration https://review.opendev.org/c/openstack/nova/+/763487
19:49:44 openstackgerrit Ghanshyam proposed openstack/nova master: Reuse code from oslo lib for JSON policy migration https://review.opendev.org/c/openstack/nova/+/763487
20:02:16 openstackgerrit Elod Illes proposed openstack/nova stable/ussuri: [stable-only] fix lower-constraints and disable qos resize https://review.opendev.org/c/openstack/nova/+/766738
20:44:11 openstackgerrit Merged openstack/nova master: libvirt: make cross cell resize spawn from snapshot image https://review.opendev.org/c/openstack/nova/+/765561
#openstack-nova - 2020-12-16
00:13:33 openstackgerrit Ghanshyam proposed openstack/placement master: Remove deprecated [placement]/policy_file config option https://review.opendev.org/c/openstack/placement/+/764409
00:13:57 openstackgerrit Ghanshyam proposed openstack/placement master: [goal] Deprecate the JSON formatted policy file https://review.opendev.org/c/openstack/placement/+/764394
00:15:14 gmann gibi: bauzas stephenfin please check these two placement patches also. for json->yaml migration community goal. https://review.opendev.org/c/openstack/placement/+/764409 https://review.opendev.org/c/openstack/placement/+/764394
00:16:04 gmann 764409 is to remove the deprecated policy_file config option which is needed for this migration
00:56:56 brinzhang gibi, bauzas, sean-k-mooney: Since Microversion 2.77, we support specifying an avaiability_zone to unshleve an instance, it allowed {'unshelve': null} and {'unshelve': {'availability_zone': }}, but we restrict the empty dict {'unshelve': {}} as the request body for unshelve. this restrict asked by gmann in https://review.opendev.org/c/openstack/nova/+/663851/16/nova/api/openstack/compute/schemas/servers.py#663
00:57:55 brinzhang gibi, bauzas, sean-k-mooney, gmann: this is the record https://review.opendev.org/c/openstack/nova/+/663851/13/nova/api/openstack/compute/schemas/servers.py#659
00:59:50 gmann brinzhang: yeah as this action API need specific request body so disallow empty make sense. for other action API there is no specific request body so allowing null or anything is what we have during legacy way
00:59:55 brinzhang and we note that in https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/schemas/shelve.py#L28-L31
01:02:15 brinzhang gmann: yes, I think gibi and bauzas's concern in the docs description, because of the availability_zone is optional, but we allowed request {'unshelve': {'availability_zone': <string>}}, if we dont chose the availability_zone, it will be {'unshelve': {}
01:03:11 gmann ok and we only allow {'unshelve': null} in that case
01:04:45 gmann brinzhang: may be we can update it in api-ref too https://docs.openstack.org/api-ref/compute/?expanded=unshelve-restore-shelved-server-unshelve-action-detail
01:04:54 brinzhang yes, I think this is reasonable too
01:05:00 gmann saying 2.77 onwards
01:05:44 brinzhang gmann: ok, let submit a patch to add a Note for the availability_zone
01:06:07 brinzhang s/let/ let me
01:06:13 gmann +1
01:06:46 brinzhang Will do, and pls review it later
01:08:04 gmann thanks
01:12:29 brinzhang gmann: how abot saying: Since microversion 2.77, we are allowed to request {'unshelve': null} and {'unshelve': {'availability_zone': <string>}}, but not allowed to request {'unshelve': {}}.
01:16:10 gmann brinzhang: yeah, little modification- Since microversion 2.77, allowed request body schema are {'unshelve': null} or {'unshelve': {'availability_zone': <string>}}, and {'unshelve': {}} request body is not allowed.
01:31:08 brinzhang gmann: and let me register a bug to trace this change
01:32:57 sean-k-mooney gmann: brinzhang we were suggesting allowing {} ealier today
01:33:10 sean-k-mooney i think its perfectly consitent to treat null and {} the same
01:35:20 gmann sean-k-mooney: {} means allowed everything or any unknown request body which is later ignored so confusing for users. where {'unshelve': null} means allow only empty request body
01:35:53 gmann sean-k-mooney: like for {}. i can pass {'unshelve': {'unknown': 'anything', ....}}
01:36:16 gmann that was the reason we thought of improving it in 2.77
01:50:51 openstackgerrit Brin Zhang proposed openstack/nova master: Improving the description for az of unshleve API https://review.opendev.org/c/openstack/nova/+/767251
01:51:52 brinzhang sean-k-mooney: yeah, I agree with gmann, while I do the change of specifying az to unshelve a server, already talked, also with mriedem
01:52:19 brinzhang gmann: done, pls review https://review.opendev.org/c/openstack/nova/+/767251
02:47:55 openstackgerrit Merged openstack/nova master: Refactoring of functional.regression.test_bug_1702454 https://review.opendev.org/c/openstack/nova/+/765997
06:11:30 openstackgerrit Wenping Song proposed openstack/nova master: Resume accelerator guest state when a host is booted https://review.opendev.org/c/openstack/nova/+/767273
08:10:11 openstackgerrit Mamduh proposed openstack/os-vif stable/ussuri: Fix - os-vif fails to get the correct UpLink Representor https://review.opendev.org/c/openstack/os-vif/+/765967
08:16:26 openstackgerrit Merged openstack/placement master: Fix l-c job and move to latest hacking 4.0.0 https://review.opendev.org/c/openstack/placement/+/766994
