Tuesday, 2026-06-30

opendevreviewsean mooney proposed openstack/cyborg master: Use request context as policy target  https://review.opendev.org/c/openstack/cyborg/+/99272211:06
opendevreviewsean mooney proposed openstack/cyborg master: Add SRBAC base rules and enforce_new_defaults override  https://review.opendev.org/c/openstack/cyborg/+/99229211:06
opendevreviewsean mooney proposed openstack/cyborg master: Migrate ARQ policies to DocumentedRuleDefault  https://review.opendev.org/c/openstack/cyborg/+/99229311:06
opendevreviewsean mooney proposed openstack/cyborg master: Migrate device policies to DocumentedRuleDefault  https://review.opendev.org/c/openstack/cyborg/+/99229411:06
opendevreviewsean mooney proposed openstack/cyborg master: Migrate deployable policies to DocumentedRuleDefault  https://review.opendev.org/c/openstack/cyborg/+/99229511:06
opendevreviewsean mooney proposed openstack/cyborg master: Migrate attribute policies to DocumentedRuleDefault  https://review.opendev.org/c/openstack/cyborg/+/99229611:06
opendevreviewsean mooney proposed openstack/cyborg master: Add SRBAC policy concepts documentation  https://review.opendev.org/c/openstack/cyborg/+/99229711:06
opendevreviewsean mooney proposed openstack/cyborg master: Add cyborg-tempest-srbac Zuul job  https://review.opendev.org/c/openstack/cyborg/+/99233211:06
opendevreviewsean mooney proposed openstack/cyborg master: Remove no-op deprecated policy bridges  https://review.opendev.org/c/openstack/cyborg/+/99546211:06
opendevreviewMerged openstack/cyborg master: Fix incorrect set syntax in placement client logging  https://review.opendev.org/c/openstack/cyborg/+/99485112:03
opendevreviewMerged openstack/cyborg master: Rename devstack CYBORG_CONF_FILE to CYBORG_CONF for consistency  https://review.opendev.org/c/openstack/cyborg/+/97980512:03
chandankumarsean-k-mooney: Hello, can you take a look at https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/992330 and https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/992328/ in your free time, not super urgent, thank you!12:12
sean-k-mooneyill update https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/992330/3 a little later today12:29
sean-k-mooneyi would liek to merge the srbac serives this week assuming there are no outstandign issues. after that i plan ot adress teh ruff/per-commit changes next on the tempest plugin side but ill update https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/992347/1 to cover all ai agents first12:31
chandankumarno all good from my side12:32
chandankumarI will take a look at docs tomorrow12:32
sean-k-mooneychandankumar: do you have any patch for teh tempest plugin you think woudl be good to land sooner rather then later?12:32
sean-k-mooneyi need to swap to spec review soon and hopefully we can move forward whith those today12:33
sean-k-mooneyso that next on my list12:33
chandankumarsean-k-mooney: I think I can drop https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/992210 in favor of your pci scenario tests.12:34
chandankumarNo other patch from my side on cyborg-tempest-plugin12:34
sean-k-mooneyoh the placement ones12:34
sean-k-mooneythose are not nessisarly reduntance but they are also not actully secnairo tests in the normal sense12:35
chandankumaryes, but it is alos more than api tests12:35
sean-k-mooneyya so i think we done really need to have placement assert the provier toplogy12:36
sean-k-mooneybut its up to you.12:36
sean-k-mooneythey add soem addtion coverage but not a lot12:36
chandankumarOk I will fine that patch and move it to api test12:36
sean-k-mooneyhave you tested those with the pci-sim?12:37
chandankumarI tested it with pci-sim12:37
