Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2020-04-08
11:56:00 sean-k-mooney brinzhang: are you using bool_from_string because x is not a vaild value
11:56:12 sean-k-mooney brinzhang: with strict 'x' should raise an error
11:56:28 brinzhang >>> from oslo_utils import strutils
11:56:28 brinzhang >>> strutils.bool_from_string('dd')
11:56:28 brinzhang False
11:56:47 brinzhang sean-k-mooney: I test in python3
11:56:59 sean-k-mooney right that is becasue you are hitting this https://github.com/openstack/oslo.utils/blob/master/oslo_utils/strutils.py#L157-L158
11:57:18 brinzhang yeah
11:57:34 sean-k-mooney the default is false and if you dont have stict then any unrecognised value returns the default
11:59:12 brinzhang yes, the strict=Fasle, if we change this to True, it will be an invalid value
12:07:13 brinzhang lyarwood:nova.exception.ValidationError: Invalid input for field/attribute delete_on_termination. Value: xxx. 'xxx' is not one of [True, 'True', 'TRUE', 'true', '1', 'ON', 'On', 'on', 'YES', 'Yes', 'yes', False, 'False', 'FALSE', 'false', '0', 'OFF', 'Off', 'off', 'NO', 'No', 'no']
12:07:53 brinzhang while I change the value to error, it cannot pass the schema check, do I necessary to do the bad value test?
12:10:00 lyarwood brinzhang: ah no then sorry, the schema check is enough.
12:10:14 lyarwood brinzhang: but we should still leave strict=True on to be safe
12:10:18 brinzhang lyarwood: yeah, I will give up this test
12:10:38 brinzhang and add body['volumeAttachment']['delete_on_termination'] = '1' to assertTure()
12:11:10 brinzhang lyarwood: yes, I was added strict=True to the bool_from_string()
12:11:27 lyarwood brinzhang: ack yeah anything to assert true being passed in would be nice to have
12:14:30 stephenfin gibi: Oh, that's a good point. It probably doesn't now :)
12:14:45 gibi \o/
12:16:13 openstackgerrit Brin Zhang proposed openstack/nova master: Convert delete_on_termination from string to boolean https://review.opendev.org/718387
12:16:15 jkulik is there anything in or around nova, that would help defragment host memory? move VMs around, so that a bigger VM can fit on a host again
12:16:28 brinzhang lyarwood, sean-k-mooney: update done.
12:16:42 brinzhang good morning all, have a nice day, bye.
12:16:56 gibi brinzhang: thanks! good night!
12:17:58 lyarwood \o
12:18:24 gibi jkulik: nova itself allows you to move VMs to other host. top of nova there is https://wiki.openstack.org/wiki/Watcher but don't know the health of the watcher project
12:20:49 jkulik thanks for the hint
12:35:31 bauzas stephenfin: question, you turned up your vote to +2 for lyarwood's, could you please tell me why ? did lyarwood provided a FUP or your comment on https://review.opendev.org/#/c/700812/21/nova/virt/libvirt/blockinfo.py@99 is wrong ?
12:35:41 bauzas because https://review.opendev.org/#/c/700812/21/nova/virt/libvirt/blockinfo.py@99 afraids me
12:35:55 stephenfin because I was expecting a follow-up
12:36:00 stephenfin the -1 was accidental
12:37:57 bauzas cool then
12:38:13 bauzas stephenfin: and you don't see a problem with https://review.opendev.org/#/c/700812/21/nova/virt/libvirt/blockinfo.py@99
12:38:13 bauzas ?
12:39:11 stephenfin they should be combined, yes, but I didn't think it was a blocking issue so long as it was resolved in a follow-up
12:39:26 stephenfin Happy to drop the +2 until that follow-up is there though
12:39:53 huaqiang @stephenfin: can you have a look at url:https://review.opendev.org/#/c/718345/
12:40:21 huaqiang if this patch is accepted, some of mixed instance patch will be easier
12:41:26 stephenfin huaqiang: We have to work without that because feature freeze has passed for oslo
12:41:47 stephenfin so it'll be Victoria before we can merge that
12:41:52 huaqiang ok
12:42:22 huaqiang I'll make workaround for now
12:43:39 openstackgerrit Stephen Finucane proposed openstack/nova master: Follow-up for flavor-extra-spec-validators series https://review.opendev.org/718357
12:47:05 openstackgerrit Stephen Finucane proposed openstack/nova master: api: Add framework for extra spec validation https://review.opendev.org/704643
12:48:49 bauzas lyarwood: left a comment giving you only a +1 on https://review.opendev.org/#/c/700812/21
12:49:37 bauzas tl;dr: I understand this can be unstable to rescue an instance with attached disks, but this is *not* unstable when we talk of a single instance with only one disk
12:50:01 bauzas (and this saved me more than a couple of times when I was operating an Essex cloud)
12:50:39 bauzas lyarwood: also, remember me, but can't you detach disks before rescuing ?
