Wednesday, 2026-10-07

opendevreviewJyotsna Lothe proposed openstack/cinder master: HPE AlletraMP driver : Online clone synchronous wait for copy increases bulk VM provisioning time  https://review.opendev.org/c/openstack/cinder/+/100804903:32
opendevreviewJyotsna Lothe proposed openstack/cinder master: HPE AlletraMP driver : Online clone synchronous wait for copy, increases bulk VM provisioning time  https://review.opendev.org/c/openstack/cinder/+/100804903:52
opendevreviewJyotsna Lothe proposed openstack/cinder master: HPE AlletraMP driver : Online clone synchronous wait for copy, increases bulk VM provisioning time  https://review.opendev.org/c/openstack/cinder/+/100804903:59
opendevreviewJyotsna Lothe proposed openstack/cinder master: HPE Alletra MP: Skip sync wait for same-size clones  https://review.opendev.org/c/openstack/cinder/+/100804904:03
opendevreviewRajat Dhasmana proposed openstack/cinder-tempest-plugin master: Scenario: Verify secondary failback writes  https://review.opendev.org/c/openstack/cinder-tempest-plugin/+/100820006:00
opendevreviewRajat Dhasmana proposed openstack/cinder master: Improve volume delete error message  https://review.opendev.org/c/openstack/cinder/+/98900406:26
auniyalzuul is not running jobs with recheck, fungi any ideas06:34
opendevreviewJyotsna Lothe proposed openstack/cinder master: HPE AlletraMP driver : Online clone synchronous wait for copy, increases bulk VM provisioning time  https://review.opendev.org/c/openstack/cinder/+/100804907:08
fricklerauniyal: do you have an example? usually this happens for changes that need a rebase07:27
auniyalhttps://zuul.opendev.org/t/openstack/status?change=99593407:29
auniyalfrickler so these 2 patch might be good examples https://zuul.opendev.org/t/openstack/status?change=995934&change=98900407:30
sp-bmilanovhey simondodsley, I see your third-party CI logs (  https://github.com/Everpure-Zuul-CI/Cinder-CI/tree/gh-pages ) .. why Github? Easiest to manage?07:34
opendevreviewRajat Dhasmana proposed openstack/cinder master: RBD: Update volume location on failover  https://review.opendev.org/c/openstack/cinder/+/100730607:35
sp-bmilanovalso, how did you set up your current iteration? is it sill SoftwareFactory?07:35
-opendevstatus- NOTICE: zuul processing is broken since about 03:00 UTC, investigation is in progress, please be patient08:01
opendevreviewMarkus Hentsch proposed openstack/cinder master: Adopt image encryption standardization  https://review.opendev.org/c/openstack/cinder/+/92629808:14
*** tobias-u1 is now known as tobias-urdin08:19
opendevreviewRajat Dhasmana proposed openstack/cinder master: RBD: Update volume location on failover  https://review.opendev.org/c/openstack/cinder/+/100730609:13
opendevreviewRaghavendra Tilay proposed openstack/cinder master: DNM - 2026_10_07 - please ignore  https://review.opendev.org/c/openstack/cinder/+/100917109:18
opendevreviewRajat Dhasmana proposed openstack/cinder master: RBD: Update volume location on failover  https://review.opendev.org/c/openstack/cinder/+/100730609:20
opendevreviewOpenStack Proposal Bot proposed openstack/cinder master: Imported Translations from Zanata  https://review.opendev.org/c/openstack/cinder/+/100739912:17
rosmaitaCinder meeting in #openstack-cinder at 1400 UTC13:56
rosmaitajungleboyj 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 13:56
Jyotsnao/13:59
simondodsleyo/14:00
eduardomorais[m]\o14:00
rosmaitawait, let me start the meeting14:00
rosmaita#startmeeting cinder14:00
opendevmeetMeeting started Wed Oct  7 14:00:53 2026 UTC and is due to finish in 60 minutes.  The chair is rosmaita. Information about MeetBot at http://wiki.debian.org/MeetBot.14:00
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.14:00
opendevmeetThe meeting name has been set to 'cinder'14:00
