Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2025-10-21
09:15:43 elodilles np! thanks too o/
09:19:57 jlejeune and if you have time, I would really appreciate that my topics (bug-2044235 and bug/2085135) moved forward until 2024.1 before that release is set to 'unmaintained' :)
09:32:37 elodilles jlejeune: i'm planning to review all stable patches that are ready to merge, but on the other hand we are very late already with the final stable release of 2024.1 Caracal, so I don't think we have time to merge there any more patches :/
09:33:09 elodilles jlejeune: note that bug fixes can land to unmaintained/2024.1 after the transition
09:34:05 elodilles jlejeune: though i understand that it is less comfortable for consumers as the releases have to be cut downstream by themselves from unmaintained/2024.1 branches
09:40:02 jlejeune ok understandable, so I would need to repush my change on new unmaintained/2024.1 branch later and abandon my changes over stable/2024.1 branches
09:58:15 elodilles jlejeune: yes, that will be the way forward
10:05:11 jlejeune ok got it
10:39:46 opendevreview Merged openstack/nova stable/2025.1: Adds regression test for bug LP#2044235 https://review.opendev.org/c/openstack/nova/+/961601
10:43:18 gibi dansmith: could you add back your +2 on https://review.opendev.org/c/openstack/nova/+/937275 ? The devstack dependency landed finally so we can good to go and test the placement optimization stuff in nova-next permanently
10:43:32 gibi s/can/are/
10:44:27 gibi and if you are around that then there is a similar patch that enables such optimizations in some of the nova func tests https://review.opendev.org/c/openstack/nova/+/963036
10:44:53 gibi cores, there is a super easy doc patch that needs a second +2 https://review.opendev.org/c/openstack/nova/+/963281
10:54:18 sean-k-mooney gibi: o/ still gong thorugh my email but ill take a look in a bit, i think i looked at the docs patch before while traveling but didn leave a review
11:16:34 opendevreview Merged openstack/nova stable/2025.1: nova-conductor puts instance in error state https://review.opendev.org/c/openstack/nova/+/961774
11:17:51 opendevreview Nicolai Ruckel proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.opendev.org/c/openstack/nova/+/959682
11:20:56 nicolairuckel sean-k-mooney: I changed my patch according to your comments. The patch is now a lot smaller but it still only works if I set the permissions manually. I was hoping that just letting libvirt do its thing would solve that. :/
11:22:28 opendevreview Balazs Gibizer proposed openstack/nova master: Rally job for eventlet-removal https://review.opendev.org/c/openstack/nova/+/960130
11:27:31 opendevreview Balazs Gibizer proposed openstack/nova master: Rally job for eventlet-removal https://review.opendev.org/c/openstack/nova/+/960130
11:32:05 opendevreview Balazs Gibizer proposed openstack/nova master: Rally job for eventlet-removal https://review.opendev.org/c/openstack/nova/+/960130
11:37:34 sean-k-mooney nicolairuckel: thats odd. i guess we can try and replicate the behvior and see if we can avoid this but. did you test this with a new vm? is it possibel that you previous attempet left the files in a non standard state?
11:43:11 sean-k-mooney gibi: by the way the numa in placemetn talk was very good. they also appremntly foudn hand fixed the placment scaling issue in there work
11:43:13 nicolairuckel Ah, that could be the problem.
11:43:20 nicolairuckel I didn't create a new VM.
