Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2023-12-13
10:29:18 bauzas elodilles: sorry, just saw your ping
10:29:23 bauzas elodilles: nope, wasn't aware of it
10:38:10 elodilles bauzas: thanks. well, i'm not even sure it's still there in qemu, and i haven't seen sign of it in libvirt either, though i thought it worth a question o:)
12:50:09 opendevreview Stephen Finucane proposed openstack/nova master: docs: Address nits in new security group doc https://review.opendev.org/c/openstack/nova/+/903590
13:38:56 gibi sean-k-mooney: an easy power management bugfix backport that has elodilles' +2 https://review.opendev.org/c/openstack/nova/+/903151
13:47:56 sean-k-mooney "raise exception.InvalidConfiguration(msg) from e" ah your using that to convert the excpetion cool
13:49:48 sean-k-mooney +2w
14:05:10 artom Ooo, `raise from` is a thing?
14:05:22 sean-k-mooney its a python 3 only thing
14:05:23 artom So I could import zombie; raise zombie from the_dead? :D
14:05:24 sean-k-mooney but yes
14:05:36 sean-k-mooney ...
14:05:56 sean-k-mooney your a terrible terrible person. a good person but a teribel one :P
14:06:11 artom :D
14:09:21 gibi :D
14:10:01 gibi artom: you need to declare the_dead before
14:10:17 gibi :)
14:15:57 sean-k-mooney artom: +2 on https://review.opendev.org/c/openstack/nova/+/899776 but add a release note please if you respin
14:18:06 sean-k-mooney gibi: if your review then ^ is small
14:18:21 artom Yeah, agree on the release note
14:19:30 sean-k-mooney its not vital but it woudl be good to advertise this. if you want to respin anyway jsut ping me and ill take a look again
14:23:20 opendevreview Artom Lifshitz proposed openstack/nova master: libvirt: Stop unconditionally enabling evmcs https://review.opendev.org/c/openstack/nova/+/899776
14:23:25 artom Done ^^ :)
14:30:54 sean-k-mooney ack +2'd
14:47:00 kashyap elodilles: Hi, I was just pointed to your question about https://patchwork.kernel.org/project/qemu-devel/patch/20180622003140.164613-2-mst@redhat.com/
14:47:11 kashyap elodilles: It's a QEMU command-line option, as you see. We don't have a BP (not that I know of) for it
14:48:02 kashyap Have you tested it w/ "<qemu:commandline>
14:48:43 kashyap ... in the guest XML? Do you see noticeable benefits?
14:49:09 sean-k-mooney we do not allow <qemu:commandline> to be used in nova
14:49:29 sean-k-mooney so even if that worked it would need to be expose via an actual libvirt element
14:49:46 sean-k-mooney its fine for testing hwoever
14:50:11 gibi artom, sean-k-mooney +2+W
14:53:51 sean-k-mooney elodilles: so i had heard about the fucntionaltiy but not looked into it much
14:54:12 sean-k-mooney elodilles: i could see use in the future enableing this by default for guest with pinned but not mixed cpus
14:54:30 sean-k-mooney since this does not seam to be per core we cant supprot mixed cpus
15:00:40 elodilles kashyap: i'm not aware of any testing, but yeah, that's probably would be a good start to see if it has any benefit.
15:01:10 kashyap sean-k-mooney: I of course know we don't allow. I explicitly said "testing" there
15:01:38 kashyap The qemu:commandline thing. It is alwys mentioned in context of developer testing in this channel for a million years now.
15:01:54 kashyap elodilles: Yeah; some data on that would be good. What made you think of that feature?
15:01:55 elodilles sean-k-mooney: ACK, yepp, only for guests with pinned CPUs, yepp. that was my understanding too.
15:02:34 sean-k-mooney elodilles: if you do try this in devstack just not i disabled the apic by default in devstack becasue of a bug in cirros 5.x
15:02:35 elodilles kashyap: well, i got a question from downstream team :)
15:02:45 kashyap Ah, makes sense :)
15:02:50 sean-k-mooney that bug nolonger exists in 6.x but i have not disabeld that workaround
15:03:01 sean-k-mooney so you might need to makesure it off in your env if you try it
15:03:19 sean-k-mooney off meaing the workaroudn is disabeld and the apic is present in the guest
15:03:50 sean-k-mooney https://docs.openstack.org/nova/latest/configuration/config.html#workarounds.libvirt_disable_apic
15:04:46 sean-k-mooney oh actully ignore that
15:05:01 sean-k-mooney its acpi that is used for power not apic
15:07:38 elodilles :)
15:08:41 kashyap Yeah, confusing acronyms
16:22:23 opendevreview Merged openstack/nova stable/2023.2: [pwmgmt]ignore missin governor when cpu_state used https://review.opendev.org/c/openstack/nova/+/903151
16:52:03 opendevreview Balazs Gibizer proposed openstack/nova stable/2023.1: [pwmgmt]ignore missin governor when cpu_state used https://review.opendev.org/c/openstack/nova/+/903553
16:53:06 gibi elodilles, sean-k-mooney: thanks for the review on the backport of the power management fix here is the Antelope backport ^^
16:53:11 gibi still clean cherry-pick
17:00:40 elodilles gibi: will look into it
17:34:11 gibi thanks!
