| opendevreview | Merged openstack/nova stable/2024.1: repoduce post liberty pre vicoria instance numa db issue https://review.opendev.org/c/openstack/nova/+/932064 | 00:38 |
|---|---|---|
| opendevreview | Amit Uniyal proposed openstack/nova master: Reproducer for cold migration on shared storage https://review.opendev.org/c/openstack/nova/+/940304 | 02:26 |
| opendevreview | Merged openstack/os-vif master: Update gate jobs as per the 2025.1 cycle testing runtime https://review.opendev.org/c/openstack/os-vif/+/935580 | 03:03 |
| gmann | mikal: approved the spice tests tempest change but melwitt reminded its depends-on nova change. if you can remove that depends-on I can quickly re-approve it https://review.opendev.org/c/openstack/tempest/+/942023 | 03:05 |
| opendevreview | Amit Uniyal proposed openstack/nova master: Reproducer for cold migration on shared storage https://review.opendev.org/c/openstack/nova/+/940304 | 04:24 |
| opendevreview | Tatsuya Hayashino proposed openstack/nova master: Allow update metadata and tag for error VM https://review.opendev.org/c/openstack/nova/+/942167 | 05:19 |
| mikal | gmann: sure, I can do that now | 09:22 |
| mikal | gmann: done | 09:25 |
| opendevreview | Rajesh Tailor proposed openstack/nova master: Handle InstanceExists exception for duplicate instance https://review.opendev.org/c/openstack/nova/+/860938 | 10:25 |
| sshmlbrg | Hi! I have a question about the 'attach volume to shelved instance' functionality. How does it work if an instance in the 'shelved_offloaded' status has no availability zone, but when attaching, the availability zones of the instance and the volume are checked? The documentation states that this functionality works starting from microversion 2.20. | 10:32 |
| opendevreview | Rajesh Tailor proposed openstack/nova master: Fix KeyError on assisted snapshot call https://review.opendev.org/c/openstack/nova/+/900783 | 10:37 |
| opendevreview | Rajesh Tailor proposed openstack/nova master: Fix case sensitive comparison https://review.opendev.org/c/openstack/nova/+/933107 | 11:21 |
| opendevreview | ribaudr proposed openstack/nova master: Add managed flag to PCI device specification https://review.opendev.org/c/openstack/nova/+/937649 | 13:20 |
| opendevreview | ribaudr proposed openstack/nova master: Update driver to deal with managed flag https://review.opendev.org/c/openstack/nova/+/938405 | 13:20 |
| opendevreview | ribaudr proposed openstack/nova master: Update driver to deal with managed flag https://review.opendev.org/c/openstack/nova/+/938405 | 13:39 |
| MengyangZhang[m] | <MengyangZhang[m]> "Do we know what would be the..." <- sean-k-mooney: Do you have an idea of how to implement this? | 14:36 |
| sean-k-mooney | i dont have context for waht your asking | 14:37 |
| opendevreview | Rajesh Tailor proposed openstack/nova master: Add support for showing finish_time https://review.opendev.org/c/openstack/nova/+/928933 | 14:42 |
| dansmith | bauzas: I'm here | 14:42 |
| bauzas | cool | 14:42 |
| frickler | MengyangZhang[m]: doing quote replies in the matrix doesn't transfer well into IRC, best avoid those | 14:43 |
| bauzas | so indeed the lazyloading is something like a good performance concern | 14:43 |
| bauzas | when we discussed with sean-k-mooney, we said eventually "yeah, but just for the moment we wouldn't working on it" | 14:43 |
| bauzas | that said, now, if you want, I can provide another dependent change for this | 14:44 |
| dansmith | bauzas: we will lazy-load every instance on every host we're considering...sequentially | 14:44 |
| dansmith | for ten hosts with 100 instances each, that's 1000 trips to the database for each schedule operation | 14:44 |
| bauzas | indeed you see this in https://4d2664479333d0a7d727-a4adcb41e06ec456a00383225f090da6.ssl.cf5.rackcdn.com/940642/9/check/nova-multi-cell/bbf5fea/controller/logs/screen-n-sch.txt | 14:45 |
