Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2023-12-19
16:20:37 bauzas so, basically, people can backport some changes to stable/yoga but we shouldn't be merging them ?
16:21:00 bauzas now we released yoga with the last number?
16:21:00 elodilles also this is a good reading for people who are interested in the topic: https://governance.openstack.org/tc/resolutions/20231114-amend-unmaintained-status.html
16:21:28 sean-k-mooney the stable/yoga branch will be deleted after the tag is created adn unmaintained/yoga is created
16:21:40 sean-k-mooney one thing we need to do is update the .gitreview file
16:21:46 sean-k-mooney once we cut the new branch
16:21:46 elodilles bauzas: we can merge, the tag will be applied at the HEAD of stable/yoga (like the *-eol tag)
16:22:04 bauzas elodilles: okay
16:22:13 bauzas good to know
16:22:58 bauzas so, basically, we can still backport some changes if we need (like some important fixes), but it would be better to wait until yoga is unmaintained
16:23:49 bauzas people interested in maintaining yoga would tho still want to backport changes to the stable branch, but we shouldn't be releasing it
16:23:49 elodilles +1
16:24:06 bauzas I'm cool with the plan then
16:24:17 bauzas moving on
16:24:19 bauzas #topic vmwareapi 3rd-party CI efforts Highlights
16:24:28 fwiesel #Info Script installs functional devstack, starting from a blank vm snapshot in ca. 500s
16:24:29 bauzas fwiesel: grandchild: if you want to tell us
16:24:44 bauzas fwiesel: bravo !
16:24:57 fwiesel So, we should be able to hook now things up, but...
16:25:04 fwiesel #Info Need to lock down network before hooking up CI (ETA: Roughly Mid-January due to holidays)
16:25:34 fwiesel For the curious ones, the stack.log I've put here: http://sap-openstack-ci.s3-website.us-east-2.amazonaws.com/logs-stack.tar.gz
16:25:39 bauzas heh, sure :)
16:25:58 fwiesel We will probably dump there also then the output of the test-runs, so they are publically visible
16:26:13 bauzas fwiesel: I got a 403 from the link
16:26:15 elodilles (fwiesel: it says 403 forbidden)
16:27:07 fwiesel Hmm, I'll have to check with the colleague then. Just five minutes I we fixed the permissions. And I do not have the means to change them myself
16:27:19 bauzas fwiesel: no rush
16:27:33 sean-k-mooney for what its worht this https://www.ovhcloud.com/en-ie/web-hosting/performance-offer/ used to be what we used to host the logs publicly in teh past
16:27:54 sean-k-mooney it is much cheaper then aws
16:28:19 bauzas I didn't wanted to play the French card :D
16:28:25 fwiesel Not if you factor in the bureaucracy to get that through in a big company
16:28:57 sean-k-mooney so its 120 euro a year so it genrelly does not require appoveal and can be expensed in most companies :)
16:29:00 sean-k-mooney but sue
16:29:07 sean-k-mooney *sure
16:29:17 bauzas we're publicly logged here so I won't speak that much, but yeah AWS is good as being the primer
16:29:21 sean-k-mooney anyway good to know
16:29:43 fwiesel We probably put it behind a dns redirect, so we can switch it whenever we want.
16:30:08 bauzas fwiesel: that's a good idea and you shouldn't publiclize what you use for your environment IMHO )
16:30:10 bauzas :)
16:30:31 bauzas even if people can see the dns redirect :)
16:30:50 bauzas anyway
16:31:04 bauzas thanks for the report, and that's very good to hear that your lab is running :)
16:31:11 bauzas kudos again
16:31:38 fwiesel bauzas: You're welcome. Thanks for giving us the opportunity to pitch in.
16:31:46 bauzas shall we move on ?
16:31:55 bauzas I have an item
16:31:56 fwiesel #Info Next update on 9th January
16:32:03 bauzas noted.
16:32:13 bauzas and it will transition to my own item actually :)
16:32:18 bauzas #topic Open discussion
16:32:24 bauzas (bauzas) Skip next two nova meetings ?
16:32:24 bauzas (bauzas) Skip next two nova meetings ?
