Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-09-07
19:26:39 mriedem and why
19:26:43 mriedem like, in the operators ML
19:27:48 Nisha_Agarwal ironic team says nova should agree to make changes for all this then its fine to do corresponding changes in ironic
19:28:29 Nisha_Agarwal mriedem, i cannot get such list but i do know one customer atleast(cannot name the customer) who does have such env
19:28:44 mriedem Nisha_Agarwal: can you start a thread about this in the openstack-operators mailing list?
19:28:51 Nisha_Agarwal and they just have ironic nodes in their datacenter
19:29:01 mriedem tag it with [nova][ironic]
19:29:21 Nisha_Agarwal is it not our usual mailing list?
19:29:32 mriedem you could cross-post to the openstack-dev mailing list
19:29:43 mriedem but i'm mostly interested in hearing from ironic operators
19:30:25 Nisha_Agarwal ok. will just start a thread. Never wrote in this mailing list though :)
19:30:36 mriedem i have faith that you can do it
19:30:38 openstackgerrit Matthew Treinish proposed openstack/nova master: Set regex flag on ostestr command for osprofiler tests https://review.openstack.org/501842
19:30:43 mtreinish mriedem: ^^^ lets see if that does it
19:32:40 mriedem i will hold my breath in anticipation
19:33:02 Nisha_Agarwal mriedem, :)
19:36:58 Nisha_Agarwal mriedem, it would be good if you could suggest the subject for the mail....i dont know what to write actually
19:37:10 Nisha_Agarwal in the subject
19:37:14 mriedem hmm
19:37:15 Nisha_Agarwal body i can fill up
19:37:43 mriedem [nova][ironic] Concerns over rigid resource class-only ironic scheduling
19:37:43 mriedem ?
19:39:59 Nisha_Agarwal mriedem, looks good :)
19:42:08 mriedem stvnoyes: well looky here http://logs.openstack.org/05/501805/2/check/gate-tempest-dsvm-multinode-live-migration-ubuntu-xenial/069e82e/console.html#_2017-09-07_17_45_00_706868
19:45:46 Nisha_Agarwal mriedem, do i need to subscribe to openstack-operators@lists.openstack.org for sending an email?
19:45:51 mriedem yeah
19:45:55 Nisha_Agarwal ok
19:46:38 mriedem ildikov: why does https://review.openstack.org/#/c/330285/ need to depend on https://review.openstack.org/#/c/498828/ ?
19:47:50 ildikov mriedem: uWSGI changes in Cinder blew up
19:48:13 mriedem why would that matter here?
19:49:00 Nisha_Agarwal mriedem, i sent an email just now. if you would want to add to any thing to the body of the mail please add it.
19:49:44 ildikov mriedem: because the url wasn't working with the microversion checks in the client
19:50:02 Nisha_Agarwal mriedem, meanwhile i would open a blueprint at nova side or i should do it after responses on the email and probably after the PTG?
19:50:13 Nisha_Agarwal just asking
19:50:55 ildikov mriedem: smcginnis might be able to describe it in details
19:52:48 smcginnis Wha?
19:53:29 smcginnis Ah
19:53:53 smcginnis mriedem: There's an issue right now with the way the public endpoint is being set and how that works for microversions.
19:54:01 smcginnis So that needs to get fixed first.
19:56:24 stvnoyes mriedem: nice to see it pass. I also figured out why I put the change in the iscsi test. the other live migrate block test was skipped dur a bug. I just removed that skip decorator - https://review.openstack.org/#/c/501828/ . If you're OK with that, I'll add the new attach checks to that test too.
19:56:48 stvnoyes dur = 'due to'
19:57:28 mriedem Nisha_Agarwal: replied with some more details from our irc discussion
19:57:31 mriedem thanks for starting that thread
19:58:10 mriedem stvnoyes: probably less worthwhile to add things to a skipped test in tempest
19:58:24 mriedem fwiw this does test that volume-backed live migration test https://review.openstack.org/#/c/481290/
19:58:34 mriedem because it depends on https://review.openstack.org/#/c/481346/
19:58:58 Nisha_Agarwal mriedem, thanks for the reply there
19:59:33 stvnoyes ah you already did it. ok I'll abandon my review since it does the same thing
19:59:50 Nisha_Agarwal mriedem, with respect to "eventually lead to scheduling failures later because of the scheduler thinking a node is available for more than one instance when it's really not." doesnt happen in the customer env where he uses non-exact filters
20:00:29 mriedem mtreinish: i'm +2 on your ostestr change, maybe you want to review this now https://review.openstack.org/#/c/498828/
20:00:48 Nisha_Agarwal mriedem, the ironic node doesnt get scheduled again if use non-exact filters
20:02:34 Nisha_Agarwal atleast with this customer we have never heard of this issue
20:02:38 mtreinish mriedem: +2
20:03:04 mtreinish although I wonder how much the tls-proxy is used now. As things move to use uwsgi the standalone proxy isn't used anymore
20:12:14 openstackgerrit Matt Riedemann proposed openstack/nova-specs master: Add spec to use cinder's new attachment API https://review.openstack.org/497552
20:16:22 ildikov mriedem: thanks ^^
20:16:57 mriedem ildikov: johnthetubaguy found the same thing i did in https://review.openstack.org/#/c/493323/
20:18:45 ildikov mriedem: I bumped the version for all the calls to 3.44 in the attach patch
20:19:50 mriedem that's not clear here, it wasn't called out in the commit message, and it would probably just be cleaner if we did it in this change
20:20:12 ildikov mriedem: I got lost on the newer version note in the review though
20:20:30 mriedem mine?