sean-k-mooneyack so ya if you move them to api i think that woudl be fien. we could keep them in secario to as you said its somewhere in between12:37
chandankumarsorry let me double check12:38
sean-k-mooneyi think it shoudl work as long as you set the correct adreses12:38
chandankumartested here: https://review.opendev.org/c/openstack/cyborg/+/977083/51/.zuul.yaml12:38
chandankumarnvmevirt patch12:38
chandankumarhttps://123b844dd0bc940cd071-d2975a8f065f970ca29e743c021b8a36.ssl.cf1.rackcdn.com/openstack/5f698c9d53e44e52bbd22509f59b8b25/testr_results.html12:39
chandankumartest_pci_device_discovery[id-b3d5e7f9-1a2c-4e6f-8b0d-2c4e6f8b0d2e]12:39
sean-k-mooneywhats the plan for that by the way12:39
sean-k-mooneyare we goign to abandon the nvmevirt patches in favor of your new nvme pci-sim series12:39
chandankumaryes correct12:40
chandankumardrop nvmevirt patches12:40
chandankumarsince we can achieve almost everyhting in first party CI12:40
chandankumarsean-k-mooney: I am not sure you saw this message https://review.opendev.org/c/openstack/cyborg/+/995079/6#message-fd9b5227f7e11e2defeda4a79e4b1b19d0f3c0b212:41
chandankumarwe can see nvme block in lsblk output in guest vm but we need a ubuntu image12:41
sean-k-mooneywell i guess it just need an image with the kernel nvme driver12:42
chandankumaryes12:42
sean-k-mooneyi might add that to the list of cirrors changes i have in flight12:42
chandankumarcool, that would be awesome12:42
sean-k-mooneywe coudl use lspci without the driver by the way12:42
sean-k-mooneyas an inital test12:43
sean-k-mooneywe just wont be able to run the lsblk part12:43
chandankumaryup, in cirros we can see nvme in lspci output12:43
sean-k-mooneyi dont realy have an issue with using a alternitive guest image for this test but ya we coudl make that part optional.12:43
chandankumarthat also sounds good12:44
sean-k-mooneydid you see my comment about needing to be abel to have 1 job with both nvme and uart pci-sim pfs12:44
chandankumaryes, will work on that12:44
sean-k-mooneycool, my hope is when we add mdev supprot to pci-sim12:44
sean-k-mooneywe will be able to validate teh 3 generic driver in one job12:45
sean-k-mooneythat will require makign the devstack pluging a little smarter but i think its quite doable12:45
sean-k-mooneyalso i dont know if you read the review guide i created. i need to actully do a deepre review myslef but im goign to go look at the spec shortly before the irc meeting12:46
chandankumarpaste.opendev.org/show/bDXSGmjCs9TH0NRXcVaK/ not yet, it is in my list.12:49
opendevreviewchandan kumar proposed openstack/cyborg-specs master: Add generic NVMe driver spec with secure cleanup  https://review.opendev.org/c/openstack/cyborg-specs/+/98534912:50
chandankumarsean-k-mooney: two more reviews good to merge https://review.opendev.org/c/openstack/cyborg/+/994816 and https://review.opendev.org/c/openstack/cyborg/+/994830 from Melanie.12:53
sean-k-mooneyah the api sample tests and api refe set ya i skimmed them breifly but ill try and review those later in the week12:58
chandankumarsure12:58
jgilaberHi all! IRC meeting will start in ~ 1 hour, add your topics to the agenda https://etherpad.opendev.org/p/openstack-cyborg-irc-meeting#L4813:00
opendevreviewchandan kumar proposed openstack/cyborg-specs master: Add generic NVMe driver spec with secure cleanup  https://review.opendev.org/c/openstack/cyborg-specs/+/98534913:37
opendevreviewTakashi Kajinami proposed openstack/cyborg master: Refactor RequestContext  https://review.opendev.org/c/openstack/cyborg/+/99519513:48
opendevreviewTakashi Kajinami proposed openstack/cyborg master: Remove unused context properties  https://review.opendev.org/c/openstack/cyborg/+/99520513:52
jgilaber#startmeeting cyborg14:00
opendevmeetMeeting started Tue Jun 30 14:00:33 2026 UTC and is due to finish in 60 minutes.  The chair is jgilaber. 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 'cyborg'14:00