| bauzas | Feb 17 19:20:02.725500 np0039881226 nova-scheduler[84264]: DEBUG nova.objects.instance [None req-6e249595-d7b0-4cbc-95d0-e2798d45bc99 tempest-ImagesOneServerNegativeTestJSON-728670221 tempest-ImagesOneServerNegativeTestJSON-728670221-project-member] Lazy-loading 'system_metadata' on Instance uuid e6f93654-3b12-429d-8f17-15ad1e605964 {{(pid=84264) obj_load_attr | 14:45 |
| bauzas | /opt/stack/nova/nova/objects/instance.py:1141}} | 14:45 |
| dansmith | yeah | 14:45 |
| bauzas | so I don't disagree with you, if you think it's a priority issue, I can try to fix it then | 14:45 |
| bauzas | or you if you want | 14:46 |
| bauzas | that would be a bottom change for the series | 14:46 |
| dansmith | I really don't think it's reasonable to do this in a filter, and definitely not one that is always enabled by default and/or would have this behavior when it's not even weighted | 14:46 |
| dansmith | can we gmeet about this for a minute? I want to make sure we're also considering the track-changes path properly | 14:47 |
| bauzas | yeah cool with me | 14:47 |
| bauzas | dansmith: yeah, sure | 14:47 |
| * bauzas just gets a coffee | 14:47 | |
| dansmith | bauzas: when you're ready: meet.google.com/aug-ibty-wrz | 14:47 |
| dansmith | sean-k-mooney: if you're interested ^ | 14:48 |
| sean-k-mooney | am sure | 14:49 |
| sean-k-mooney | be right there | 14:49 |
| sean-k-mooney | give me a sec i need to sign in to sso... | 14:50 |
| sean-k-mooney | bauzas: so looks like melwitt gave me feedback on the release not so ill proably respine for that https://review.opendev.org/c/openstack/nova/+/940835/7 if you have any other comments then i can adress them at the same time | 15:17 |
| sean-k-mooney | im goign to squash the two commit and reuplload in a few minutes | 15:19 |
| sean-k-mooney | if people are around later i want to merge https://review.opendev.org/c/openstack/os-vif/+/941585 and https://review.opendev.org/c/openstack/os-vif/+/941568 today if we can too so we can do a release of os-vif before the non clinet lib freeze tomorrow | 15:22 |
| sean-k-mooney | ill work on the nova deprecation patch later today as well but that is less urgent | 15:23 |
| sean-k-mooney | as a psa firday is a company day at redhat so many of us wont be around | 15:24 |
| bauzas | sean-k-mooney: gibi: you probably lost context about https://review.opendev.org/c/openstack/nova/+/845757/8 but I now rebased it and I think it would help many users, particularly for evacuation ccases | 15:29 |
| bauzas | sean-k-mooney: I can dig into the os-vif patches | 15:29 |
| sean-k-mooney | i cna review that later. i still have some of the context | 15:29 |
| bauzas | sean-k-mooney: I wonder why I did split the change in two patches, one for the doc change and the other one for the code change | 15:30 |
| bauzas | maybe I should squash that, particularly you had concerns about the doc | 15:31 |
| opendevreview | Dan Smith proposed openstack/nova master: Add fill_metadata() to InstanceList https://review.opendev.org/c/openstack/nova/+/942214 | 15:47 |
| dansmith | sean-k-mooney: bauzas: ^ | 15:49 |
| bauzas | thanks | 15:50 |
| opendevreview | Merged openstack/nova master: Respect supplied arguments in novncproxy_base_url https://review.opendev.org/c/openstack/nova/+/928839 | 16:00 |
| opendevreview | ribaudr proposed openstack/nova master: Augment the LiveMigrateData object https://review.opendev.org/c/openstack/nova/+/942143 | 17:05 |
| opendevreview | ribaudr proposed openstack/nova master: Add live_migratable flag to PCI device specification https://review.opendev.org/c/openstack/nova/+/942144 | 17:05 |
| opendevreview | ribaudr proposed openstack/nova master: Update manager to allow vfio pci device live migration https://review.opendev.org/c/openstack/nova/+/942145 | 17:05 |
| opendevreview | ribaudr proposed openstack/nova master: Update conductor and filters allowing migration with SR-IOV devices https://review.opendev.org/c/openstack/nova/+/942146 | 17:05 |
