| *** mhen_ is now known as mhen | 02:13 | |
| *** auniyal681 is now known as auniyal68 | 04:21 | |
| opendevreview | Anthony Galica proposed openstack/cinder master: [Hitachi] Filter/sorting during group searches https://review.opendev.org/c/openstack/cinder/+/1009405 | 06:45 |
|---|---|---|
| opendevreview | Rutuja Pawar proposed openstack/cinder master: Fixed IBM Storwize/FlashSystem GMCV volume resize failure by waiting for replication relationship synchronization and added configurable sync timeout and polling options. BUG ID : https://launchpad.net/bugs/2155894 https://review.opendev.org/c/openstack/cinder/+/1004999 | 07:46 |
| opendevreview | Rutuja Pawar proposed openstack/cinder master: Fixed IBM Storwize/FlashSystem GMCV volume resize failure by waiting for replication relationship synchronization and added configurable sync timeout and polling options. Closes-Bug: #2155894 https://review.opendev.org/c/openstack/cinder/+/1004999 | 09:16 |
| opendevreview | Liron Kuchlani proposed openstack/cinder-tempest-plugin master: Add consistent volume group replication tests https://review.opendev.org/c/openstack/cinder-tempest-plugin/+/1014046 | 09:30 |
| opendevreview | Merged openstack/cinder master: Make pylint job voting https://review.opendev.org/c/openstack/cinder/+/1010487 | 09:31 |
| opendevreview | Rutuja Pawar proposed openstack/cinder master: IBM SVf: Add unit tests for multiattach force detach https://review.opendev.org/c/openstack/cinder/+/1014061 | 10:37 |
| opendevreview | Takashi Kajinami proposed openstack/cinder-tempest-plugin master: Make failing jobs non-voting https://review.opendev.org/c/openstack/cinder-tempest-plugin/+/1011526 | 10:37 |
| opendevreview | Takashi Kajinami proposed openstack/cinder-tempest-plugin master: Add stable/2026.2 job https://review.opendev.org/c/openstack/cinder-tempest-plugin/+/1005308 | 10:37 |
| opendevreview | Medha Choudhary netapp proposed openstack/cinder master: NetApp: Use REST for CG failover target selection https://review.opendev.org/c/openstack/cinder/+/1008054 | 10:39 |
| abhishekk | greetings o/ need some help in moving this forward, https://review.opendev.org/c/openstack/cinder/+/1011471, Thank you!! | 13:46 |
| rosmaita | Cinder meeting in #openstack-cinder at 1400 UTC | 14:00 |
| rosmaita | jungleboyj rosmaita smcginnis tosky whoami-rajat m5z e0ne geguileo eharney jbernard hemna fabiooliveira yuval tobias-urdin adiare happystacker dosaboy hillpd msaravan sp-bmilanov Luzi sfernand simondodsley zaubea nileshthathagar flelain wizardbit agalica lutimura kaisers aloke_dev Anoop_Shukla erlon Jyotsna rakshay ^^ | 14:00 |
| yuval | 0/ | 14:00 |
| rosmaita | #startmeeting cinder | 14:01 |
| opendevmeet | Meeting started Wed Nov 4 14:01:00 2026 UTC and is due to finish in 60 minutes. The chair is rosmaita. Information about MeetBot at http://wiki.debian.org/MeetBot. | 14:01 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 14:01 |
| opendevmeet | The meeting name has been set to 'cinder' | 14:01 |
| rosmaita | #topic roll call | 14:01 |
| rosmaita | hello yuval | 14:01 |
| pungrumpy | o/ | 14:01 |
| simondodsley | o/ | 14:01 |
| aloke_dev | o/ | 14:01 |
| harsh | o/ | 14:01 |
| fqi | o/ | 14:01 |
| yuval | whats new? | 14:01 |
| jayaanand | o/ | 14:01 |
| hvlcchao1 | o/ | 14:01 |
| Sai | o/ | 14:02 |
| rakshay | o/ | 14:03 |
| Jyotsna | hi | 14:04 |
| rosmaita | hello everyone, guess we should get started | 14:05 |
| *** auniyal68 is now known as auniyal | 14:05 | |
| rosmaita | i was just looking at a patch where the releasenotes job seemed to fail for reasons nothing to do with the patch | 14:05 |
| rosmaita | but luckily the build history shows that was the only patch where that happened recently | 14:06 |
| rosmaita | anyone else with gate observations? | 14:06 |
| rosmaita | i think the cinder-tempest-plugin gate is still borked, there are some patches up to address that | 14:06 |
| Anoop_Shukla | O/ | 14:07 |
| rosmaita | ok, i will be cautiously optimistic that the gates are OK this week | 14:07 |
| auniyal | I see we have to do lot recheck these - mostly unit py314 job | 14:07 |
