Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2020-04-24
16:02:34 bauzas but only have nits for the moment
16:03:43 bauzas gibi: thanks, I have a virtual drink with friends in a couple of mins ;)
16:03:58 gibi \o/
16:13:34 gmann dansmith: good point. anyways api-ref has those changes documented. good to go from me.
16:14:09 dansmith cool
16:18:27 spatel sean-k-mooney: how are you, hope you doing well
16:18:45 spatel I have question related cpu mode="host-passthrough" vs cpu mode="host-model"
16:19:24 spatel As you know host-passthrough good for performance but do we know how much performance its going to improve
16:19:49 spatel as i know host-passthrough is just exposing more flags to VM but does that improve performance?
16:26:05 mgoddard bauzas: ok, good start
16:36:53 openstackgerrit Merged openstack/nova-specs master: Update SPEC: Using PUT instead of PATCH API to update delete_on_termination https://review.opendev.org/722194
16:47:47 kashyap artom: Saw your comments; thank you. I'll respond on Monday. But a quick note here after reading your ocmments:
16:48:40 kashyap artom: No, we shouldn't mark 'pc' as deprecated -- the story is "complex" here -- but it remains a valid use for many workloads, 'legacy' _and_ non-legacy
16:49:35 kashyap artom: So, not changing the default in Nova itself is a valid option, for the reason libvirt did it. When you get a minute, take time read this commit message from DanPB:
16:49:38 kashyap https://libvirt.org/git/?p=libvirt.git;a=commit;h=26cfb1a3cd
16:49:43 kashyap I need to head out, but thanks, again!
16:50:05 kashyap (I'll expand more on the spec.)
16:54:26 openstackgerrit Merged openstack/nova stable/stein: Make RBD imagebackend flatten method idempotent https://review.opendev.org/708615
17:03:36 artom kashyap, akc, thanks!
17:06:51 openstackgerrit Merged openstack/nova stable/stein: Reject boot request for unsupported images https://review.opendev.org/708578
17:12:22 kashyap artom: Actually, I responded now. Couldn't resist. Some of your comments just "invited me out to respond" :D
17:12:46 artom kashyap, yeah, I have that troll'ish effect :)
18:17:03 openstackgerrit Merged openstack/python-novaclient master: Switch to using TOX_CONSTRAINTS_FILE https://review.opendev.org/722795
18:32:37 masterpe Hi the cache mode of a vm of e existing vm is it stored in a database?
18:55:59 openstackgerrit Merged openstack/python-novaclient master: Bump default tox env from py37 to py38 https://review.opendev.org/722885
20:43:25 masterpe I'm trying to change the cache mode of running instances from writeback to writethrough. I have changed the nova.conf to use writethrough. This works find for new instances. But for existing instances after live migration the mode on the new compute node it is still writeback? How can I fix this?
20:52:42 melwitt masterpe: I think you will need to stop/start or hard reboot or migrate the instances to make them generate new xml and take the new cache mode
20:53:48 melwitt oh but you're saying after you changed the conf and live migrated, it kept the old cache mode? maybe cold migration only will regenerate xml. sorry I don't know for sure
21:03:58 artom melwitt, yeah, I don't see anything about cache in libvirt/migration.py, I'm pretty sure we don't update that
21:05:09 melwitt artom: ah, thanks for checking that. I didn't know if it was magic inside libvirt that moves the guest (xml) over or if it was something in nova. good to know
21:07:28 artom melwitt, it's all Nova
21:08:21 melwitt thanks
21:34:44 openstackgerrit melanie witt proposed openstack/nova master: zuul: Switch to the Zuulv3 grenade job https://review.opendev.org/704364
22:23:05 masterpe so libvirt generated the xml file and nova commands libvirt to migrate an instance to migrate an instance from hostA to hostB?