| opendevreview | ribaudr proposed openstack/nova master: WIP: Update driver to map the targeted address for SR-IOV PCI devices https://review.opendev.org/c/openstack/nova/+/942147 | 17:05 |
| ratailor | gibi, I already submitted changes for sdk and client here https://review.opendev.org/q/topic:%22bp/image-props-in-server-show%22 but the tests are failing, because not sure how to add 'image_properties' attr to class Server in openstacksdk with 2.98 version specific. | 17:06 |
| ratailor | gibi, could you please check and provide your feedback, when you have time ? | 17:07 |
| opendevreview | sean mooney proposed openstack/nova master: move nova-ovs-hybrid-plug to deploy with spice and fix qxl default https://review.opendev.org/c/openstack/nova/+/940873 | 17:09 |
| sean-k-mooney | melwitt: bauzas ^ squased into one comment, with release note updated. this is the review that was used for depends on before and i have abandoned the first patch | 17:11 |
| sean-k-mooney | ok so next on my list is dans patch | 17:11 |
| melwitt | sean-k-mooney: cool, thanks. will look | 17:12 |
| sean-k-mooney | dansmith: bauzas im good with https://review.opendev.org/c/openstack/nova/+/942214 | 17:40 |
| sean-k-mooney | just waiting on zuul beofre +w | 17:41 |
| sean-k-mooney | with that said | 17:41 |
| sean-k-mooney | it would be nice to see this used in https://review.opendev.org/c/openstack/nova/+/940642 | 17:41 |
| sean-k-mooney | before we merge | 17:41 |
| sean-k-mooney | so perhaps we shoud wait to merge until the other patches are on top | 17:42 |
| dansmith | yep, fine to wait until it's used | 17:42 |
| bauzas | sure, I'm working on the rebase | 17:42 |
| bauzas | and then I'll need to end for the day | 17:42 |
| gmann | In case anyone interested, TC is short of one candidate for this election. nomination ending in ~6 hrs from now https://governance.openstack.org/election/ | 17:56 |
| dansmith | gmann: a non-redhat person is needed I assume :) | 18:00 |
| gmann | dansmith: I think we still have one member capacity for RH member. slaweq is not re-running and no new RH member nominated yet | 18:02 |
| MengyangZhang[m] | sean-k-mooney: I may need some help with implementing this min version check you mentioned for this patch https://review.opendev.org/c/openstack/nova-specs/+/932653. I have two ideas to do this:... (full message at <https://matrix.org/oftc/media/v1/media/download/AViM5KuzefB-19EfrDbw71k64XnKJJpP_G9sobaeS8NH-AJhP8DfPwD665n9DmazUuyKEk2lk1KiimNHIqjb0xpCeVZ31wDgAG1hdHJpeC5vcmcvcUdQVElMSnlsTUNLTWFJUVFDaFpmdERR>) | 18:07 |
| sean-k-mooney | MengyangZhang[m]: oh so you cant really do that properly. my expection was cidner qos would never change out side of spawn/attach or perhaps volume retype | 18:09 |
| sean-k-mooney | we coudl do it for spawn and attach in https://github.com/openstack/nova/blob/master/nova/compute/api.py | 18:09 |
| sean-k-mooney | but cidner will end up bypassing any proection we put in place... | 18:09 |
| sean-k-mooney | so im not sure what to do about that | 18:10 |
| sshmlbrg | Hi? | 18:10 |
| sean-k-mooney | sshmlbrg: o/ | 18:11 |
| sshmlbrg | I have a question about the 'attach volume to shelved instance' functionality. How does it work if an instance in the 'shelved_offloaded' status has no availability zone, but when attaching, the availability zones of the instance and the volume are checked? The documentation states that this functionality works starting from microversion 2.20. | 18:11 |
| sean-k-mooney | sshmlbrg: i dont think we supprot attach ro deation for shleve_offloaded but i would have to check | 18:12 |
| sshmlbrg | ok, thanks | 18:13 |
| sean-k-mooney | maybe we do i might be confusing that with snapshosts | 18:13 |
| sshmlbrg | https://docs.openstack.org/api-ref/compute/?expanded=attach-a-volume-to-an-instance-detail#attach-a-volume-to-an-instance | 18:14 |