| rosmaita | auniyal: are they actual test failures, or the job isn't running? | 14:08 |
| auniyal | but nothing realted to patch, or anything blocking - we can proceed I guess | 14:08 |
| Jyotsna | some timeout frequently and after running recheck it works | 14:09 |
| auniyal | mostly either waiting or 500 on API - but nothing repetative which I could remember | 14:09 |
| rosmaita | Jyotsna, auniyal: thanks, that's what i have been seeing too | 14:09 |
| rakshay | nothing of that on my recent patch though! | 14:10 |
| vijay-jatti | Hi | 14:10 |
| rosmaita | looks like it's not just us: https://zuul.opendev.org/t/openstack/builds?job_name=openstack-tox-py314&branch=master&skip=100 | 14:11 |
| rosmaita | but it also looks like it's mostly passing | 14:12 |
| rosmaita | or at least the failures are spread out enough that it hasn't attracted infra's attention | 14:12 |
| rosmaita | just a general reminder to people: if you see the job failing, please take a look at the result and make sure it's not a "real" test failure before rechecking | 14:13 |
| rosmaita | and in your comment on the patch say something like | 14:13 |
| erlon | \o | 14:14 |
| rosmaita | recheck openstack-tox-py314 - job timed out, not related to patch | 14:14 |
| rosmaita | #topic announcements | 14:14 |
| rosmaita | Festival of Reviews on friday! | 14:14 |
| rosmaita | #link https://etherpad.opendev.org/p/cinder-festival-of-reviews | 14:14 |
| rosmaita | same google meet room, same time 1400-1600 UTC | 14:15 |
| rosmaita | upcoming deadlines | 14:15 |
| jayaanand | Tox-py314 is failing for multiple patches | 14:15 |
| rosmaita | jayaanand: it seems to be mostly passing, though: https://zuul.opendev.org/t/openstack/builds?job_name=openstack-tox-py314&project=openstack%2Fcinder&branch=master&skip=0 | 14:16 |
| rosmaita | that failure looks for real on 1014061,1 | 14:17 |
| jayaanand | rosmaita: thanks for confirming. I will initiate recheck | 14:17 |
| rosmaita | btw, in that listing, the jobs whose result is RETRY are automatically retried by zuul | 14:18 |
| jayaanand | ok | 14:18 |
| rosmaita | so the situation isn't quite as bad as it looks | 14:19 |
| rosmaita | still, it is annoying | 14:19 |
| rosmaita | oh, yeah, so back to announcements | 14:19 |
| rosmaita | spec freeze is Nov 20 | 14:20 |
| rosmaita | that week is also milestone-1 for Indri | 14:20 |
| rosmaita | we will want (or the release team will require us) to do an early os-brick release at M-1 | 14:21 |
| rosmaita | it would be good to have the nvme-of patches reviewed so we can get an early release out with that support | 14:21 |
| rosmaita | for people to test with | 14:21 |
| aloke_dev | simondodsley, I have addressed all review comments on https://review.opendev.org/c/openstack/os-brick/+/986070. Please have a look when you have a moment. Also enabled InfoScale CI for the os-brick changes. | 14:23 |
| aloke_dev | Wanted to mention it because it is related to os-brick. | 14:23 |
| rosmaita | aloke_dev: also, that is holding up the infoscale driver? | 14:24 |
| Anoop_Shukla | Need some attention on the patch: https://review.opendev.org/c/openstack/cinder/+/1010351 from core reviewers. This addresses a fix for FC similar to one done for iSCSI recently. | 14:24 |
| aloke_dev | Yes. rosmaita. InfoScale Cinder driver has dependency on os-brick patch | 14:25 |
| rosmaita | yes, and iirc, InfoScale driver is a priority for M-1, as we discussed at the PTG | 14:25 |
| rosmaita | that's all the announcements | 14:27 |
| rosmaita | #topic read-only roles working in unexpected ways | 14:27 |
| rosmaita | #link https://review.opendev.org/c/openstack/cinder/+/1012421 | 14:28 |
| rosmaita | is pungrumpy here? | 14:28 |
| rosmaita | let me say something about this patch/bug | 14:28 |
| pungrumpy | hi, thanks. this is https://review.opendev.org/c/openstack/cinder/+/1012421 for LP #1522264. today the only way to give a role read access to every project's volumes/snapshots/backups is to add it to context_is_admin, and then every is_admin check passes for it, deletes included. | 14:29 |