20:20:31 ildikov I was debating where to do it and thought to try to keep the patches clean
20:20:47 mriedem i noticed the skip_version_check=True thing,
20:20:55 mriedem which is wrong here, as i noted
20:21:20 ildikov because I haven't bumped all versions?
20:21:41 mriedem i'd just bump the version for attachment_create to be 3.44 as john pointed out, update whatever unit tests and make a comment about it in the commit message that we need the minimum microversion for creating the attachment to equal the microversoin needed for complete
20:21:57 mnaser abaranov https://review.openstack.org/#/c/501820/1 looks lik ethis failed somewhere along the line
20:22:05 mnaser it doesnt seem unrelated but does anyone know the bug # offhand
20:22:12 ildikov mriedem: that's what I planned to do
20:23:21 mriedem mnaser: unrelated, just recheck it
20:23:26 ildikov mriedem: as I said, I had two choices on where to bump and I chose what seemed cleaner to me, but agree with this variation too so will change when I get there later today
20:23:31 mnaser done, ty
20:23:53 mriedem mnaser: btw, wanted to ask, how do you pick up changes from stable branches? do you wait for a release or just pull from the branch?
20:24:03 mnaser mriedem rdo trunk stable builds
20:24:14 mriedem ok so don't wait for release
20:24:20 mnaser in cases like this, no
20:24:32 mriedem do you patch rpms until the changes are merged upstream then?
20:24:51 mnaser in very urgent scenarios yes, with puppet after the rpm is installed (not pretty)
20:25:00 mnaser but for the most we try to get it into stable branch asap and then deploy those packages
20:25:11 mriedem yeah
20:25:52 mriedem dansmith: these stable backports are up when you've got the time https://review.openstack.org/#/q/topic:bug/1715462+branch:stable/pike
20:26:05 mriedem mtreinish: could you also go through these ^ ?
20:26:11 mriedem nasty regressions for pike
20:26:21 mriedem and we're short on stable cores atm
20:26:29 mriedem that aren't in europe anyway
20:29:46 openstackgerrit Merged openstack/nova-specs master: Add spec to use cinder's new attachment API https://review.openstack.org/497552
20:41:13 mtreinish mriedem: sure, I'll take a look
21:00:50 openstackgerrit Merged openstack/nova master: Make eventlet hub use a monotonic clock https://review.openstack.org/434327
21:14:42 openstackgerrit Matt Riedemann proposed openstack/nova master: Drop support for the Cinder v2 API https://review.openstack.org/501874
21:15:10 smcginnis Woot! ^
22:05:01 openstackgerrit Eric Fried proposed openstack/nova master: Use ksa adapter for placement conf & requests https://review.openstack.org/492247
22:05:02 openstackgerrit Eric Fried proposed openstack/nova master: nova.utils.get_ksa_adapter() https://review.openstack.org/488137
22:09:09 openstackgerrit Ildiko Vancsa proposed openstack/nova master: Add attachment_complete call to volume/cinder.py https://review.openstack.org/493323
22:09:10 openstackgerrit Ildiko Vancsa proposed openstack/nova master: Tweak connection_info translation for the new Cinder attach/detach API https://review.openstack.org/493324
22:09:11 openstackgerrit Ildiko Vancsa proposed openstack/nova master: Implement new attach Cinder flow https://review.openstack.org/330285
22:19:04 openstackgerrit Merged openstack/nova master: Transform keypair.delete notification https://review.openstack.org/463946
22:26:04 openstackgerrit Eric Fried proposed openstack/nova master: Use ksa adapter for placement conf & requests https://review.openstack.org/492247
22:27:59 cdent efried: I don’t think you are being sufficiently critical of yourself in your comments
22:28:32 efried cdent I know what I mean, and I'm VICIOUS.
22:28:42 cdent :)

Earlier   Later