rosmaita#topic roll call14:01
agalica_o/14:01
fqio/14:01
harsho/14:01
rosmaita(glad to see people are so excited to get the meeting started)14:01
hvlcchao1o/14:01
eduardomorais[m]\o14:01
sp-bmilanovo/14:01
Jyotsna0/14:02
rosmaitahello everyone14:03
rosmaitanot much on the agenda today except for the PTG14:03
rosmaitabut there is quite a bit to discuss there14:03
rosmaita#topic announcements14:03
rosmaitaIndri Project Team Gathering is next week, Oct 12-1614:03
rosmaitago to https://openinfra.org/ptg/ and hit the "Register Now!" button14:03
Anoop_ShuklaO/14:04
rosmaitayou will get a free ticket, that you don't really need for anything14:04
rosmaitai think the point of registering is so that you read and agree to the code of conduct14:04
rosmaitalink to that is at the bottom of the registration page14:04
rosmaitabut i don't think that you are required to say that you read the code of conduct to get your ticket14:05
rosmaitaanyway, the code of conduct is basically don't be a jerk during the PTG14:05
rosmaitaalso, the foundation likes to know how many people are participating14:05
rosmaitaso, please register14:06
simondodsley(so we can plan for in-person in the future hopefully)14:06
rosmaitawould definitely be nice to have in-person PTGs again!14:06
rosmaitaok, that's all for announcements14:07
simondodsleythe more people that say this to the Foundation the better14:07
rosmaitasimondodsley: that is a good point, there is usually a feedback etherpad for general PTG comments14:07
rosmaitamaybe you can remind us at some point during the week to leave a comment about in-person meeting on it14:08
agalica_So I can then post my dissent, haha14:08
rosmaitaagalica_: that is your right14:08
simondodsleymisery14:09
rosmaitasimondodsley: this would be a good time for you to explain why people in the cinder community are referred to as "Argonauts"14:09
rosmaitai wonder how many people here know14:10
simondodsleylol - yes14:10
simondodsleybut i think i forgot... let me try and remember as it was my idea in the first place i think14:11
rosmaitait was!14:11
rosmaitai can explain14:11
rosmaitaif you look on our launchpad page, you can see the cinder mascot: https://launchpad.net/cinder14:12
simondodsleyoh - i remember14:12
agalica_Simon has been away from contact for so long that his brain has begun to calcify!!! Argh!!!14:12
simondodsleyIt's to do with the OpenStack release Xena14:12
rosmaitawhich is a horse presenting his backend to the viewer, as a reminder that the backends are really important for the block storage service14:12
simondodsleyXena's horse awas called Argo14:12
sp-bmilanov:D14:13
agalica_lol14:13
rosmaitathat's right, we had a vote at the xena ptg to name the mascot, and "Argo" won14:13
rosmaitaand as a bonus, people in the cinder community can be called "Argonauts"14:13
simondodsleyAll we need is a PTL called Jason14:14
rosmaitai don't remember the other names proposed, but the "Argonauts" connection seemed to be a deciding factor14:14
rosmaitawe definitely need a PTL named Jason!14:14
simondodsleyit was 5 years ago - hence why I had swapped that part out14:15
rosmaitayeah, 10 development cycles ago is a long time14:15
rosmaitaok, now you are all clued in to the Argonauts references14:16
rosmaita#topic PTG organization14:16
rosmaitaok, so most people will have this channel open in IRC during the PTG, so you can probably find people here14:17
rosmaitabut the general PTG channel is #openinfra-events14:17
rosmaitaprojects will announce what discussions are happening, stuff like that14:17
rosmaitaso it's good to keep that one open, too14:18
rosmaitathe schedule for the entire thing, list of etherpads, etc is at14:18
rosmaita#link https://ptg.opendev.org/14:18
