| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-cyborg - 2026-07-21 | |||
| 14:02:51 | jgilaber | let's start | |
| 14:03:11 | jgilaber | jumping directly to reviews since the agenda is empty | |
| 14:03:16 | jgilaber | #topic Reviews | |
| 14:03:16 | rlandy | o/ | |
| 14:03:25 | jgilaber | first reviews are from chandankumar | |
| 14:03:36 | jgilaber | #link https://review.opendev.org/q/topic:%22fpga_program%22 | |
| 14:03:39 | chandankumar | yes | |
| 14:03:44 | jgilaber | anything you want to highlight? | |
| 14:04:06 | chandankumar | I was working on improving the tempest test coverage and fixes around fgpa program interface | |
| 14:04:29 | chandankumar | I have proposed couple of reviews for rename and refactoring | |
| 14:04:38 | chandankumar | Use GenericDriver.update() as the FPGA programming interface - https://review.opendev.org/c/openstack/cyborg/+/991598 | |
| 14:04:52 | chandankumar | https://review.opendev.org/c/openstack/cyborg/+/998004: Move create() from FPGADriver to GenericDriver | |
| 14:05:06 | chandankumar | both reviews are tested with https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/991081 | |
| 14:05:14 | chandankumar | scenario test for FPGA programming with FakeDriver | |
| 14:05:24 | chandankumar | It boots a server with an | |
| 14:05:24 | chandankumar | FPGA device profile backed by FakeDriver, locates the fake deployable, and calls the program API to verify that it successed. | |
| 14:06:13 | chandankumar | Feel free to have a look at these reviews. | |
| 14:06:29 | jgilaber | ack, I had reviewed the first one in the chain, but not the other two, I'll add them to my review list | |
| 14:06:40 | chandankumar | thank you jgilaber! | |
| 14:06:48 | chandankumar | Next one is also mine | |
| 14:06:59 | chandankumar | #link https://review.opendev.org/q/topic:%22bug/2157586%22 | |
| 14:07:12 | chandankumar | few days back I opened https://bugs.launchpad.net/openstack-cyborg/+bug/2157586 | |
| 14:07:31 | chandankumar | Cyborg Device Profile API rejects OWNER_CYBORG trait because it rejects non custom traits | |
| 14:08:23 | chandankumar | https://review.opendev.org/c/openstack/cyborg/+/997945 removes the conditional for non custom traits and allow owner_cyborg traits to b used with device profole | |
| 14:08:55 | jgilaber | that bug only affects traits but not resource classes right? | |
| 14:08:59 | chandankumar | I have also added scenario test to validate cyborg placement integration https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/992210 | |
| 14:09:26 | chandankumar | Based on each device profile, it troes to find out triats based on vendor and product id | |
| 14:10:21 | chandankumar | yes only traits | |
| 14:10:51 | jgilaber | ack, this seems like an important bug to fix, thanks for working on it | |
| 14:11:09 | chandankumar | that's it from my side on reviews | |
| 14:11:37 | jgilaber | thanks chandankumar | |
| 14:12:00 | jgilaber | any other question on these reviews? | |
| 14:12:34 | jgilaber | moving on | |
| 14:12:49 | jgilaber | I have a couple of patches to hightligth | |
| 14:12:55 | jgilaber | #link https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/997751 | |
| 14:13:33 | jgilaber | This first one tries to improve the problems with the guest reboot test | |
| 14:14:57 | chandankumar | yes, thank you for working on this one. | |
| 14:15:01 | jgilaber | and the next one in the chain refactors the lifecycle tests | |
| 14:15:13 | jgilaber | so that they can be reused for multiple drivers | |
| 14:15:14 | chandankumar | Do we want to do more recheck on this one to make sure reboot test is not failing? | |
| 14:15:15 | jgilaber | #link https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/997936 | |
| 14:15:31 | jgilaber | yes, I want to recheck it some more | |
| 14:15:50 | jgilaber | and check if the failures in https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/997936 are because the first patch does not solve the problem completely | |
| 14:16:05 | jgilaber | or that I'm reintroducing the problem in the refactor somehow | |
| 14:16:14 | jgilaber | I'll double check that after the meeting | |
| 14:16:33 | jgilaber | any other question/comment on this? | |
| 14:16:39 | chandankumar | on second patch reboot is still failing, it is timing out when instance is coming back after reboot | |
| 14:16:55 | chandankumar | in temepst-py3 job. | |
| 14:17:22 | jgilaber | yes, I'm not sure if it's the exact same failure or is a new problem from the refactor | |
