| *** mhen_ is now known as mhen | 02:03 | |
| jbernard | #startmeeting cinder | 14:01 |
|---|---|---|
| opendevmeet | Meeting started Wed Jul 9 14:01:17 2025 UTC and is due to finish in 60 minutes. The chair is jbernard. 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 |
| jbernard | #topic roll call | 14:01 |
| jbernard | o/ | 14:01 |
| Sai | o/ | 14:01 |
| harsh | o/ | 14:01 |
| Luzi | o/ | 14:01 |
| rosmaita | o/ | 14:01 |
| hvlcchao1 | o/ | 14:02 |
| gireesh | o/ | 14:03 |
| jbernard | welcome everyone | 14:05 |
| jbernard | #topic annoucements | 14:05 |
| tosky_ | o/ | 14:05 |
| hvlcchao1 | o/ | 14:05 |
| jbernard | membership (openinfra) renewal | 14:05 |
| jbernard | #link https://openinfra.org/a/renew-membership | 14:06 |
| jbernard | this seems to be important | 14:06 |
| harsh | done :) | 14:06 |
| jbernard | there was a mail earlier this week, deadline is in about 10 days | 14:06 |
| rosmaita | i couldn't find the email, but it was easy to renew with the link | 14:06 |
| jbernard | mail subject is "Action Required: Renew Your OpenInfra Individual Membership by July 19" | 14:07 |
| rosmaita | word on the street is that if your foundation membership had already lapsed, you won't get an email | 14:07 |
| jbernard | ahhj | 14:07 |
| jbernard | so in summary, fixup your memberships :) | 14:08 |
| rosmaita | :D | 14:08 |
| jbernard | #topic followup | 14:09 |
| jbernard | simondodsley asked about missing sdk commands in teh last meeting | 14:09 |
| Vivek__ | o/ | 14:09 |
| jbernard | #link https://etherpad.opendev.org/p/cinder-openstacksdk-gaps | 14:09 |
| jbernard | i think rajat would probably know best but I haven't gotten any feedback yet | 14:10 |
| * jungleboyj sneaks in late | 14:10 | |
| jbernard | rosmaita: ^ any insights that i may have missed? | 14:10 |
| rosmaita | not really, what was the source of that etherpad? | 14:11 |
| jbernard | #link https://docs.google.com/spreadsheets/d/1yetPti2XImRnOXvvJH48yQogdKDzRWDl-NYbjTF67Gg/edit?pli=1&gid=1463660729#gid=1463660729 | 14:11 |
| jbernard | rosmaita: i thinkm simon created it, listing the ones he noticed were missing | 14:11 |
| rosmaita | ok, just curious | 14:12 |
| jbernard | i haven't had any time to look at the sdk/cli recently | 14:12 |
| jbernard | in summary, this is still something i need to followup on | 14:12 |
| jbernard | #topic image encryption tempest tests | 14:14 |
| jbernard | #link https://review.opendev.org/c/openstack/barbican-tempest-plugin/+/952699 | 14:14 |
| jbernard | this is very nice to have | 14:14 |
| Luzi | mhen finished his work on this | 14:15 |
| jbernard | we would like to land the image encryption patch | 14:15 |
| jbernard | #link https://review.opendev.org/c/openstack/cinder/+/926298 | 14:15 |
| jbernard | i was able to test it sucessfully locally | 14:15 |
| jbernard | but I think now others can review without having to do that, as we now have coverage in the gate | 14:16 |
| jbernard | Luzi: am I accurate about that^ ? | 14:16 |
| Luzi | am i don't know about patches that depend on each other | 14:17 |
| Luzi | i mean the tempest patch clearly depends on the cinder patch, the cinder patch currently does not depend on the tempest patch | 14:17 |
| Luzi | is that correct? | 14:18 |
| Luzi | btw: the glance patch is needed for this too | 14:18 |
| jbernard | that's what im seeing also | 14:18 |
| jbernard | basically, we'd like to get this code in sooner as opposed to later in the cycle, so please review if you have the time | 14:19 |
| jbernard | Luzi: what is the glance status? | 14:19 |
| Luzi | https://review.opendev.org/c/openstack/glance/+/926295 there is 2x +1 | 14:20 |
| jbernard | #topic ibm driver phone-home | 14:22 |
| jbernard | #link https://review.opendev.org/c/openstack/cinder/+/951829 | 14:22 |
