| opendevreview | Merged openstack/cinder master: Storwize: stop tests spawning cinder-manage subprocesses https://review.opendev.org/c/openstack/cinder/+/994981 | 00:32 |
|---|---|---|
| opendevreview | Rajat Dhasmana proposed openstack/cinder master: NetApp: Fix terminate_connection crash on cleanup https://review.opendev.org/c/openstack/cinder/+/999786 | 06:24 |
| opendevreview | Rajat Dhasmana proposed openstack/cinder master: NFS: Fix race condition in capacity calculation https://review.opendev.org/c/openstack/cinder/+/999709 | 06:33 |
| opendevreview | Anoop Kumar Shukla proposed openstack/cinder master: NetApp: default HTTPS transport and TLS verify https://review.opendev.org/c/openstack/cinder/+/999310 | 10:18 |
| opendevreview | Eric Harney proposed openstack/cinder master: Run Cinder's Backup service in native threads https://review.opendev.org/c/openstack/cinder/+/977851 | 12:17 |
| opendevreview | Eric Harney proposed openstack/cinder master: Run Cinder's Backup service in native threads https://review.opendev.org/c/openstack/cinder/+/977851 | 12:22 |
| opendevreview | Anoop Kumar Shukla proposed openstack/cinder master: NetApp: enrich periodic Cinder EMS payloads https://review.opendev.org/c/openstack/cinder/+/996205 | 12:26 |
| opendevreview | Fernando Ferraz proposed openstack/cinder master: Run Ceph Backup driver in native threads https://review.opendev.org/c/openstack/cinder/+/978013 | 12:36 |
| opendevreview | Eric Harney proposed openstack/cinder master: mypy: Correct futurist.GreenThreadPoolExecutor name https://review.opendev.org/c/openstack/cinder/+/999821 | 12:59 |
| opendevreview | Rajat Dhasmana proposed openstack/cinder master: NFS: Fix race condition in capacity calculation https://review.opendev.org/c/openstack/cinder/+/999709 | 13:12 |
| mhen | no meeting today? | 14:07 |
| rosmaita | meeting cancelled, but people are around if you want to discuss something in irc | 14:08 |
| Jyotsna | HPE Alletra MP new driver patch https://review.opendev.org/c/openstack/cinder/+/969373 (Recieved +2 and workflow +1 and ready for merge) | 14:09 |
| Jyotsna | https://review.opendev.org/c/openstack/requirements/+/974322 dependent patch, we have address comment from zigo and responded to Erlon on his feedback. | 14:09 |
| Jyotsna | I request reviewers attention on providing +2 and merge approval on requirement patch | 14:09 |
| rosmaita | Jyotsna: looking | 14:10 |
| Jyotsna | Thanks Brian | 14:10 |
| Jyotsna | we need 2 +2 from reviewers | 14:10 |
| Sai | Hi, | 14:15 |
| Sai | Waiting for reviwes on below changes, got approval from Gireesh/Anoop/Jayaanand from NetApp. | 14:16 |
| Sai | https://review.opendev.org/c/openstack/manila/+/998581 | 14:16 |
| rosmaita | Jyotsna: i think you can release a new version of hpe-storage-flowkit-py and update the requirements patch | 14:16 |
| rosmaita | you shouldn't have to change anything in the cinder patch, because that has "hpe-storage-flowkit-py>=0.9.3", which the new version will satisfy | 14:16 |
| Sai | https://review.opendev.org/c/openstack/cinder/+/998810 | 14:17 |
| rosmaita | so i don't think you will lose any +2/+W on the driver patch | 14:17 |
| zigo | Jyotsna: Hi ! | 14:17 |
| Jyotsna | ok thanks Brian | 14:17 |
| Jyotsna | hi zigo | 14:17 |
| rosmaita | oh good zigo is here to tell you | 14:17 |
| zigo | Let me check. | 14:17 |
| Jyotsna | yes sure | 14:18 |
| zigo | To me, upstream's pyproject.toml is still wrong. | 14:18 |
| zigo | I can read: | 14:19 |
| zigo | [tool.setuptools] | 14:19 |
| zigo | include-package-data = true | 14:19 |
| zigo | packages = [ | 14:19 |
| zigo | "hpe_storage_flowkit_py", | 14:19 |
| zigo | "hpe_storage_flowkit_py.v1" | 14:19 |
| zigo | ... | 14:19 |
| zigo | There's no such folder as "hpe_storage_flowkit_py" in upstream repo. | 14:19 |
| zigo | Using upstream repo will lead to a broken python / debian package. | 14:19 |