| sean-k-mooney | ok so yes we do suprot it basedon the api refe as you said form 2.20 | 18:14 |
| sean-k-mooney | you were asking about aviablity zone checks | 18:15 |
| sshmlbrg | but it doesn't work | 18:15 |
| sean-k-mooney | so by default htere are no aviablity zone checks for nova and cinder voluems. | 18:15 |
| sean-k-mooney | we have a config option to opt into the non default beahvior of preventign cross az attach | 18:15 |
| sean-k-mooney | its possible this only works when cross az atach is allow or when an instance is pinned to an az | 18:16 |
| sean-k-mooney | ill see if i can quickly test it | 18:19 |
| sshmlbrg | But this function lacks a cross-AZ check, and we proceed to compare the availability zone of the volume and the instance. https://github.com/openstack/nova/blob/master/nova/compute/api.py#L5143 | 18:20 |
| sean-k-mooney | so cross_az attach is a comput enode option | 18:22 |
| sean-k-mooney | a shevle offloaded instance does not have a compute service | 18:22 |
| sean-k-mooney | so you cant actuly reason about if this shoudl or shoudl not work | 18:22 |
| sean-k-mooney | the az check i enabeld by this check https://github.com/openstack/nova/blob/master/nova/compute/api.py#L5065 | 18:23 |
| sean-k-mooney | *this default | 18:23 |
| sean-k-mooney | because we are not passsing that variable | 18:24 |
| opendevreview | Sylvain Bauza proposed openstack/nova master: Add a new ImagePropertiesWeigher https://review.opendev.org/c/openstack/nova/+/940642 | 18:24 |
| opendevreview | Sylvain Bauza proposed openstack/nova master: per-property ImageMetaPropsWeigher https://review.opendev.org/c/openstack/nova/+/941601 | 18:24 |
| bauzas | dansmith: sean-k-mooney: rebased as asked | 18:24 |
| * bauzas needs to leave | 18:24 | |
| sshmlbrg | Then, should we disable this option so that attach works for shelved_offloaded VMs? | 18:25 |
| dansmith | bauzas: okay did you want me to make the mods to use the new method? | 18:25 |
| sean-k-mooney | there is more context here | 18:26 |
| sean-k-mooney | sshmlbrg: https://github.com/openstack/nova/commit/07a24dcef7ce6767b4b5bab0c8d3166cbe5b39c0 | 18:26 |
| sshmlbrg | i need a few minutes to get it :) | 18:27 |
| sean-k-mooney | sshmlbrg: it looks like the curent check was added to fix https://bugs.launchpad.net/nova/+bug/1694844 | 18:28 |
| sean-k-mooney | but it may have broken seveled atatch in the process | 18:28 |
| sean-k-mooney | skimming the patch it maybe looks like the intent was to fall back to the default schduler zone | 18:29 |
| sean-k-mooney | i think in this sepcific case it woudl be valide to pass false | 18:30 |
| sean-k-mooney | if an donly if instnace.az was None in either the request spec or the instance table in teh cell db | 18:31 |
| sean-k-mooney | basically if the isntance is shelve_offloaded. and it not pinend to any az | 18:31 |
| sean-k-mooney | then we can possibly disable the afinity check | 18:32 |
| sean-k-mooney | but that may lead to issue unshevling | 18:32 |
| sean-k-mooney | unless we pinned the shleved instance to the relevent az | 18:32 |
| sean-k-mooney | that is only vailid to do however if cros_az_attach=false | 18:33 |
| sean-k-mooney | since we can now unshlve to an az | 18:34 |
| sshmlbrg | Yes, and also, we can unshelve the VM to another AZ, and I’m not sure how the volumes will behave in that case. | 18:34 |
| sean-k-mooney | the workaround for tha twould be to return an error and make the user explictly request the az they want to unshvle to and that woudl have to match the volume az | 18:34 |
| sean-k-mooney | sshmlbrg: its blocked if you use the non default beahvior | 18:35 |
| sshmlbrg | ok | 18:35 |
| sean-k-mooney | sshmlbrg: teh reals fix is to delete the ablity for nova to prevent cross_az entirly | 18:35 |
| sean-k-mooney | that never shoudl have been a per compute config option | 18:35 |