| 14:17:38 | chandankumar | I have not taken a look at the refactored code, will take a look tomorrow | |
| 14:17:55 | chandankumar | here py3 job is faoling while ipv6 is passing | |
| 14:18:04 | jgilaber | ack, thanks chandankumar | |
| 14:18:16 | jgilaber | yes, the error is not very consistent | |
| 14:18:32 | chandankumar | yes | |
| 14:18:43 | jgilaber | anything else on this? | |
| 14:18:53 | jgilaber | if not, I have one last review | |
| 14:19:02 | jgilaber | #link https://review.opendev.org/c/openstack/cyborg/+/997979 | |
| 14:19:37 | jgilaber | I noticed that in cyborg we had a test file to test database migrations with classes that were supposed to run them with both mariadb and postgres | |
| 14:19:48 | jgilaber | but both were skipped since the dbs were not setup | |
| 14:20:07 | jgilaber | and the checks for the migrations were not being called | |
| 14:20:33 | jgilaber | so I added test requirements for mariadb in this patch to at least run them there | |
| 14:20:59 | jgilaber | the proposal is to add mariadb for now and enable also postgres in the future if we want to | |
| 14:21:24 | jgilaber | any questions on this? | |
| 14:21:59 | chandankumar | I have added 998101: Drop postgres migration unit tests | https://review.opendev.org/c/openstack/cyborg/+/998101to drop prostes test | |
| 14:22:41 | chandankumar | Based on code search https://codesearch.openstack.org/?q=TestMigrationsPostgreSQL&i=nope&literal=nope&files=&excludeFiles=&repos= I did not find except cyborg and ironic having postgres coverage | |
| 14:23:23 | chandankumar | +1 for enabling it | |
| 14:24:05 | jgilaber | thanks, lgtm | |
| 14:24:22 | jgilaber | we did not enable them in watcher either | |
| 14:25:28 | chandankumar | ok | |
| 14:25:39 | jgilaber | anything else on this topic? | |
| 14:26:15 | jgilaber | moving to bugs | |
| 14:26:20 | jgilaber | #topic Bugs | |
| 14:26:29 | jgilaber | #link https://bugs.launchpad.net/openstack-cyborg/+bug/2161354 | |
| 14:27:36 | jgilaber | this seems like a valid bug and reading the report | |
| 14:27:55 | jgilaber | adding support for this in the pci driver might be better in the long term | |
| 14:28:26 | jgilaber | melwitt, do you know what would we need to change to support those devices? | |
| 14:30:34 | jgilaber | in the meantime, I think we can triage this as medium? | |
| 14:31:39 | chandankumar | yes medium sounds good | |
| 14:32:08 | jgilaber | ack, done | |
| 14:32:30 | jgilaber | that's the last thing we had in the agenda, any last minute topic? | |
| 14:33:21 | jgilaber | moving to the last topic then | |
| 14:33:42 | jgilaber | #topic Volunteers to chair next meeting | |
| 14:33:45 | jgilaber | any volunterr? | |
| 14:33:57 | chandankumar | I can volunteer for next meeting | |
| 14:34:31 | jgilaber | thanks chandankumar ++ | |
| 14:34:44 | jgilaber | that's all for today, thanks all! | |
| 14:34:48 | jgilaber | #endmeeting | |
| 14:34:48 | opendevmeet | Meeting ended Tue Jul 21 14:34:48 2026 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | |
| 14:34:48 | opendevmeet | Minutes: https://meetings.opendev.org/meetings/cyborg/2026/cyborg.2026-07-21-14.00.html | |
| 14:34:48 | opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/cyborg/2026/cyborg.2026-07-21-14.00.txt | |
| 14:34:48 | opendevmeet | Log: https://meetings.opendev.org/meetings/cyborg/2026/cyborg.2026-07-21-14.00.log.html | |
| 14:35:02 | chandankumar | thank you jgilaber ! | |
| 15:06:17 | melwitt | jgilaber_: hi sorry I had to step away in the middle of the meeting. on https://bugs.launchpad.net/openstack-cyborg/+bug/2161354, for the shorter term it would be a small change to leverage the same exclude mechanism in the NIC driver (duplication), | |
| 15:07:46 | melwitt | for the long term moving the NIC driver into the PCI driver would be a larger change with the main thing being we would need to move/add SRIOV VF enumeration into the PCI driver. along with device type awareness | |
| 15:19:44 | jgilaber | melwitt, no worries I was just curious | |
| 15:20:24 | jgilaber | the sriov is currently in progress, what kind of device type awareness would we need, do we need to do some setup on the device? | |
| 15:22:52 | melwitt | just to be able to differentiate between say a gpu vs a nic | |
| 15:23:36 | melwitt | so check the type to know whether to do the sriov VF enumeration for example | |
| 15:24:34 | melwitt | (sorry I am not so knowledgeable in the drivers yet, maybe that is already there) | |
| 15:42:14 | jgilaber | ah ok | |