| pungrumpy | the change adds volume:get_all:all_tenants, volume:get_all_snapshots:all_tenants and backup:get_all:all_tenants, all defaulting to rule:admin_api, so nothing changes until an operator grants them. no new microversion, no API change. | 14:29 |
| pungrumpy | two questions: is this approach OK, and does it need a spec or blueprint first? there is an old bp allow-all-tenants-query (2022) asking for the same. a follow-up for transfers/groups/group snapshots/attachments is ready and waits on the answer. | 14:29 |
| rosmaita | so here's the situation | 14:29 |
| rosmaita | we still do not have complete tempest coverage for secure RBAC in the cinder-tempest-plugin | 14:29 |
| rosmaita | there are several patches that have been proposed, but are passing the gate because they skip tests | 14:30 |
| rosmaita | but, it seems like the failing tests are a test issue, because we haven't been able to reproduce manually | 14:30 |
| rosmaita | so | 14:30 |
| rosmaita | the issue pungrumpy wants to address is to clean out some of the is_admin database layer checks that shouldn't be necessary | 14:31 |
| rosmaita | but | 14:31 |
| simondodsley | aloke_dev: i'll take another look - thanks | 14:31 |
| rosmaita | i am scared to clean these out until we have full test coverage and know that they really are unnecessary | 14:31 |
| erlon | we should be as rigorous as possible with tests since, they 'cost' less to write now, and the amount of codes to review is becoming overwhelming | 14:32 |
| rosmaita | erlon: ++ | 14:32 |
| auniyal | we should add functional tests too for this | 14:32 |
| aloke_dev | simondodsley: Thank you! | 14:32 |
| pungrumpy | rosmaita: one clarification, this change does not remove any is_admin check from the db layer. cinder/db is untouched, @require_admin_context stays. | 14:32 |
| pungrumpy | it only replaces the is_admin test above the db that decides all_tenants, with a policy rule. the unit tests assert db.volume_get_all still raises AdminRequired for the reader. | 14:32 |
| rosmaita | ok, that is good to know | 14:33 |
| rosmaita | but i guess my wider point is that i would kind of like to hold your patch hostage until we get the secure RBAC tests fixed | 14:34 |
| rosmaita | which is admittedly unfair to you, but this is some technical debt that we need to get addressed | 14:35 |
| pungrumpy | rosmaita: understood, thanks for being upfront. let me check with my team and reply on the review. | 14:36 |
| pungrumpy | which patches or tests should I look at for the secure RBAC gap? I would like to read them before we answer. | 14:36 |
| rosmaita | and it is always hard to find time to address technical debt, unless there is a highly motivated party | 14:36 |
| rosmaita | pungrumpy: let's talk after the meeting ... to be clear, you aren't going to have to fix everything yourself | 14:36 |
| pungrumpy | rosmaita: sure, I will stay after the meeting. thanks. | 14:37 |
| rosmaita | but it would be helpful to have a motivated person around to drive the effort | 14:37 |
| rosmaita | pungrumpy: ty | 14:37 |
| rosmaita | #open discussion | 14:37 |
| rosmaita | well, that's not a command | 14:37 |
| rosmaita | #topic open discussion | 14:37 |
| vijay-jatti | greetings o/ need some help in moving this forward,https://review.opendev.org/c/openstack/cinder/+/1006944 Active HPE AlletraMP : Feature Automate Failback cinder replication recovery, Thank you!! | 14:38 |
| Jyotsna | 1008049: Skip sync wait for same-size clones - https://review.opendev.org/c/openstack/cinder/+/1008049 : (rework done suggested by auniyal, lost previous +1 and +2, so requires attention from core reviewers again for +2) 1001281: Improved host proximity for ActivePP support https://review.opendev.org/c/openstack/cinder/+/1001281: (+2 pending , Zuul CI passed) 1001152: Improved failover/failback with Active PP support ht | 14:39 |
| rosmaita | ok, looks like no discussion is desired; please grab your now free 20 minutes and review something! | 14:41 |
| rosmaita | #endmeeting | 14:41 |
