Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2021-07-13
16:10:55 tosky elodilles: I'd say it's worth it: it should help people who will keep on eye on that branch when a failure will happen
16:12:23 elodilles i see.... it's just.... really not that nice of a patch :]
16:12:52 elodilles for backport :X
16:13:32 gibi elodilles: so far we backported that to newer stable branches to unblock gate there?
16:13:56 elodilles well, it became like that in ussuri
16:14:11 elodilles in newer branches it was a *bit* nicer
16:14:37 elodilles but yes, we did backport it so far
16:14:47 elodilles until ussuri
16:14:58 elodilles since gate was broken back to ussuri
16:15:50 gibi so the gate is OK in train, so this becomes a nice to have compared to ussure where it was a must have
16:16:00 gibi anyhow I let the stable core team to decide
16:17:40 gibi let me know if you stuck on deciding and need my final ruling :)
16:17:57 gibi moving on
16:18:04 gibi the placement gate status looks green
16:18:16 gibi any other topic about the gate?
16:19:25 gibi #topic Release Planning
16:19:29 gibi Milestone 2 is Thursday (15 of July) which is spec freeze
16:19:45 gibi https://review.opendev.org/q/project:openstack/nova-specs+status:open
16:19:45 gibi we have couple of open specs
16:19:55 gibi melwitt: has two re-propose specs that needs a second core
16:20:33 gibi nova-audit https://review.opendev.org/c/openstack/nova-specs/+/800570
16:20:41 gibi and
16:20:43 gibi consumer types https://review.opendev.org/c/openstack/nova-specs/+/800569
16:21:14 gibi these are pretty easy to get in before the deadline I think
16:21:32 gibi there is one more spec with only positive feedback https://review.opendev.org/c/openstack/nova-specs/+/787458
16:21:40 gibi Integration With Off-path Network Backends ^^
16:22:15 gibi I sean-k-mooney said in the review that we wait for bauzas to review
16:22:31 gibi but I think bauzas is already on PTO
16:22:45 stephenfin he is
16:22:46 sean-k-mooney i think part of the concern there is the neutorn apporval and ovn support
16:23:08 sean-k-mooney we could approve it for this cycle but im not sure the depencies will line up before feature freeze
16:23:34 sean-k-mooney so we could leave that to the implemantion reivew and be optimistic or defer
16:23:44 sean-k-mooney until everyting is ready
16:23:44 gibi sean-k-mooney: could you please summarize this dependency complication in the spec review?
16:23:55 sean-k-mooney gibi: its already captured in the review
16:24:05 gibi ohh, cool then I missed that
16:24:18 sean-k-mooney https://review.opendev.org/c/openstack/nova-specs/+/787458/9/specs/xena/approved/integration-with-off-path-network-backends.rst#654
16:24:43 sean-k-mooney we are not expecting the nova design to change
16:25:07 sean-k-mooney but we likely want to wait for the depencies to be resolved before merging the nova code
16:25:49 sean-k-mooney gibi: i agree on nova-audit and consumer types
16:26:00 gibi sean-k-mooney: thanks, I will read the last part of the discussion in the spec tomorrow and then try to decide if I upgrade my vote
16:26:00 sean-k-mooney i think those could be progressed before spec freeze
16:26:58 gibi the rest of the open specs has negative feedback on them so they have small chance to land until Thursday but I will keep an eye on them
16:27:32 gibi anything else about the coming spec freeze?
16:28:50 gibi #topic Stable Branches
16:28:54 gibi stable gates are not blocked
16:28:59 gibi stable releases have been proposed:
16:29:08 gibi victoria: https://review.opendev.org/800476
16:29:08 gibi wallaby: https://review.opendev.org/800475
16:29:12 gibi ussuri: https://review.opendev.org/800477
16:29:16 gibi EOM (from elodilles )
16:29:48 gibi I've already looked at the stable release patches and they look good to me. I will wait for lyarwood before approve
16:29:53 elodilles lyarwood: thanks in advance :)
16:30:02 lyarwood ack np, sorry I didn't get to them already
16:30:09 gibi lyarwood: thanks!