22:51:22 openstackgerrit melanie witt proposed openstack/nova stable/rocky: Lowercase ironic driver hash ring and ignore case in cache https://review.opendev.org/723050
22:58:13 melwitt masterpe: yes and nova passes new xml to use on the destination https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L8846
23:04:46 melwitt so to generate new xml you have to do server stop + start, or hard reboot, or cold migration, or rebuild
23:05:26 melwitt or resize, shelve + unshelve
23:22:27 openstackgerrit melanie witt proposed openstack/nova stable/queens: Lowercase ironic driver hash ring and ignore case in cache https://review.opendev.org/723054
23:27:17 openstackgerrit melanie witt proposed openstack/nova stable/pike: Lowercase ironic driver hash ring and ignore case in cache https://review.opendev.org/723055
#openstack-nova - 2020-04-25
08:14:29 openstackgerrit Merged openstack/python-novaclient master: Add py38 package metadata https://review.opendev.org/722753
08:33:21 openstackgerrit OpenStack Proposal Bot proposed openstack/nova master: Imported Translations from Zanata https://review.opendev.org/722644
10:56:41 openstackgerrit Takashi Natsume proposed openstack/python-novaclient master: doc: Update Testing document https://review.opendev.org/723078
15:18:06 openstackgerrit Merged openstack/nova master: Add Python3 victoria unit tests https://review.opendev.org/722523
23:10:38 openstackgerrit Ghanshyam Mann proposed openstack/nova-specs master: [Community goal] Update contributor documentation https://review.opendev.org/722393
#openstack-nova - 2020-04-26
01:16:12 openstackgerrit xuyuanhao proposed openstack/nova master: loads pyc files failed,sys.modules support loads py and pyc files as modules https://review.opendev.org/723130
06:35:12 openstackgerrit Qiu Fossen proposed openstack/python-novaclient master: Remove Babel requirement https://review.opendev.org/723139
06:40:56 openstackgerrit Qiu Fossen proposed openstack/python-novaclient master: Remove Babel requirement https://review.opendev.org/723139
07:18:48 openstackgerrit Takashi Natsume proposed openstack/python-novaclient master: Use unittest.mock instead of third party mock https://review.opendev.org/723152
07:19:42 openstackgerrit Takashi Natsume proposed openstack/python-novaclient master: Remove future imports https://review.opendev.org/723153
07:44:04 openstackgerrit OpenStack Proposal Bot proposed openstack/nova stable/ussuri: Imported Translations from Zanata https://review.opendev.org/723160
07:51:30 openstackgerrit OpenStack Proposal Bot proposed openstack/nova master: Imported Translations from Zanata https://review.opendev.org/722644
10:12:26 openstackgerrit Takashi Natsume proposed openstack/python-novaclient master: Use unittest.mock instead of third party mock https://review.opendev.org/723152
10:41:38 openstackgerrit Qiu Fossen proposed openstack/python-novaclient master: Remove Babel requirement https://review.opendev.org/723139
11:16:43 openstackgerrit Andreas Jaeger proposed openstack/nova master: Remove translation sections from setup.cfg https://review.opendev.org/723206
11:43:32 openstackgerrit Andreas Jaeger proposed openstack/nova master: Remove translation sections from setup.cfg https://review.opendev.org/723206
13:27:01 openstackgerrit jayaditya gupta proposed openstack/nova master: Support for --force flag for nova-manage placement heal_allocations command https://review.opendev.org/715395
19:34:10 openstackgerrit jayaditya gupta proposed openstack/nova master: Support for --force flag for nova-manage placement heal_allocations command https://review.opendev.org/715395
20:57:50 openstackgerrit Luigi Toscano proposed openstack/nova master: DNM a test for a grenade fix https://review.opendev.org/723266
#openstack-nova - 2020-04-27
00:14:00 openstackgerrit Ghanshyam Mann proposed openstack/nova master: zuul: Switch to the Zuulv3 grenade job https://review.opendev.org/704364