| opendevmeet | Meeting ended Wed Nov 4 14:41:46 2026 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 14:41 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/cinder/2026/cinder.2026-11-04-14.01.html | 14:41 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/cinder/2026/cinder.2026-11-04-14.01.txt | 14:41 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/cinder/2026/cinder.2026-11-04-14.01.log.html | 14:41 |
| pungrumpy | rosmaita: thanks. I would like to help drive the secure RBAC tests, it is what my change needs anyway. I can't promise timing yet, I will confirm that on the review. | 14:41 |
| pungrumpy | where should I start? which patches in cinder-tempest-plugin are the skipped ones, and which failing tests do you want looked at first? | 14:41 |
| opendevreview | Merged openstack/cinder master: Upgrade to pylint 4.1.0 https://review.opendev.org/c/openstack/cinder/+/1010692 | 14:42 |
| rosmaita | pungrumpy: give me a minute to take a look | 14:42 |
| opendevreview | Merged openstack/cinder master: fix: remove unused reservations list and related logic https://review.opendev.org/c/openstack/cinder/+/980192 | 14:42 |
| opendevreview | Merged openstack/cinder master: Remove comment to describe behavior in Python 2 https://review.opendev.org/c/openstack/cinder/+/927711 | 14:43 |
| auniyal | pungrumpy this might be helpful https://review.opendev.org/c/openstack/cinder/+/989543/2/cinder/api/middleware/auth.py | 14:44 |
| pungrumpy | auniyal: thanks, I will read it. | 14:44 |
| pungrumpy | rosmaita: one more for when you have a moment: do you have a rough idea of when the secure RBAC tests could be done, and what would be enough to look at 1012421 again? full coverage, or the parts it touches (volumes, snapshots, backups)? | 14:44 |
| rosmaita | pungrumpy: these are the patches we haven't merged yet: topic:"secure-rbac" project:openstack/cinder-tempest-plugin status:open | 14:46 |
| rosmaita | a real link would be helpful | 14:46 |
| rosmaita | https://review.opendev.org/q/topic:%22secure-rbac%22+project:openstack/cinder-tempest-plugin+status:open | 14:46 |
| rosmaita | a list like that is kind of demotivating | 14:47 |
| rosmaita | if you look at the patches, you see a bunch of skip decorators in there | 14:48 |
| rosmaita | for example https://review.opendev.org/c/openstack/cinder-tempest-plugin/+/889379/22/cinder_tempest_plugin/rbac/v3/test_volume_actions.py | 14:48 |
| rosmaita | scroll down to around line 406 | 14:49 |
| rosmaita | one problem is that if you haven't worked much with tempest, the design pattern is confusing (at least I think so) | 14:50 |
| rosmaita | my suspicion is that there is some flaw in the test structure such that once it's figured out, it may be relatively easy to fix | 14:50 |
| rosmaita | so one approach could be to pick the patch that seems to be most closely related to your change, and see if you can make any sense of it | 14:51 |
| rosmaita | and we may be able to set up a video call with a tempest expert and discuss and get some pointers | 14:52 |
| rosmaita | the person I have in mind is in utc-8:00 time zone | 14:53 |
| rosmaita | not sure where you are? | 14:53 |
| pungrumpy | rosmaita: I'm in Thailand, UTC+7. so their late afternoon is my early morning, that could work. thanks, I'll read 889852 and 889379 first so I have something concrete to ask. | 14:57 |
| rosmaita | ok, thanks ... and don't forget, you don't have to do this all yourself | 14:57 |
| opendevreview | Anthony Galica proposed openstack/cinder master: [Hitachi] Filter/sorting during group searches https://review.opendev.org/c/openstack/cinder/+/1009405 | 14:58 |
| pungrumpy | rosmaita: thanks, that helps. is there a target for getting the tests done, like a milestone in Indri? and for 1012421, does it need all of them, or would volumes, snapshots and backups be enough? | 14:58 |
| opendevreview | Merged openstack/cinder master: Ensure driver is initialized before attach https://review.opendev.org/c/openstack/cinder/+/794001 | 14:59 |
| rosmaita | pungrumpy: i need to look at your patch more closely, but if you are right about the db layer, good progress on the most closely related tempest tests would be enough | 14:59 |