11:55:16 opendevreview Ahmad Hassan proposed openstack/nova-specs master: Offline VM migration using https https://review.opendev.org/c/openstack/nova-specs/+/964444
11:58:41 opendevreview Merged openstack/nova master: [doc]PCI in Placement tuning https://review.opendev.org/c/openstack/nova/+/963281
12:11:46 opendevreview Julien LE JEUNE proposed openstack/nova stable/2024.2: nova-conductor puts instance in error state https://review.opendev.org/c/openstack/nova/+/961775
12:16:56 opendevreview Julien LE JEUNE proposed openstack/nova stable/2024.2: Update start_service() function in test https://review.opendev.org/c/openstack/nova/+/963848
12:16:57 opendevreview Julien LE JEUNE proposed openstack/nova stable/2024.2: Adds regression test for bug LP#2085135 https://review.opendev.org/c/openstack/nova/+/963849
12:16:58 opendevreview Julien LE JEUNE proposed openstack/nova stable/2024.2: Reset the mapped field of nodes at service deletion https://review.opendev.org/c/openstack/nova/+/963850
12:36:45 gibi sean-k-mooney: I looked at their slidset. What I decoded form there is that they found something similar but maybe not exactly the same as they mentioned same_subtree which was not part of our problem or our solution for the PCI case
12:36:59 gibi sean-k-mooney: I left some comments / question on the PTG etherpad related to this
12:37:24 sean-k-mooney gibi im not sure if its exactly the same, related or independent fo what you fixed
12:38:03 sean-k-mooney but yes it would be nice to see what there solution was and if its an addtional enhancement or not needed now
12:42:32 gibi I have a hunch that we could move the same_subtree check also inside the product() call if we want that might help pruning the searchspace in case same_subtree is used for their NUMA solution
12:42:55 gibi for PCI we don't need that but for NUMA RP with multiple children under it it is needed
12:43:06 sean-k-mooney yep maybe.
12:43:48 sean-k-mooney we can cross that bridge when we evaulate there solution
12:43:56 gibi yepp
12:43:58 sean-k-mooney they hope to upstream it
12:44:07 sean-k-mooney which would be cool if it work well
12:44:25 opendevreview Julien LE JEUNE proposed openstack/nova stable/2024.1: nova-conductor puts instance in error state https://review.opendev.org/c/openstack/nova/+/961778
12:44:26 sean-k-mooney they apprently have backported it to yoga internaly had have custoemr using it in production
12:44:46 sean-k-mooney so at least for there use cases it apprenlty funcitonal at the scale they are using it at
12:44:58 gibi yeah if it not spread or exaggerate the placement performance issue then lets get it, if it is then we need to think a bit more and assess the risk
12:45:49 sean-k-mooney i also think i agree with them about the ux issue
12:46:09 sean-k-mooney i.e. the workaround they had to do for member_of ectra related to the unamed group
12:46:34 opendevreview Julien LE JEUNE proposed openstack/nova stable/2024.1: Update start_service() function in test https://review.opendev.org/c/openstack/nova/+/963851
12:46:35 opendevreview Julien LE JEUNE proposed openstack/nova stable/2024.1: Adds regression test for bug LP#2085135 https://review.opendev.org/c/openstack/nova/+/963852
12:46:36 opendevreview Julien LE JEUNE proposed openstack/nova stable/2024.1: Reset the mapped field of nodes at service deletion https://review.opendev.org/c/openstack/nova/+/963853
12:46:37 sean-k-mooney i forget the explcit excption they said they got but i dont think they shoudl have got that
12:47:02 sean-k-mooney it would be good to get that into a placmenet bug report so we can assess it properly
12:47:36 sean-k-mooney with out going into the detail is sounded like a design oversight that could be condiered a bug
12:49:49 gibi sure, I'm happy to read incoming bugreports
12:50:06 gibi the slideset is low on details so most of I know are based on guesses and extrapolations
12:54:05 sean-k-mooney we invited them to the ptg. they have some pto this week so they proably wont have time to prepare much. bauzas added the topic on there behalf but we can try and connect with them on irc or the list seperatly as well
12:55:00 sean-k-mooney gibi: bauzas offered to help them with engaging with the upstream comunity so i assume if they submit a spec they will act as the feature liason or at least helping them if they have any issues with that
12:59:57 Uggla Hi sean-k-mooney, I hope you had a nice trip back to your home town. Can you point me to the patch about image encryption (lucks), we need to discuss next weak ?
13:00:37 sean-k-mooney am sure. the oen that is is currently kind fo blocked is on the glance side one sec
13:01:36 sean-k-mooney and ya the trip back was pretty ok. of couse when we got back to ireand it was raining and the car was parked in a 1 inch high puddel of water but other hten that it was fine
13:02:47 sean-k-mooney i more or less went to bed after getting back and slept until 2 am. i went back to sleep but got up again at liek 4 so my slep patther is off but not unexpected
13:02:57 sean-k-mooney Uggla: https://review.opendev.org/c/openstack/glance/+/926295
13:03:26 sean-k-mooney Uggla: how was your train? did you and sylvain travel back together?