| zigo | They are obviously not using what they've pushed in the repo to push stuff to pypi. | 14:19 |
| zigo | Wrote in https://github.com/hpe-storage/hpe_storage_flowkit_py/issues/1 to ask for a fix. | 14:21 |
| Jyotsna | we have build the pypi with same contents which is present in https://github.com/hpe-storage/hpe_storage_flowkit_py | 14:21 |
| Jyotsna | there is no outside dependency in building pypi | 14:21 |
| zigo | How come the pyproject.toml looks wrong then? | 14:21 |
| Jyotsna | hpe_storage_flowkit_py itself is root folder when we checkout and inside it we have structure v1/src/.... | 14:22 |
| Jyotsna | so the pyproject.toml is telling correct structure | 14:23 |
| zigo | As much as I know, the [tool.setuptools]/packages directive is relative to where the pyproject.toml is. | 14:23 |
| Jyotsna | ok | 14:24 |
| zigo | If it works... well it'd be the first time I see something like this ! :) | 14:24 |
| zigo | Also, you do not need to list packages whenever there's an __init__.py normally. | 14:25 |
| sailaja_ | Please review the bug fixes submitted for the NetApp Cinder driver- https://review.opendev.org/c/openstack/cinder/+/992427 & https://review.opendev.org/c/openstack/cinder/+/991534 | 14:26 |
| zigo | rosmaita: Could you review my 2 patches ? https://review.opendev.org/c/openstack/cinder/+/979641 https://review.opendev.org/c/openstack/cinder/+/971710 | 14:27 |
| zigo | First one is IMO an urgent fix that would deserve backports. | 14:27 |
| rosmaita | zigo: looking | 14:27 |
| rosmaita | zigo: they both look like important changes, i will raise the priority now, and try to review later today | 14:34 |
| Jyotsna | zigo : If we change the folder structure now, we have import changes in driver patch ? We have tested this throughly and it is been like this for driver for so long. Do you still see any concern ? | 14:34 |
| zigo | Jyotsna: I have a the concern that building a Debian package with this pyproject.toml wont work. | 14:37 |
| zigo | rosmaita: Thanks ! | 14:37 |
| zigo | The "improve heartbeat" patch was even more important in the Eventlet time, as the main thread could appear dead. With newer native threading, there's probably less risk that this happens, though still, I think it's a nice improvement to check the health of the main thread. | 14:38 |
| Jyotsna | zigo: can you please describe what exact issue or error you are facing? we can try and get back on this | 14:39 |
| zigo | Jyotsna: Issue: your Python packaging is wrong, please fix it. :) | 14:40 |
| zigo | Jyotsna: Usually, you'd have a structure like project_name/{v1,v2,tests,whatever}, and some __init__.py in each folder, then only a single packages = [ "project_name"]. | 14:41 |
| Jyotsna | zigo: when we install pypi package we get same structure | 14:41 |
| zigo | Then it will end up in /usr/lib/python3/dist-packages/project_name/{v1,v2,tests,whatever} correctly. | 14:41 |
| zigo | In such case, if you switch to what I'm suggesting, everything will continue to work then, no ? | 14:42 |
| Jyotsna | ls -l /usr/local/lib/python3.12/dist-packages/hpe_storage_flowkit_py/ total 32 -rw-r--r-- 1 root root 11358 Jul 28 03:37 LICENSE.txt -rw-r--r-- 1 root root 5081 Jul 28 03:37 README.md -rw-r--r-- 1 root root 32 Jul 28 03:37 requirements.txt drwxr-xr-x 4 root root 4096 Jul 28 03:37 v1 drwxr-xr-x 4 root root 4096 Jul 28 03:37 v3 | 14:45 |
| Jyotsna | can you check this ? | 14:45 |
| Jyotsna | this is folder listing after we install pypi package | 14:46 |
| zigo | I can try again to package it, but most likely : it will *not* work, and I'll have to patch pyproject.toml. | 14:46 |
| zigo | Got to go see a colleague, I'll be back in a few minutes. | 14:46 |
| Jyotsna | please give it try and let us know on the issue in repo | 14:47 |
| Jyotsna | Thanks for your feedback and support till now | 14:48 |
| Jyotsna | <jayaanand> Thanks for your feedback and patience in reviewing the patch till now | 14:48 |