| pungrumpy | rosmaita: that is very clear, thank you. I'll work on those. | 14:59 |
| pungrumpy | separate question from my team: capacity-based QoS on the RBD back-end (*_per_gb and *_per_gb_min applied on the ceph side, recomputed on extend and retype). upstream only computes those for consumer=front-end. | 15:00 |
| rosmaita | pungrumpy: M-2 would be good, but we can be flexible | 15:00 |
| pungrumpy | we already run a patch for this in our deployment and are ready to send it upstream if the team wants it. is the team open to that? should it go through a spec, or revive 940275 (bp cinder-rbd-qos-total-iops-per-gb)? | 15:00 |
| *** agalica__ is now known as agalica_ | 15:00 | |
| agalica_ | Oh, the meeting starts at 6am my time doesn’t it. Lol | 15:00 |
| agalica_ | Forgot time change | 15:00 |
| rosmaita | agalica_: good morning! | 15:00 |
| agalica_ | Morning! | 15:01 |
| rosmaita | pungrumpy: so this would be backend ceph qos? | 15:01 |
| pungrumpy | rosmaita: yes, consumer=back-end. the computed limit is written to the rbd image as rbd_qos_*_limit. upstream rbd.py maps only the flat keys and *_max today, so per_gb has no effect there. | 15:05 |
| rosmaita | pungrumpy: i don't think you would need a spec, the LP blueprint and https://review.opendev.org/c/openstack/cinder/+/940275 should be fine | 15:07 |
| rosmaita | looks like there hasn't been any action on that patch in over a year; you could drop the author an email and ask if it's ok to take it over | 15:09 |
| pungrumpy | rosmaita: great, no spec is good news. I'll take that back to my team. | 15:09 |
| pungrumpy | last one, since spec freeze is Nov 20: same answer for 1012421, is bp allow-all-tenants-query enough, or would you want a spec for that one? | 15:09 |
| rosmaita | pungrumpy: that one is kind of an API change, so i think it probably would need a spec | 15:12 |
| rosmaita | but hopefully it can be short, you're proposing to add some finer-grained policies than we currently have | 15:14 |
| pungrumpy | rosmaita: ok, I'll write one before the Nov 20 freeze. would a spec-lite be enough, or a full spec? | 15:14 |
| rosmaita | pungrumpy: i don't think we have spec-lite, so a full spec | 15:15 |
| rosmaita | but you can keep it short | 15:16 |
| pungrumpy | rosmaita: ok, a full spec, kept short. thanks a lot for your time today, I'll follow up on the review. | 15:16 |
| rosmaita | pungrumpy: something like this one: https://specs.openstack.org/openstack/cinder-specs/specs/2026.1/openapi.html | 15:17 |
| pungrumpy | rosmaita: perfect, I'll use that as the template. thanks! | 15:17 |
| rosmaita | np | 15:17 |
| tkajinam | hi could anyone please help me merge https://review.opendev.org/c/openstack/cinder-tempest-plugin/+/1011526 ? this is becoming more urgent since 2025.1 branch was transitioned to unmaintained phase... | 16:18 |
| opendevreview | Leo Jenkins proposed openstack/cinder-specs master: Support for Hostname-based Ceph Mon References https://review.opendev.org/c/openstack/cinder-specs/+/1000721 | 16:19 |
| opendevreview | Sebastien Piechurski proposed openstack/cinder master: Update SPDK driver to take into account param changes in bdev_lvol_create and bdev_lvol_resize from size to size_in_mib. https://review.opendev.org/c/openstack/cinder/+/981785 | 17:45 |
| opendevreview | Harsh Ailani proposed openstack/cinder master: [IBM SVf] Added FlashCopy 2.0 snapshot support for individual volumes, including create, restore, delete, and clone operations. https://review.opendev.org/c/openstack/cinder/+/1010161 | 20:16 |
| opendevreview | Simon Dodsley proposed openstack/cinder master: Add a lazy option to the privsep umount helper https://review.opendev.org/c/openstack/cinder/+/1014185 | 20:47 |
| opendevreview | Simon Dodsley proposed openstack/cinder master: Add Everpure FlashBlade NFS volume driver https://review.opendev.org/c/openstack/cinder/+/1014186 | 20:47 |
| opendevreview | Merged openstack/os-brick master: Add openstack-python3-next-jobs https://review.opendev.org/c/openstack/os-brick/+/998182 | 21:31 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!