Anoop_ShuklaI have a question on the security side of things - We flagged some drivers security stand during Hibiscus. There are some gaps that were also brought up - CHAP authentication be default. Also many drivers do not support DH-HMAC-CHAP for NVMe. In PTG, can we decide what is our stand on these two open items?14:19
rosmaitacinder will meet in the "ocata" room from tuesday through friday14:19
rosmaitaAnoop_Shukla: yes, that is a good topic14:19
simondodsley@Anoop_Shukla: that will be part of the ironic cross-project I suspect14:19
rosmaita#link https://etherpad.opendev.org/p/cinder-indri-ptg-planning14:20
rosmaitaok, so as far as what times we are meeting ...14:20
rosmaita#link https://etherpad.opendev.org/p/cinder-ptg-timeslots-fall14:20
rosmaitaso if you open that page, you can see what our options are14:21
Anoop_ShuklaThanks @simondodsley - missed that part of the etherpad.14:21
Anoop_ShuklaBut I did not get - why it’s part of ironic-cinder discussion and not on cinder itself? Do we have a dependency?14:22
simondodsleyprobbaly not - but whatever ironic need to do will affect os-brick which we control14:22
simondodsleyif we decided to do HMAC before TLS and they went the other way, that would be a problem14:23
Anoop_ShuklaOkay got it14:23
rosmaitaso, that is a good point about scheduling, namely that we have to schedule some sessions with other teams14:24
rosmaitaif you look on the etherpad linked earlier, line 3614:24
rosmaitaboth ironic and manila want to meet in the 1700-1800UTC time slot14:25
rosmaitathose are tuesday and thursday14:25
rosmaitaso to keep the day to 4 hours, i think we should start at 1400UTC on tuesday and thursday14:25
rosmaitawhich is good for agalica_ but maybe not for people on the other end of the time zone spread14:26
simondodsleyi won't be able to attend on Tuesday, so anything I have suggested needs to be Thursday14:26
rosmaitaso the question here is what to do about wednesday and friday, start at 1300 or 1400 ?14:26
simondodsleystick with the same time - 140014:27
auniyal2it would be nice to be done by 16 UTC14:27
agalica_I vote 2300 across the board14:27
auniyal2so I prefer to start as early as possible - 1314:28
auniyal2in case it get streched 14:28
simondodsleytbh i have no preference - i ws just trying to suggest consistency14:28
agalica_I’m picturing auniyal2 getting stretched on one of those medieval torture devices now14:29
auniyal2:D 14:29
rosmaitaif we meet > 4 hours on any one day, you don't need to imagine a special kind of torture device14:29
Anoop_ShuklaCan we keep the same time as IRC meetings this works for the other side of hemisphere more or less as well14:30
rosmaitaAnoop_Shukla: that's a good point, so that would be 1400 UTC14:30
auniyal2yeah, 1400 is fine if all agree14:31
agalica_1400 pacific is fine by me14:31
agalica_:D14:32
rosmaitaagalica_: you are not helping14:32
rosmaitaok, let's reserve the 1400 through 1700 blocks (that is, ending at 1800 UTC) for tuesday-friday14:33
rosmaitai will try to schedule only 3 hours on wednesday and friday14:33
agalica_Will Friday be “open discussion” like normal, or is it a packed ptg?14:34
rosmaitashould be pretty open, plus it's a Festival of Reviews day14:34
agalica_Right14:34
rosmaitaok, thanks14:35
rosmaitaok, next question is where to meet14:35
rosmaitai am inclined to stick with google meet but am willing to entertain other suggestions14:36
rosmaitaexcept from agalica_14:36
agalica_Lol14:36
rosmaita:D14:36
simondodsleymessaging in google meet is blocked for me14:36
rosmaitajust the messaging?14:36
simondodsleydon't ask why - it makes zero sense14:36
simondodsleyyep14:36
rosmaitai was about to ask why14:37
rosmaitaok14:37
rosmaitawell, we tend to take notes in the etherpad14:37
simondodsleytrue14:37
rosmaitaand we have IRC for messaging14:37
simondodsleyin that case i'm good with google14:37