00:46:30 openstackgerrit Kevin Zhao proposed openstack/nova master: [WIP] CI: add tempest-integrated-compute-aarch64 job https://review.opendev.org/714439
01:28:19 openstackgerrit Brin Zhang proposed openstack/python-novaclient stable/train: Update master for stable/train https://review.opendev.org/723290
02:03:42 openstackgerrit Brin Zhang proposed openstack/nova stable/train: Update master for stable/train https://review.opendev.org/723295
02:44:24 openstackgerrit xuyuanhao proposed openstack/nova master: failed to loads pyc file's classes https://review.opendev.org/723301
03:40:47 openstackgerrit Arthur Dayne proposed openstack/os-resource-classes master: Add new resource class: PSSD, VSSD https://review.opendev.org/723303
04:46:47 openstackgerrit Arthur Dayne proposed openstack/os-resource-classes master: Add new resource class: SSD https://review.opendev.org/723303
05:20:48 openstackgerrit Kevin Zhao proposed openstack/nova master: [WIP] CI: add tempest-integrated-compute-aarch64 job https://review.opendev.org/714439
05:53:51 openstackgerrit Arthur Dayne proposed openstack/os-resource-classes master: Add new resource class: SSD https://review.opendev.org/723303
06:15:27 bauzas gibi: I'm taking an half-day PTO this morning, see you this afternoon
06:46:17 openstackgerrit Andrey Volkov proposed openstack/nova master: [WIP] Image auto signature https://review.opendev.org/723320
07:05:34 gibi bauzas: good morning. ACK.
07:33:31 zigo root@C1-z-controller-1>_ ~ # openstack flavor create --format shell octavia_65 --private --id 65 --ram 2048 --disk 4 --vcpus 1
07:33:31 zigo With ussuri, I'm getting:
07:33:32 zigo Policy doesn't allow os_compute_api:os-flavor-manage:create to be performed. (HTTP 403) (Request-ID: req-30fe38ae-deb9-451f-9234-58edd691696b)
07:33:56 zigo Is there something wrong with the scope enforcement? It's by default to False as it should be ...
07:36:16 zigo When I set rule:admin_api instead, then it works. So something's wrong in oslo.policy or what?
07:38:19 gibi zigo: there was a wide change in the policy handling in Ussuri but the new behavior should be off by default
07:38:44 zigo Which is what I'm saying, it shouldn't be enforced by default, but it looks like it is!
07:45:29 gibi zigo: did you use a admin token in the failed case?
07:45:48 zigo I'm doing this as an admin user indeed.
07:49:56 gibi I have default policy config and it works for me http://paste.openstack.org/show/792738/
07:52:09 zigo gibi: You can try with the Debian packages, maybe ? :)
07:52:42 zigo deb http://buster-ussuri.debian.net/debian buster-ussuri-backports-nochange main
07:52:42 zigo deb http://buster-ussuri.debian.net/debian buster-ussuri-backports main
07:52:42 zigo For Buster:
07:52:53 zigo Or if you are more adventurous, just from Experimental! :P
07:56:46 gibi zigo: I have to jump on a call for a while, gmann, johnthetubaguy, stephenfin if you can help zigo in the meantime that would be appreciated
07:57:37 gibi zigo: I would check oslo.policy version as there was a late change there as well
07:57:57 gibi zigo: and you can try to rollback https://review.opendev.org/#/c/714822 maybe, but I'm not sure
07:58:18 zigo python3-oslo.policy 3.1.0-1~bpo10+1
07:58:18 zigo # dpkg-query -W python3-oslo.policy
07:59:33 zigo So that's latest release ...
07:59:41 gibi yepp, that is the last one
08:09:10 zigo Reverting that patch doesn't fix the problem.
08:09:46 gibi zigo: ack, then I'm out of ideas at the moment
08:10:04 gibi and sitting on a call so will be slow responding
08:20:50 zigo I can switch the packaging from rule:system_admin_api to rule:admin_api in the default policy.conf, but obviously, something is wrong that needs to be fixed.
08:21:33 zigo Or is there a way to give the system_scope:all to my admin user?
08:24:37 frickler zigo: you can call for system scoped tokens in the openstack client command
08:25:03 zigo frickler: How?

Earlier   Later