18:01:39 sean-k-mooney looks liek bobcat one already merged. that was fastere then i expected which is nice
18:02:43 sean-k-mooney the feature was added in antelope
18:02:43 sean-k-mooney i rememeber correctly this shoudl be the final backport correct
18:02:43 sean-k-mooney i agree its a clean backport and the commit ectra is all correct so +2w
18:03:10 sean-k-mooney so once that merges the issue will be closed across all affected branches
18:32:20 opendevreview Artom Lifshitz proposed openstack/nova master: POC: glanceclient streaming uploads https://review.opendev.org/c/openstack/nova/+/903611
18:32:54 sean-k-mooney that sound like a new feature
18:33:31 sean-k-mooney hum just enabing an exisitng glance one. i wonder if that needs to be a config option however
18:34:05 sean-k-mooney i think streaming upload/download may have some depenices on how glance or the webserver/loadbalance its deployed behind are configured
18:34:48 sean-k-mooney it was added by https://opendev.org/openstack/glance/commit/7f761e16ce69ec549be1a66bd2797e53d0a1d988?style=unified&whitespace=show-all&show-outdated=
18:36:46 sean-k-mooney application/octet-stream that shoudl work
18:37:05 sean-k-mooney its the same content type as a normal upload
19:53:56 opendevreview Stephen Finucane proposed openstack/nova master: docs: Address nits in new security group doc https://review.opendev.org/c/openstack/nova/+/903590
19:59:15 opendevreview Artom Lifshitz proposed openstack/nova master: POC: glanceclient streaming uploads https://review.opendev.org/c/openstack/nova/+/903611
20:00:42 opendevreview Merged openstack/nova master: Use SDK for remaining ironic driver calls https://review.opendev.org/c/openstack/nova/+/867795
20:03:06 artom sean-k-mooney, so chunked is apparently not the same as streaming
20:04:08 artom Though if I'm honest I'm not sure I understand the difference on the wire. The mechanics appear to be - pass a generator to Python requests = chunked, pass a file handler = streaming
20:04:37 artom Maybe I should check what Nova currently passes... I kinda assumed it just yeeted the entire image in the POST body
20:05:07 artom Hrmp, no mention of 'chunked' anywhere in our git log
20:06:52 opendevreview Jay Faulkner proposed openstack/nova master: [ironic] Partition & use cache for list_instance* https://review.opendev.org/c/openstack/nova/+/900831
20:32:17 opendevreview Merged openstack/nova master: docs: Revamp the security groups guide https://review.opendev.org/c/openstack/nova/+/903507
20:37:01 opendevreview Merged openstack/nova master: docs: Address nits in new security group doc https://review.opendev.org/c/openstack/nova/+/903590
20:46:36 JayF gibi: When you get a chance, https://review.opendev.org/c/openstack/nova/+/867797/ and the children below it are almost ready to land in that big Ironic<>Nova openstacksdk migration chain; I've been trying to usher these through the gate so a review and approval there (and on the other patches in the stack, as appropriate) would be appreciated. Thank you!
23:46:10 opendevreview Merged openstack/nova master: [codespell] fix typos in tests https://review.opendev.org/c/openstack/nova/+/897213
#openstack-nova - 2023-12-14
00:07:27 opendevreview Merged openstack/nova master: Use SDK for deprecated baremetal proxy API https://review.opendev.org/c/openstack/nova/+/867796
01:01:36 opendevreview Takashi Kajinami proposed openstack/nova master: Remove deprecated [api] use_forwarded_for https://review.opendev.org/c/openstack/nova/+/903339
02:31:05 opendevreview Nobuhiro MIKI proposed openstack/nova master: scheduler: Request COMPUTE_ADDRESS_SPACE_* traits https://review.opendev.org/c/openstack/nova/+/895761
03:31:55 opendevreview OpenStack Proposal Bot proposed openstack/nova master: Imported Translations from Zanata https://review.opendev.org/c/openstack/nova/+/903427
05:41:14 opendevreview Takashi Kajinami proposed openstack/nova master: Remove deprecated [api] use_forwarded_for https://review.opendev.org/c/openstack/nova/+/903339
08:11:44 gibi JayF: Done
08:22:57 bauzas gibi: sean-k-mooney: it would be nice if you could review my bugfix series https://review.opendev.org/q/topic:%22bug/2041519%22
10:10:07 bauzas it would also be nice if cores could look at my mdev live-migration spec https://review.opendev.org/c/openstack/nova-specs/+/900636
10:17:29 bauzas wow
10:37:19 noonedeadpunk hey there! I have quite "trivial" question about limit_tenants_to_placement_aggregate / placement_aggregate_required_for_tenants / AggregateMultiTenancyIsolation things just to confirm I understand that correctly.
10:38:34 noonedeadpunk So basically "user story" is to make set of computes be available only for specific tenant, meaning that it should be able to spawn instances _only_ on that aggregate, while all other tenants should be able to spawn anywhere _except_ this aggregate
10:38:54 bauzas noonedeadpunk: shit, I need to leave in a sec
10:39:01 noonedeadpunk Am I right that I should then use limit_tenants_to_placement_aggregate: true, placement_aggregate_required_for_tenants: false and ignore AggregateMultiTenancyIsolation
10:39:25 bauzas there are multiple ways to support tenancy limitations
10:39:30 noonedeadpunk Or, use AggregateMultiTenancyIsolation but then add another property and set of flavors for this tenant?
10:39:38 bauzas either by the filter or by our pre-filters
10:41:02 bauzas noonedeadpunk: there is another change from sahid related to this https://review.opendev.org/c/openstack/nova/+/896512
10:42:00 noonedeadpunk I need just 1 tenant to be isolated, so it shouldn;t be an issue...
10:42:03 noonedeadpunk sure...

Earlier   Later