rosmaitai will see if i can turn off messaging on my end so that people don't start doing it by mistake14:37
rosmaitai believe that all current contributors have can access google meet?14:38
rosmaitamy primary reason for using it is that the recording works really well14:38
simondodsleywill we be allowed to use AI transcription14:39
simondodsleysorry, SI14:39
Anoop_Shuklagoogle meet works for us at netapp14:39
Jyotsnait works for us as well 14:39
agalica_Google meet works for me, preferred in fact since it’s already set up14:39
rosmaitaI guess we can use the transcription feature if that is useful14:40
simondodsleymight be for for historical purposes14:40
rosmaitai guess it's not that different from having a recording14:41
Anoop_ShuklaThe only crib I have with google meet is, sharing is pretty complex..14:42
Anoop_ShuklaSharing screen*14:42
rosmaitai guess i haven't noticed that14:43
rosmaitai think, let's stick with google meet for this PTG, and if anyone has an alternative, maybe we could test it out on Friday at the end of the ptg14:45
rosmaitaand then use the alternative in the future14:45
rosmaitaok, guess we've got it all covered14:47
rosmaita#topic open discussion14:47
Anoop_Shukla@simondodsley: We were working on boot volume backup proposal some time back. Do you think it’s of importance to take up this ptg?14:47
simondodsleythere is a lot planned for the PTG, so we could add it, but at a lower priority14:48
rosmaitathere is a high-priority change that removes usage of sqlalchemy-utils (which we only use for one particular unit test class)14:48
Anoop_ShuklaYeah I agree.14:48
rosmaita#link https://review.opendev.org/c/openstack/cinder/+/100902914:48
simondodsleyI'll take a look14:48
JyotsnaWe have few patches on priority which requires immediate attention from reviewers  1) https://review.opendev.org/c/openstack/cinder/+/1008049 - Skip Online clone synchronous wait for copy, agalica,auniyal and jayaanand already review this patch and rework submitted. Required Re-review and +2 2) https://review.opendev.org/c/openstack/cinder/+/1001152 - Improved failover/failback with Active PP support - jayaanand already 14:48
rosmaitaanyway, if you want to dig into some esoteric sqlalchemy usage, you will find https://review.opendev.org/c/openstack/cinder/+/1009029 to be of interest14:49
eharneywill take a look there14:49
Jyotsna3) https://review.opendev.org/c/openstack/cinder/+/1001281 -Improved host proximity for ActivePP support +1 already and requires +2 from core reviewers14:49
Jyotsnacan we include them in festival of reviews - Friday meeting 14:50
simondodsleyJyotsna: there is no rush now - we are at the start of the Indri cycle - these missed Hibiscus14:50
rosmaitathe other item already on today's agenda is "General advice and strategy on deprecating an option"14:50
Anoop_ShuklaWe have a high priority patch we are looking to get reviewed for ASAr2 support: https://review.opendev.org/c/openstack/cinder/+/97117414:50
rosmaita#link https://review.opendev.org/c/openstack/cinder/+/973595/14:50
sp-bmilanovHey, that's me -- I am not looking for any precise reviews yet, I just want to confirm that is generally the approach to carry on with14:51
sp-bmilanovin order to skip radical refactoring back-and-forth once it has been fleshed out14:52
simondodsleyi think this is a good PTG topic14:52
Jyotsna@simondodsley : 1) https://review.opendev.org/c/openstack/cinder/+/1008049 - Skip Online clone synchronous wait for copy... this patch is more interest to customers using concurrent deployment so we need this on priority as soon as indri opens 14:52
rosmaitasp-bmilanov: you want to completely remove the option, not rename it or something?14:53
sp-bmilanovsimondodsley:  hmm.. ok, I will have a look a look how to plug into a PTG slot.. I didn't think it warrants that much attention14:53