16:32:35 sean-k-mooney i think yes
16:32:56 bauzas so, I'm off from Christmas to Jan 8
16:33:10 bauzas (Jan 8 not included, tbc)
16:33:25 gibi +1 on skipping
16:33:29 gibi I'm back on the 15th
16:33:31 elodilles +1 from me, too
16:33:48 bauzas anyone having concerns and wanting to discuss the next day after Jan 1 ? :D
16:34:17 bauzas looks not
16:34:42 bauzas #agreed Next two nova meetings (Dec 26 and Jan 2) will be SKIPPED. Next meeting will be Jan 9
16:34:55 bauzas #action bauzas to communicate this over ML
16:35:05 bauzas okay
16:35:11 bauzas so, we have another item
16:35:18 bauzas that we already discussed last week
16:35:24 bauzas (JayF/johnthetubaguy) Specless blueprint for ironic guest metadata
16:35:28 bauzas https://blueprints.launchpad.net/nova/+spec/ironic-guest-metadata
16:35:33 JayF I've put a significant amount of detail in this since last week
16:35:39 bauzas JayF: fancy taking the baton again ?
16:35:42 bauzas cool
16:35:43 JayF I'm happy to discuss here, or async in the blueprint/mailing list/etc
16:36:11 JayF This is mostly based on conversations with sean-k-mooney, especially the part about generalizing the libvirt code
16:36:55 sean-k-mooney just readign it now
16:37:10 bauzas JayF: I just have one question
16:37:43 JayF I either have an answer or a clever deflection technique ;)
16:37:46 bauzas with a SLURP environment consisted of Caracal controllers, Caracal computes but also Antelope computes, how Ironic could use this ?
16:38:07 JayF Ironic is going to just look for the metadata provided by Nova in instnace_info
16:38:16 sean-k-mooney im not sure that will matter given this will be entrily in the driver
16:38:18 JayF if it's there; Ironic will add the lessee; if not, it won't
16:38:39 JayF if someone is migrating, they'll migrate *after* the upgrade is complete and because the migration script is additive only, it'll make the environment consistent
16:39:00 sean-k-mooney if the ironic compute are running antelope code it just wotn set that data
16:39:03 bauzas JayF: here, I just want to hear that you understand that the metadata will be optional, at least until we upgrade to 2024.2
16:39:11 sean-k-mooney if its upgraded it can check the ironic api version if needed
16:39:22 JayF Absolutely; 100%, ironic will opportunistically use that metadata if it exists, we will not make a hard dependency on it
16:39:54 JayF it'd be an API break for us to do anything else
16:40:03 sean-k-mooney the ironic script that will be created for backfilling can also be used to reconsile everythign after the upgrade is fully done
16:40:22 opendevreview Stephen Finucane proposed openstack/nova master: Resolve mypy error https://review.opendev.org/c/openstack/nova/+/903530
16:40:23 opendevreview Stephen Finucane proposed openstack/nova master: pre-commit: Add mypy https://review.opendev.org/c/openstack/nova/+/896228
16:40:23 opendevreview Stephen Finucane proposed openstack/nova master: tox: Use pre-commit for pep8 target https://review.opendev.org/c/openstack/nova/+/896227
16:40:24 opendevreview Stephen Finucane proposed openstack/nova master: pre-commit: Bump linter versions https://review.opendev.org/c/openstack/nova/+/903749
16:40:36 bauzas JayF: then I'm cool
16:40:41 JayF woo
16:40:49 sean-k-mooney i think im oke with moving forward with this as a specless bluepint and movign the dicussion to the implementaion
16:40:51 bauzas from a nova pov, this is just a small thing
16:41:12 JayF That's sorta how I feel about it too, but it's hard for me to tell :)
16:41:18 bauzas but again, I don't want Ironic to have a bug because we continue to support Antelope computes, that's it :)
16:41:40 bauzas if the plan is already sorted, then yay, sounds specless
16:41:53 bauzas anyone concerned by any design problem ?
16:42:10 bauzas anyone thinking of some nit that would need to be discussed ?
16:42:29 bauzas apparently no

Earlier   Later