sean-k-mooneyo/14:00
jgilaberHello! Who is around today?14:00
chandankumaro/14:00
jgilaberWhile we gather, feel free to add topics to the agenda https://etherpad.opendev.org/p/openstack-cyborg-irc-meeting#L4814:01
rlandyo/14:01
jgilabercourtesy ping: amoralej bogdando14:01
morenodo/14:01
amoralejo/14:01
jgilaberlet's give folks a minute and we'll start with the first topic14:01
jgilaber#topic Reminder: Spec freeze on July 214:02
jgilaberwe have two specs open14:02
jgilaber#link https://review.opendev.org/c/openstack/cyborg-specs/+/98227614:02
jgilaber#link https://review.opendev.org/c/openstack/cyborg-specs/+/98534914:02
melwitto/14:03
jgilaberI'll do another pass of the nvme spec after the meeting but I think that one is good to go14:03
chandankumarI will take a look at mdev spec tomorrow14:03
sean-k-mooneyjgilaber: i just left comments on yours +1 with minor change im happy to merge it14:03
sean-k-mooneychandankumar: yours is next on my list14:04
sean-k-mooneyfor after the meeting14:04
jgilaberack thanks sean-k-mooney I'll look after the meeting14:04
chandankumarthank you!14:04
jgilaberlooks like we're in good shape, anything else on this topic?14:04
jgilaberok, next topic14:05
jgilaber#topic Freezing python-cyborgclient improvements14:05
jgilaberthis one if from rlandy I believe14:05
jgilaber#link https://review.opendev.org/c/openstack/python-cyborgclient/+/99469614:05
rlandyyes...14:05
rlandythis change started with  bug: https://bugs.launchpad.net/python-cyborgclient/+bug/214635114:06
rlandywhich requires removing the old v1 API14:06
rlandyfrom cyborgclient14:06
rlandybut there is also the piece of sdk accelerator v2 support14:07
rlandylooking at the description, the bug now contains many parts14:07
rlandyand we need to rethink the approach here14:08
rlandyas initially the plan was the keep using the openstacksdk14:08
rlandyso chandankumar and I are bringing this to the team to establish a direction14:08
rlandybecause this is probably bigger than one bug14:08
rlandyand we need to decide what should be addressed immediately vs taking a longer design14:09
jgilaberproblem 3 in the bug report seems backwards to me14:09
jgilaberI don't think we want to move the accelerator proxy to the client14:09
chandankumaryes, that needs to be fixed14:10
sean-k-mooneyrlandy: so the python-cybrogclient14:10
sean-k-mooneyalrady uses the sdk14:10
sean-k-mooneyfor v214:10
sean-k-mooneyso this was orgianly intened to track only the v1 removal14:10
melwittyeah what's in the description there looks like specless blueprint material14:10
rlandyok - so that is possibly the right direction - to make the bug just cover v1 removal14:11
sean-k-mooneyyes14:11
rlandyand move the rest to some spec/blueprint etc.14:11
sean-k-mooneywe did dicss this in the ptg14:11
sean-k-mooneyso the propsoal was to remove v1 this cycle and deprecate teh shell14:11
rlandyso I would propose to abandon this change  - and it complicates things14:12
rlandyand just start over with the v1 removal14:12
sean-k-mooneyby shell i mean cybrog command14:12
rlandytaking the rest to PTG?14:12
sean-k-mooneysure we can do that14:12
rlandyok - chandankumar  - you in agreement here?14:12
sean-k-mooneymy hope was to deprecate the python-cybrogclient repo next cycle entirly if we end up upstreamign ti to osc14:12
chandankumarBut for this release, Do we want to fix existing interface and not add any coverage there14:13
sean-k-mooneyand if not  resuding it down to only the osc plugin in 2027.214:13
chandankumarwith v1 removal14:13
rlandythe issue is there is a lot in common14:13
sean-k-mooneychandankumar: which interface?14:13
chandankumarI mean device enable /disable interface14:13