simondodsleyi think it would be good to have a strategy for futrure reference14:53
sp-bmilanovrosmaita: yes, the option is wrong in general, and there is an added issue that the Cinder scheduler schedules based on that14:54
sp-bmilanovit should have been at least 'storpool:template'14:54
Anoop_ShuklaWhat happens to upgrades when an option is removed? Shouldn’t it be deprecated and users should get time to migrate to an alternative if it exists?14:54
sp-bmilanovif we need to do the offical dance, there would be two phases -- one with the change that handles both in release X14:55
rosmaitaright, the usual practice is deprecate in a .1 release, and remove in the next release14:55
sp-bmilanovphase 2 -- remove the option in release Y14:55
sp-bmilanovnot actual X and Y, that would be toooooo far in the future :D14:55
sp-bmilanovyeah, what rosmaita said 14:55
sp-bmilanovI might also add the related idea of removing pool support from a driver as well14:56
rosmaitawell, the slight complication is that with SLURP releases, operators may only upgrade from 2026.1 to 2027.114:56
Anoop_ShuklaWe had a similar option introduced for qos in Manila recently..we took the deprecation path which works. We notified the deprecation and in a future release will remove the older options in support of new..14:56
rosmaitaso a deprecation in a .2 release might not be seen14:57
rosmaitawhat Anoop_Shukla said ... plus there are some oslo.config features that may help14:57
sp-bmilanovthat sounds reasonable.. would it be possible to squeeze all changes in if we migrate our users beforehand?14:58
sp-bmilanovsounds reasonable = I mean introducing the deprecation in a .1 and then removing the option in the next .114:59
JyotsnaOne more query on openstack-ansible, can anyone confirm if this repo is still actively used and are vendors allowed to push changes ? 14:59
rosmaitawell, you can remove the option in the next .2 ... you just can't deprecate in .214:59
sp-bmilanovhuh, that's confusing .. isn't deprecated = pls don't use? and remove = it does not exist15:00
sp-bmilanovhow would removal then precede deprecation?15:00
rosmaitayes, but if people are only using .1 releases, they won't see a deprecation notice from .215:01
simondodsleyJyotsna: yes - it is still active - i have recently had changes commited there. the process is a little slow as the governance of that project is thin15:01
rosmaitaso the idea is all deprecations have to be announced in .115:01
Jyotsna<simondodsley> any guidelines to submit patches? also is it supported in RHOSP as well?15:02
rosmaitaso, deprecate in 2027.1, remove in 2027.2, and that also removes it from 2028.115:02
rosmaitanot sure if that made sense -- we can continue to discuss after the meeting15:02
sp-bmilanovah, right, no, I meant that it sounds reasonable to deprecate in 2027.1 and remove in 2028.115:02
rosmaitaok, thanks everyone for attending15:03
rosmaitasee you next week at 1400 utc on Tuesday15:03
rosmaita#endmeeting15:03
opendevmeetMeeting ended Wed Oct  7 15:03:19 2026 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)15:03
opendevmeetMinutes:        https://meetings.opendev.org/meetings/cinder/2026/cinder.2026-10-07-14.00.html15:03
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/cinder/2026/cinder.2026-10-07-14.00.txt15:03
opendevmeetLog:            https://meetings.opendev.org/meetings/cinder/2026/cinder.2026-10-07-14.00.log.html15:03
simondodsleyJyotsna: the usual way through gerrit. RHOSP uses tripleo for deployment and is EOL anyway15:03
sp-bmilanovthanks rosmaita, I think I got it15:03
sp-bmilanovSimon! Are you still around? :)15:04
sp-bmilanovI am not sure if you read my messages here from earlier, but I was wondering why you chose Github for the Pure CI logs15:05