| Leo[m] | Hello. I was hoping to discuss whether the Cinder community would be open to the idea of using hostname-based Ceph Mon references in the future. Currently, ceph mon dump is used (see here: https://github.com/openstack/cinder/blob/1736706bddccd2732374a2cee40ed7f922501112/cinder/volume/drivers/rbd.py#L656) to get the IP addresses but we would like to use hostnames instead to make it easier to replace hardware and migrate to IPv6. We | 14:49 |
| Leo[m] | have a loose commitment from the Ceph community to add a --hostname parameter to ceph mon dump but are open to other solutions such as reading the addresses from ceph.conf. Is this something that the community would be open to supporting? If so, what would my next steps be? Should I write a blueprint or spec? Thank you for attention | 14:49 |
| Leo[m] | * Hello. I was hoping to discuss whether the Cinder community would be open to the idea of using hostname-based Ceph Mon references in the future. Currently, ceph mon dump is used (see here: https://github.com/openstack/cinder/blob/1736706bddccd2732374a2cee40ed7f922501112/cinder/volume/drivers/rbd.py#L656) to get the IP addresses but we would like to use hostnames instead to make it easier to replace hardware and migrate to IPv6. We | 14:52 |
| Leo[m] | have a loose commitment from the Ceph community to add a --hostname parameter to ceph mon dump but are open to other solutions such as reading the addresses from ceph.conf. Is this something that the community would be open to supporting? If so, what would my next steps be? Should I write a blueprint or spec? Thank you for your attention | 14:52 |
| Leo[m] | I can also wait till next weekly meeting if need be | 14:59 |
| rosmaita | Leo[m]: if you could bring it up next week, that would be good ... off the top of my head, the ceph.conf solution seems better until the --hostname param is implemented for ceph mon dump | 16:04 |
| rosmaita | maybe you could put a brief proposal in an etherpad for discussion, and then you'll be able to use that when you need to write a spec | 16:04 |
| *** Unknown123 is now known as Mike-- | 16:28 | |
| opendevreview | Rajat Dhasmana proposed openstack/cinder master: NetApp: Fix terminate_connection crash on cleanup https://review.opendev.org/c/openstack/cinder/+/999786 | 17:44 |
| -opendevstatus- NOTICE: The Gerrit service on review.opendev.org will be offline momentarily at 20:00 UTC (an hour from now) while we restart for a patch upgrade, but should return within a few minutes. | 19:02 | |
| -opendevstatus- NOTICE: The Gerrit service on review.opendev.org will not be restarted at 20:00 UTC after all, due to unanticipated image build regressions, and will be rescheduled to a later time/date. | 19:13 | |
| opendevreview | Simon Dodsley proposed openstack/cinder master: LVM: add Changed Block Tracking backup support for thin volumes https://review.opendev.org/c/openstack/cinder/+/995738 | 22:00 |
| opendevreview | Simon Dodsley proposed openstack/cinder master: LVM: add Changed Block Tracking backup support for thin volumes https://review.opendev.org/c/openstack/cinder/+/995738 | 22:16 |
| opendevreview | Simon Dodsley proposed openstack/cinder master: Add Changed Block Tracking (CBT) backup framework and changed-blocks API https://review.opendev.org/c/openstack/cinder/+/990923 | 22:33 |
| opendevreview | Simon Dodsley proposed openstack/cinder master: Fix data-to-zero corruption restoring an incremental backup chain https://review.opendev.org/c/openstack/cinder/+/999911 | 22:33 |
| opendevreview | Simon Dodsley proposed openstack/cinder master: Add Changed Block Tracking (CBT) backup framework and changed-blocks API https://review.opendev.org/c/openstack/cinder/+/990923 | 22:38 |
| opendevreview | Simon Dodsley proposed openstack/cinder master: [Pure Storage] Implement Tiramisu group replication actions https://review.opendev.org/c/openstack/cinder/+/996497 | 23:58 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!