16:30:16 gibi anythin else on stable land?
16:30:32 elodilles nothing from me
16:31:12 gibi I'm skipping the libvirt subteam topic as we dont have bauzas
16:31:17 gibi #topic Open discussion
16:31:25 gibi nothing on the agenda
16:31:30 gibi but I have one note
16:31:47 gibi I sent out a doodle about the coming PTG
16:32:00 sean-k-mooney ah i was going to ask about that i saw the ooo one
16:32:15 gibi #link http://lists.openstack.org/pipermail/openstack-discuss/2021-July/023560.html
16:32:35 gibi in the ML thread I have a question about the amount of slots we would need
16:32:54 gibi and also we have brand new PTG etherpad too https://etherpad.opendev.org/p/nova-yoga-ptg
16:33:21 sean-k-mooney i do think that 4 days with 4 hour slot is likely better yes
16:33:40 sean-k-mooney if it was in person we likely owuld do 3 8 hour slots - breaks
16:35:07 gibi yeah my experience tells me we could use more time than 3x4 hours
16:35:18 sean-k-mooney when do you need to inform the organisers
16:35:36 gibi 21st of July
16:35:41 gibi so next week
16:35:42 sean-k-mooney ack
16:35:53 gibi but I think I can always add more slots then remove it later
16:36:08 sean-k-mooney yep
16:36:20 gibi any other topic for today?
16:36:23 alistarle Hello guys, concerning this blueprint https://review.opendev.org/c/openstack/nova/+/794837, do you have any question about it, and what is the next step ?
16:37:21 sean-k-mooney looks liek lee has a -1
16:37:32 sean-k-mooney has this been disucsed as a specless blueprint in the past
16:38:13 alistarle Yep, but I already answer in previous comment I think, but I can sum up for lee I for sure
16:38:30 lyarwood alistarle: yeah sorry did I miss the justification for this?
16:38:44 alistarle What do you mean about a speechless blueprint ? Because yes, there is a blueprint for that: https://blueprints.launchpad.net/nova/+spec/configure-direct-snapshot
16:38:47 lyarwood alistarle: I couldn't work out how you would be able to do this without manually flattening disks
16:39:09 gibi sean-k-mooney: I think this is the time alistarle asks for the approval of the specless bp
16:39:43 sean-k-mooney alistarle: that is a blueprint not a spec. a spec is a serate document that describe the usecase in detial and the design
16:39:45 sean-k-mooney gibi: ack
16:40:01 sean-k-mooney alistarle: can you summerise what you want to enable
16:40:50 alistarle Sure, I want to make the direct snapshot feature configurable, because for now, nova will not honor the glance configuration when creating a snapshot, but trying to guess it from the actual nova backend
16:41:18 dansmith what does that mean? "not honor the glance configuration" ?
16:42:10 sean-k-mooney alistarle: is this a glance multistore envoinment? or a case where glance does not use ceph for storage?
16:42:22 alistarle Let's say you 1. Store your image on ceph, and use copy on write to create VM, so you are storing your VM disk as images child in RBD, but during life of your cloud you want to move out from ceph. So 2. You remove your rbd backend from glance and put a swift one instead
16:43:21 sean-k-mooney alistarle: are we talking about cinder volumes for the vm or the rbd image backend?
16:43:21 alistarle With this configuration, nova will never call glance to check which backend is enabled, and he will always store the snapshot in ceph, even if it is not a declared glance backend
16:43:29 sean-k-mooney in nova.conf
16:43:32 lyarwood sean-k-mooney: rbd image backend
16:44:01 alistarle It it for a case when you used RBD as a glance store, then you replace it with a swift one, and you want to progressively move out from ceph
16:44:10 sean-k-mooney so the request is to flaten the snapshots effectivly in ceph and then have new shapshots go to swift
16:44:18 dansmith alistarle: okay I thought I asked you this question on the spec
16:44:41 sean-k-mooney alistarle: well one thing we do not support is changing the nova image_types backend
16:45:01 sean-k-mooney including via a move operation like cold migrate

Earlier   Later