sp-bmilanovwe have to get the logs out in some way and I was looking for options/opinions15:05
sp-bmilanov(the logs out of the StorPool CI)15:06
opendevreviewEric Harney proposed openstack/cinder master: Tests: Be more specific about request asserts in Hitachi tests  https://review.opendev.org/c/openstack/cinder/+/100928615:14
opendevreviewSimon Dodsley proposed openstack/cinder master: [Everpure] Keep volume group QoS and volume QoS apart  https://review.opendev.org/c/openstack/cinder/+/100929115:34
opendevreviewSimon Dodsley proposed openstack/cinder master: [Everpure] Separate volume group and volume QoS  https://review.opendev.org/c/openstack/cinder/+/100929115:39
opendevreviewEric Harney proposed openstack/cinder master: Bump mypy to 2.4.x  https://review.opendev.org/c/openstack/cinder/+/100930816:34
opendevreviewPratik Shah proposed openstack/cinder master: [Hitachi] Add volume group replication  https://review.opendev.org/c/openstack/cinder/+/100853416:42
opendevreviewBrian Rosmaita proposed openstack/cinder master: [doc-only] Fix typo in release note  https://review.opendev.org/c/openstack/cinder/+/100931617:07
harshsimondodsley: Hi Simon, if you are around could you please review the comments I resolved for https://review.opendev.org/c/openstack/cinder/+/98261717:35
harshThere was a +2 already so hoping to get this merged as we have quite a few dependent patches in a chain :D 17:36
opendevreviewTakashi Kajinami proposed openstack/python-cinderclient master: Update jobs according to tested runtimes for 2027.1  https://review.opendev.org/c/openstack/python-cinderclient/+/100766817:36
opendevreviewSimon Dodsley proposed openstack/cinder stable/2025.1: Pure Storage: fix revert_to_snapshot argument order  https://review.opendev.org/c/openstack/cinder/+/100933418:08
simondodsleyharsh: done18:15
opendevreviewWalt proposed openstack/cinder master: Implement graceful shutdown for Cinder services  https://review.opendev.org/c/openstack/cinder/+/98983618:34
harshsimondodsley: thank you for the quick review and the +1. I will resolve the nits and publish a new PS in a few minutes.18:49
opendevreviewHarsh Ailani proposed openstack/cinder master: [IBM SVf Driver] Added support of clone operation for volumegroup  https://review.opendev.org/c/openstack/cinder/+/98261719:44
harshsimondodsley: it's done :) 19:46
harshagalica: i'm really sorry but you would have to put the +2 again :D 19:46
opendevreviewDamian Dąbrowski proposed openstack/cinder master: Add backup_default_availability_zone config opt  https://review.opendev.org/c/openstack/cinder/+/97813520:20
opendevreviewSimon Dodsley proposed openstack/cinder master: Everpure: check the resize when cloning to a larger size  https://review.opendev.org/c/openstack/cinder/+/100936320:55
opendevreviewSimon Dodsley proposed openstack/cinder master: Everpure: keep the pod of a managed pod volume in provider_id  https://review.opendev.org/c/openstack/cinder/+/100936420:55
opendevreviewSimon Dodsley proposed openstack/cinder master: Everpure: use the right type for a failover volume group  https://review.opendev.org/c/openstack/cinder/+/100936520:55
opendevreviewSimon Dodsley proposed openstack/cinder master: Everpure: bound the wait for a ghost pod to stretch  https://review.opendev.org/c/openstack/cinder/+/100936620:55
agalicaharsh: done21:35
*** priteau_ is now known as priteau21:38
opendevreviewSimon Dodsley proposed openstack/cinder-specs master: Add spec for FlashArray pods as Cinder pools  https://review.opendev.org/c/openstack/cinder-specs/+/100938221:52
opendevreviewWalt proposed openstack/cinder master: Implement graceful shutdown for Cinder services  https://review.opendev.org/c/openstack/cinder/+/98983622:53

Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!