12:51:05 bauzas that won't make your workloads happy for sure, but you can still access the guest
12:52:12 bauzas it looks to me very like booting as single user, you only get a very limited access but you got it to
12:52:14 bauzas tho*
13:02:26 lyarwood bauzas / stephenfin ; I somehow missed that rebase error, I can get rid of that now assuming stephenfin has had to push changes to his series I'm based on
13:03:41 stephenfin lyarwood: That or a follow-up. I'm hoping I don't have to rebase the rest of my series but I think Gerrit will force me to
13:04:27 lyarwood bauzas: I can remove this use of unstable here and just call it a rescue. It's still unstable in the sense of disk ordering even with a single disk instance but I can see how this term might confuse people.
13:05:40 lyarwood bauzas: you can detach non-root disks but not bfv root disks.
13:06:22 lyarwood bauzas: and yeah stable rescue essentially gives you the same experience as you would have with bare metal machines entering into rescue mode
13:06:28 lyarwood bauzas: everything attached in the same order etc.
13:06:52 lyarwood stephenfin: kk fup might be easier in that case
13:07:45 openstackgerrit sean mooney proposed openstack/nova master: cyborg evacuate support https://review.opendev.org/715326
13:08:57 sean-k-mooney gibi: dansmith ^ just fixed the pep8 issues and a minor style tweek but tha is otherwise the same as the one i pushed last night
13:09:18 gibi sean-k-mooney: ack
13:09:48 sean-k-mooney dansmith: gibi i tested that local last night and it apears to correcly handel rebuild and evac including all the arq binding and placementa allocations
13:10:20 sean-k-mooney im not sure if we want to included it in ussuri but it should be read for review
13:12:14 gibi stephenfin: fyi relevant testcase failed in the FUP of extra_specs
13:15:18 sean-k-mooney lyarwood: :)
13:16:10 openstackgerrit Arthur Dayne proposed openstack/nova master: libvirt:driver:Disallow AIO=native when 'O_DIRECT' is not available https://review.opendev.org/682772
13:18:01 bauzas lyarwood: yup, I knew for the flipped ordering
13:18:18 bauzas lyarwood: but again, this saves a guest OS
13:18:40 bauzas maybe I'm bikeshedding on some naming this being said
13:19:10 sean-k-mooney lyarwood: for what its worth on a normal machine if you boot form usb the disk order used to change as it would become sda :P
13:19:40 openstackgerrit Stephen Finucane proposed openstack/nova master: api: Add support for new cyborg extra specs https://review.opendev.org/716222
13:19:48 openstackgerrit Stephen Finucane proposed openstack/nova master: Drop concept of '?validation' parameter https://review.opendev.org/717789
13:20:01 sean-k-mooney at least on linux on window its proably different
13:20:04 lyarwood sean-k-mooney: I've never seen that tbh, was that under a particular BIOS?
13:20:04 openstackgerrit Stephen Finucane proposed openstack/nova master: api: Add microversion for extra spec validation https://review.opendev.org/708436
13:20:16 openstackgerrit Stephen Finucane proposed openstack/nova master: docs: Add documentation for flavor extra specs https://review.opendev.org/710037
13:20:31 sean-k-mooney lyarwood: its what used to happen on the servers i used at intel
13:20:38 lyarwood sean-k-mooney: weird
13:20:51 lyarwood sean-k-mooney: I guess that's one way to change the boot order
13:20:53 sean-k-mooney i have seen it about 50% of the other laptops and servers have at home
13:21:33 lyarwood bauzas: no issues, my use of unstable is just confusing there so I'll remove it.
13:21:49 openstackgerrit Stephen Finucane proposed openstack/nova master: Follow-up for flavor-extra-spec-validators series https://review.opendev.org/718357
13:21:52 sean-k-mooney lyarwood: i dont think it hapens for uefi for what its worth
13:21:59 sean-k-mooney only legacy boot mode
13:22:03 bauzas lyarwood: thanks for your good comments on https://review.opendev.org/#/c/715490/ I'll upload a FUP
13:22:13 bauzas lyarwood: if you don't disagree
13:22:30 stephenfin gibi: Respun that. Good spot. /o\
13:22:33 bauzas (I just want to avoid a new zuul CI respin)
13:22:57 lyarwood bauzas: yeah np
13:23:04 bauzas ack thanks
13:24:05 sean-k-mooney stephenfin: bauzas lyarwood since ye are all here care to review the final few patches for cyborg. the first two are real staight forward and dont require much knoladge of cyborg
13:24:23 bauzas sean-k-mooney: I can take a look today
13:24:30 sean-k-mooney https://review.opendev.org/#/c/716185/6 https://review.opendev.org/#/c/716186/6 and https://review.opendev.org/#/c/715326/7
13:24:32 lyarwood sean-k-mooney: yup I can
13:24:55 sean-k-mooney i think sundar is nolonger working on this stuff so if there are any issue ill respin the patches
13:25:17 bauzas I'm mostly done on reviewing lyarwood's stable rescue thingies, so either I go with pinned/non-pinned stephenfin's patches or cyborg ones
13:25:32 bauzas but I think cyborg is moar important upstream
13:25:50 bauzas stephenfin: amirite ?
13:26:19 sean-k-mooney they are shorter. the first one is the release note/api docs the second is a minor fix and the last is rebuild/evacuate

Earlier   Later