| sean-k-mooney | it should be a attibute on the cinder volume | 18:36 |
| sean-k-mooney | i.e. when attaching a volume you shoudl be able to choose if cross az storage is allowed or not | 18:36 |
| sean-k-mooney | the reason it snot that way, beyond legacy | 18:36 |
| sean-k-mooney | is as a end user you dont knwo if the strage backend is accable from the current host | 18:36 |
| sean-k-mooney | its tricky | 18:37 |
| sshmlbrg | ok, i hope i got it, need to summarize it. Is it bug or ...? | 18:38 |
| sean-k-mooney | its proably a but im just not sure how to fix it | 18:49 |
| sean-k-mooney | im inclided to say pass false and then requrie pepoel to unshlve to the same az fi thats how the clodu is confiugred | 18:50 |
| sean-k-mooney | we shoudl have a better solution to this, that the problem with config driven api bevhior and why we dont add it anymore if we can avoid it at all costs | 18:51 |
| sshmlbrg | I was thinking about storing the old AZ, writing the AZ to it when shelving, and releasing it when unshelving. Then, when attaching the volume, we could check that specific availability zone. | 18:53 |
| sean-k-mooney | so that not correct | 18:54 |
| sean-k-mooney | instnace that dont request a az shoudl not be pinned to one by default | 18:54 |
| sshmlbrg | understand why | 18:54 |
| sean-k-mooney | if they have a cidner volume and land on a host with cross_az_attach=false and it does not violate the requirement nova will pin it | 18:55 |
| sean-k-mooney | sshmlbrg: did the instance have a voluem before it was shelved? | 18:56 |
| sshmlbrg | yes, volume-backed instance | 18:56 |
| sean-k-mooney | ok then it should be pinned already | 18:57 |
| sean-k-mooney | so the bug is that its proably not usign the az form the request spec if its not set in the host | 18:57 |
| sean-k-mooney | * in the instance table | 18:57 |
| sshmlbrg | yes | 18:58 |
| sean-k-mooney | ok so tha that specific edge case is fixable | 18:58 |
| sean-k-mooney | we proably cant fix the case where it did not have a volume before it was shelved eaisly | 18:59 |
| sean-k-mooney | but we coudl fix this case. | 18:59 |
| sean-k-mooney | can you write that up as a bug | 18:59 |
| sean-k-mooney | sshmlbrg: im not sure if its somethign you feel comfortable trying to fix | 18:59 |
| sshmlbrg | get availability_zone from RequestSpec for instance? | 19:00 |
| sean-k-mooney | if its shelve offloaded | 19:01 |
| sshmlbrg | I would like to fix it under your guidance :) I will create a bug tomorrow. | 19:01 |
| sean-k-mooney | and if its not pinend then just proceed as if it passed i think | 19:01 |
| sshmlbrg | sean-k-mooney: Thanks a lot for your attention. Stay in touch | 19:03 |
| sean-k-mooney | sshmlbrg: no worreis, https://github.com/openstack/nova/blob/88a36a5a0099aa52f9ea93905c7c56911b2bf0b3/nova/objects/request_spec.py#L781-L791 is your starting point | 19:04 |
| sean-k-mooney | you can use Requestspec.get_by_instance_uuid() to get the request spec object if its shelve offloaded | 19:05 |
| sean-k-mooney | and then grab the az https://github.com/openstack/nova/blob/88a36a5a0099aa52f9ea93905c7c56911b2bf0b3/nova/objects/request_spec.py#L76 | 19:05 |
| sean-k-mooney | if that is not none then you need to compare to the volume | 19:05 |
| sean-k-mooney | if it is then i think returnign form the function might be ok | 19:06 |
| sean-k-mooney | we will need to see what others think | 19:06 |
| mikal | Morning | 19:11 |
| sean-k-mooney | mikal: o/ | 19:12 |
| opendevreview | Mengyang Zhang proposed openstack/nova-specs master: Add Burst Length Support to Cinder QoS https://review.opendev.org/c/openstack/nova-specs/+/932653 | 19:30 |
| opendevreview | sean mooney proposed openstack/nova master: Deprecate linux bridge support https://review.opendev.org/c/openstack/nova/+/942291 | 20:58 |
| sean-k-mooney | bauzas: ^ thats the nova side fo the deprecation | 21:00 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!