| jbernard | im not sure who added this | 14:22 |
| Vivek__ | hi jbernard | 14:22 |
| Vivek__ | This is vivek, Im part of IBM SVf cinder driver team. I have recently started contributing to the community. | 14:23 |
| jbernard | Vivek__: welcome | 14:23 |
| Vivek__ | One of my team mate uploaded the patch | 14:23 |
| Vivek__ | thanks | 14:23 |
| jbernard | I took a look at this recently, and i think brian did as well | 14:24 |
| rosmaita | yep | 14:24 |
| Vivek__ | We can discuss on setting the callhome as disabled | 14:24 |
| jbernard | im not personally a fan of phone-home, but the enabled-by-default is difficult for me | 14:24 |
| Vivek__ | We are enabling Callhome support since we are not capturing any sensitive data. The metadata collected includes the OS version, Cinder driver version, deployment type, PBHA status, and replication feature usage. | 14:25 |
| rosmaita | It's good that it's not sensitive, but i think it should be "off" by default | 14:26 |
| Vivek__ | This information helps us understand which features are actively being used by customers, allowing us to better prioritize future enhancements for our driver. | 14:26 |
| rosmaita | Yes, and i imagine that most of your customers will decide to opt-in | 14:27 |
| rosmaita | but it would be good to give them a choice | 14:27 |
| jungleboyj | rosmaita: ++ | 14:27 |
| rosmaita | are you thinking that if you ship with it "off", no one will bother to turn it on, and hence you won't get any data? | 14:28 |
| Vivek__ | Yes, we have introduced a toggle switch in conf to turn it "off". | 14:28 |
| jbernard | one issue is that the toggle as written has no discoverability, you have to read the code to know how to do that | 14:29 |
| rosmaita | yes, i talked about that in my comment also ... this needs to be defined like a "real" option | 14:30 |
| harsh | Hi.. i am Vivek's colleague. his connection seems to have disconnected. | 14:31 |
| harsh | rosmaita : yes. thats why we want to ship it with enabled. so that if a customer is unsure about data sharing, can turn it off. | 14:31 |
| harsh | jbernard : yes, as there is no readibility except from the code, we want to ship it as enabled and update our driver document : https://docs.openstack.org/cinder/latest/configuration/block-storage/drivers/ibm-storwize-svc-driver.html | 14:31 |
| rosmaita | harsh: i think it needs to be documented like any other option | 14:32 |
| harsh | correct. we will add it in the documents. | 14:32 |
| rosmaita | not just the documentation, it should show up in the sample config file | 14:33 |
| jungleboyj | ++ | 14:33 |
| rosmaita | see my comment from yesterday on the patch | 14:33 |
| harsh | we dont want to miss out on the customer info to ma | 14:33 |
| jbernard | i am personally against on-by-default, i think packagers/vendors can set it differently, but the upstream/opensource version should not default to enabled | 14:33 |
| rosmaita | personally, i do not like the opt-out model ... i suggest that you ship with it off by default, but all your sample docs can show turning it on | 14:34 |
| harsh | to enhance the driver as per the features that are actually being utiliised by the user. | 14:34 |
| Vivek__ | sorry my chat session got disconnected due to some network issue. | 14:35 |
| harsh | okay.. thanks for your suggestions rosmaita and jbernard . I will take your inputs to my team | 14:35 |
| rosmaita | i guess a key issue here is, do we have any other examples of a vendor call-home feature in openstack that is shipped "on" by default | 14:35 |
| harsh | We don't have that info as of now. anyone else has any idea ? | 14:36 |
| jungleboyj | I am not aware of any. | 14:36 |
| rosmaita | i am not aware of anything in cinder, but neutron and manila have third party backends too | 14:36 |