08:20:51 openstackgerrit Mamduh proposed openstack/os-vif stable/train: Fix - os-vif fails to get the correct UpLink Representor https://review.opendev.org/c/openstack/os-vif/+/765970
08:22:46 gibi brinzhang, gmann: thanks for the historical context on unshelve: {}
08:23:52 gibi so the goal was to only allow unshelve:null but not unshelve:{'unknow': ''} but this still not mean we have to forbid unshelve:{}
08:33:21 openstackgerrit Mamduh proposed openstack/os-vif stable/stein: Fix - os-vif fails to get the correct UpLink Representor https://review.opendev.org/c/openstack/os-vif/+/765974
08:56:04 openstackgerrit Mamduh proposed openstack/os-vif stable/rocky: Fix - os-vif fails to get the correct UpLink Representor https://review.opendev.org/c/openstack/os-vif/+/765977
09:00:51 openstackgerrit Mamduh proposed openstack/os-vif stable/queens: Fix - os-vif fails to get the correct UpLink Representor https://review.opendev.org/c/openstack/os-vif/+/765983
09:01:40 gibi I think what what brinzhang proposed in the comment of https://review.opendev.org/c/openstack/nova/+/663851/13/nova/api/openstack/compute/schemas/servers.py#659 allows unshelve: {}, unshelve:{"availability_zone": "foo"}, unshelve: null, but forbid unshelve: {"foo": "bar"}
09:01:52 gibi bauzas: you too ^^
09:02:07 gibi which I think what is bauzas after
09:05:33 brinzhang gibi: yes, that what I want to do privous, but after talked with gmann and mriedem, I was restrict the unshelve:{} request after PS13
09:06:50 gibi brinzhang: I see that gmann requested the restriction of unshelve: {},
09:07:07 gibi I don't see mriedem requesting that though
09:07:54 gibi gmann: could you elaborate on 01:59 < gmann> brinzhang: yeah as this action API need specific request body so disallow empty
09:07:56 brinzhang maybe talked in IRC, sorry, I dont have the cleayly remember
09:07:57 gibi make sense. for other action API there is no specific request body so allowing
09:08:00 gibi null or anything is what we have during legacy way
09:08:24 gibi gmann: what do you mean by specific request body? how this is differ from the lock action?
09:08:36 gibi brinzhang: no worries, we have gmann to help :)
09:08:56 gibi (it is just the timezone differences that will make it hard to get a decisions)
09:09:26 brinzhang gibi: and Isubmit a patch for update the description of unshleve az parameter https://review.opendev.org/c/openstack/nova/+/767251
09:28:51 elod lyarwood if you have a minute: can you hit this one? >>> https://review.opendev.org/c/openstack/nova/+/766497 (bandit 1.7.0 is not yet yanked, so maybe it's worth to push the bandit cap fixes, too... though that's also not a quick fix...)
09:29:58 elod the swift side has already merged
09:41:18 lyarwood elod: ACK'd lets do this
09:53:09 bauzas gibi: morning, saw your pings with brinzhang
09:54:13 gibi lyarwood: about the lower contraint jobs, after elod summarized the situation to me as a nova PTL I think if we got blocked on the lower constraints jobs and we have no time to fix it then make it non-voting and ask the help of the deployers relying on that job to fix it
09:56:18 elod and just an addition: lower-constraints jobs are (mostly) fixed, so this ^^^ is only valid for the future I think
09:56:59 elod lyarwood: thx \o/
09:59:04 bauzas does anyone know how to fix a blank page from Zuul with Firefox ?
09:59:10 bauzas it does work fine with Chrome
09:59:20 bauzas and I flushed cookes
09:59:22 bauzas cookies*
10:00:15 gibi bauzas: havn't seen that problem yet in my firefox
10:00:30 bauzas that's weird
10:00:33 bauzas I get the 200s
10:00:53 bauzas but looks like javascript is blocked or something like this
10:01:13 bauzas ah, gotcha
10:01:16 bauzas got a console exception
10:01:23 bauzas Exception { name: "NS_ERROR_FILE_CORRUPTED", message: "", result: 2152857611, filename: "https://zuul.opendev.org/static/js/main.a8854164.chunk.js"
10:02:14 bauzas gibi: could you try refreshing your local cache ?
10:02:27 bauzas anyhow, the gate looks busted
10:02:40 gibi let me check in a private mode
10:03:01 bauzas all the jobs are in "queued" state
10:03:42 gibi bauzas: this openes for me in a private mode or in an empty firefox profile https://zuul.opendev.org/t/openstack/build/fce6478a6842495f851b8c88336f0049/log/compute1/logs/screen-c-vol.txt
10:04:15 bauzas super weirdo
10:05:00 gibi there are jobs running https://zuul.opendev.org/t/openstack/status
10:05:08 bauzas gibi: could you please tell me which main.js version you have ?
10:05:11 gibi but we have 155 check queu
10:05:19 bauzas mine https://zuul.opendev.org/static/js/main.a8854164.chunk.js is failing
10:05:30 gibi bauzas: on which page?
10:05:57 gibi I have this https://zuul.opendev.org/static/js/main.67746290.chunk.js
10:06:27 bauzas not the same version
10:06:41 gibi I'm not even sure it is them same main
10:06:46 gibi who knows
10:08:15 bauzas gibi: this JS was loaded on https://zuul.opendev.org/t/openstack/build/fce6478a6842495f851b8c88336f0049/log/compute1/logs/screen-c-vol.txt
10:08:22 gibi I have this as well https://zuul.opendev.org/static/js/main.a8854164.chunk.js
10:08:27 gibi in another pag
10:08:28 gibi e
10:08:37 bauzas which FF version ?
10:08:53 bauzas mine is 83.0
10:09:15 gibi 83.0 (64-bit)
10:09:18 bauzas and the check pipeline is super late, indeed

Earlier   Later