13:03:38 Uggla do you think we need to discuss that at the glance/nova cross session ?
13:04:01 sean-k-mooney ya i think we should
13:04:32 sean-k-mooney the featuer is addign a new luks disk_format and we need to be able to supprot that in nova
13:05:58 sean-k-mooney i dont really see how we can supprt dans request ot bea able to valdiate that its a ` luks-encrypted GPT disk image` without breaking the convidental comptue usecases
13:06:22 sean-k-mooney but i get the concer as well
13:07:33 Uggla sean-k-mooney, it was ok a minor delay, but I managed to be at home around midnight(sunday). Sylvain took the train on Monday morning.
13:07:50 nicolairuckel sean-k-mooney, I have the same problem with a fresh VM.
13:09:34 bauzas sean-k-mooney: I already told the train story earlier in the chat. Long story short, never ever take the Lyon-Grenoble line if you want to visit the Alps
13:10:30 sean-k-mooney bauzas: hehe ill keep that in mind :)
14:12:47 elodilles bauzas Uggla : sorry, a bit late, but finally i've proposed the Nova stable release patches (flamingo-caracal). please review them. note that all of them contain some api related code changes, hence all bumps MINOR versions. please review them with extra care that they are OK to release (especially Caracal, which will be the final release, after this point, we won't be able to do any further
14:12:53 elodilles release!) -> https://review.opendev.org/c/openstack/releases/+/964471 (and its parents)
14:13:21 bauzas elodilles: ack, I'm on vtpm review mode but I'll look
14:13:28 bauzas don't be afraid of harassing me :)
14:14:05 Uggla elodilles, o/ I'm gonna have a look either later or tomorrow morning, thanks!
14:15:41 elodilles bauzas Uggla : thanks both of you o:) and sorry for already harassing you, but the releases should merge soon, well, the sooner the better, as we should move Caracal to unmaintained in less than a week o:)
14:17:16 elodilles (sorry for not proposing the release patches last week, in that way, i could have harass you IRL at the Summit o:))
14:17:53 Uggla elodilles 👍
14:36:36 bauzas elodilles: physical harassment is not accepted by the Summit Code of Conduct unfortunately
14:36:59 bauzas but you could have bribed with beers
14:38:22 opendevreview Balazs Gibizer proposed openstack/nova master: Rally job for eventlet-removal https://review.opendev.org/c/openstack/nova/+/960130
14:43:16 opendevreview Balazs Gibizer proposed openstack/nova master: Rally job for eventlet-removal https://review.opendev.org/c/openstack/nova/+/960130
14:48:22 opendevreview Yateesh Kumar MN proposed openstack/nova master: Don't destroy TPM during hard reboot https://review.opendev.org/c/openstack/nova/+/964475
15:02:14 elodilles bauzas: ahh, i've missed the bribe :/ maybe next time then! ;)
15:20:52 Uggla Nova meeting in ~40mn
15:33:24 ali I have a question.
15:33:24 ali I have two aggregates named general and dedicated. The availability zone of general is nova, and the zone of dedicated is another one.
15:33:24 ali I wanted to live migrate an instance from the nova zone to the other zone.
15:33:24 ali To do this, I removed the host from the dedicated aggregate and added it to the general aggregate. After the live migration completed, I moved the compute host back to the dedicated aggregate.
15:33:25 ali After that, I tried to resize the instance to its own flavor, but I got an error related to FilterScheduler and zone specifications. I found that in the nova_api database, the availability_zone was updated correctly, but in the nova database, the availability_zone was not updated.
15:33:30 ali Why did this happen?
16:00:31 Uggla #startmeeting nova
16:00:31 opendevmeet Meeting started Tue Oct 21 16:00:31 2025 UTC and is due to finish in 60 minutes. The chair is Uggla. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:31 opendevmeet Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:31 opendevmeet The meeting name has been set to 'nova'
16:00:40 fwiesel o/
16:00:49 r-taketn hello
16:01:22 nicolairuckel o/

Earlier   Later