| rosmaita | i'm mentioning it because i think without a precedent, it will be a difficult sales job to the community to have this thing "on" by default | 14:37 |
| harsh | i understand | 14:38 |
| rosmaita | ok | 14:38 |
| harsh | we try to find out if there is any precedent.. | 14:38 |
| Vivek__ | Netapp have activeIQ and Dell have Unisphere / Telemetry, we may need to check how they are enabling it | 14:39 |
| rosmaita | that would be good info to have | 14:39 |
| rosmaita | anyone here today from netapp or dell? | 14:39 |
| Anoop_Shukla | Yes. NetApp telemetry is pushed to ONTAP via EMS and when we collect ASUP data, traces of telemetry can be mined out. | 14:42 |
| Anoop_Shukla | ASUP - AutoSupport | 14:43 |
| Vivek__ | So support for NetApp telemetry is set to enabled or disabled while driver init ? | 14:44 |
| jbernard | #link https://github.com/openstack/cinder/blob/master/cinder/volume/drivers/netapp/utils.py#L570 | 14:44 |
| Anoop_Shukla | As far as I know, customers can toggle ASUP on ONTAP storage or be on dark sites. Cinder/Manila would always post EMS to ONTAP. | 14:46 |
| harsh | So cinder is on-by-default as i understand.. | 14:47 |
| Vivek__ | Similarly we are setting it to "on" in order we can capture the metadata and derive the feature details. | 14:47 |
| jbernard | ok, i at least need some time to look at netapp and think on things, lets revisit this in the next meeting. we may need to take a vote, or elevate this to the TC for input, more thought is needed | 14:51 |
| jbernard | i do understand the argument | 14:52 |
| jbernard | #action sdk gaps | 14:52 |
| harsh | yes sure.. thankyou :) we can have a followup in the next meeting. | 14:52 |
| jbernard | #action call home defaults followup | 14:52 |
| Vivek__ | thanks | 14:53 |
| jbernard | #topic powerscale driver rename in manilla | 14:53 |
| jbernard | #link https://review.opendev.org/c/openstack/manila/+/952736 | 14:53 |
| jbernard | im not clear on how this relates to cinder, can someone take this? | 14:54 |
| jbernard | i suspect a similar rename is needed in cinder | 14:55 |
| jbernard | which shouldn't be a problem | 14:55 |
| jbernard | so this may not need discussion | 14:55 |
| rosmaita | hopefully not, these renames are a real PITA | 14:56 |
| gireesh | https://www.irccloud.com/pastebin/AuzWFywF/ | 14:58 |
| gireesh | There are couple of review pending for NetApp driver please have a look into those | 14:59 |
| gireesh | https://www.irccloud.com/pastebin/Fov4BdmM/ | 14:59 |
| jbernard | we're nearly at time, last call | 15:00 |
| jbernard | ok, thanks everyone, i know there are reviews waiting, we've been getting to some of them, still many to go but we're makign progress, appreciate the patience | 15:01 |
| jbernard | #endmeeting | 15:01 |
| opendevmeet | Meeting ended Wed Jul 9 15:01:28 2025 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 15:01 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/cinder/2025/cinder.2025-07-09-14.01.html | 15:01 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/cinder/2025/cinder.2025-07-09-14.01.txt | 15:01 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/cinder/2025/cinder.2025-07-09-14.01.log.html | 15:01 |
| jayaanand_ | also https://review.opendev.org/c/openstack/cinder/+/944964 | 15:01 |
| jungleboyj | Thanks! | 15:01 |
| jayaanand_ | https://review.opendev.org/c/openstack/cinder/+/928486 | 15:01 |
| jayaanand_ | thank you! | 15:01 |
| harsh | thankyou :) | 15:02 |
| jbernard | jayaanand_: fix zuul on the second one | 15:02 |
| jayaanand_ | sure... thank you! | 15:03 |
| jbernard | jayaanand_: if a reviewer looks at your patch and it's failing zuul, they're likely to move to another one | 15:03 |
| jayaanand_ | ok | 15:03 |
| *** vhari_ is now known as vhari | 19:03 | |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!