sean-k-mooneyoh that is a sperate bug14:14
chandankumarhttps://bugs.launchpad.net/python-cyborgclient/+bug/215179214:14
sean-k-mooneywe shoudl fix that but not as part of this14:14
sean-k-mooneywe shoudl fix htat by preposoing the sdk change14:14
jgilaberit looks to me that with the v1 removal https://bugs.launchpad.net/python-cyborgclient/+bug/2146351 should be done14:15
chandankumarok14:15
sean-k-mooneyso proeblem 1 and 2 are really the scope of this i think14:16
sean-k-mooney"Problem 3: Migrate accelerator proxy from openstacksdk into cyborgclient" wans never intended to be done14:16
chandankumarsounds good to me drop v1 from cyborgclient and rest of enahncement/fixes goes to sdk change14:16
sean-k-mooneyso i think that where the confustion came form14:16
jgilaber+1 sean-k-mooney 14:16
chandankumarsrry that is Add support from accelerator proxy from openstacksdk into cyborgclient14:17
sean-k-mooneyyes14:17
sean-k-mooneywe do not want to "Migrate the SDK accelerator proxy into cyborgclient"14:17
chandankumaryup14:18
sean-k-mooneyso are folk ok if i update teh bug to just remove problem 314:18
chandankumarI will drop that section from the bug and keep it to v1 removal to clear confson14:18
sean-k-mooneythen i htink we are good14:18
sean-k-mooneyack14:18
sean-k-mooneyworks for me14:18
rlandyok- thank you14:18
rlandyI will submit a new change for the 1 and 2 and abandon the current change14:19
rlandyjgilaber,  thank you - I think we are ok here14:19
jgilaberack, thanks14:19
jgilaber#topics Reviews14:19
jgilaberwe don't have anything listed14:19
jgilaberis there any patch that anyone wants to call out?14:19
melwittthink you need #topic without a s (sorry)14:20
jgilaberoh right, thanks melwitt 14:20
jgilaber#topic Reviews14:20
jgilaberanything? otherwise we can move to bug triage14:20
melwittI wanted to mention about, I have a API test samples patch I tried out here if anyone wants to take a look and see what they think https://review.opendev.org/c/openstack/cyborg/+/99483014:21
sean-k-mooneynothing urgent just an fyi takashi has proposed some cleanup of nova code in cybrog https://review.opendev.org/c/openstack/cyborg/+/995195/414:21
sean-k-mooneymelwitt: oh you removed the WIP :)14:21
melwittyeah just now :) seems ... pretty ok so ready for some scrutiny14:21
chandankumarwill take a look at those this week.14:22
sean-k-mooneyi approved the irrelevent files change this morning https://review.opendev.org/c/openstack/cyborg/+/994816/114:22
melwittah thanks14:22
jgilaberthanks, I'll add them to my review list as well14:23
jgilaberlet's move on 14:23
jgilaber#topic Bugs14:23
jgilaberI saw we have two untriaged14:23
jgilaber#link https://bugs.launchpad.net/openstack-cyborg/+bug/215805814:23
chandankumarwe discussed about this one in last meeting14:24
chandankumar [cyborg-tempest-plugin][fake driver]Improve the scenario test coverage for instance operation in fake driver14:24
chandankumarsean-k-mooney proposed https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/994580 pci driver based instance lifecycle test14:25
chandankumarwe will reuse the same format14:25
chandankumarfor this one, can we mark them as a wishlist?14:25
jgilaberdoes that patch completely address the bug report or will we need more?14:25
chandankumarnono, we need to duplicate the same tests for fake driver based on above patch14:26
jgilaberah ok, I see14:26
jgilaberwishlist/medium seems good to me14:27
sean-k-mooneyso it kind of superceed it14:27
sean-k-mooneyi actully think there is very littel point in useing the fake driver for the majoryit of the lifecycle operations14:27
sean-k-mooneybecause the only thing that tests is the arq creation14:27
sean-k-mooneyon the nova side we entirly drop the arq info14:28
melwittyeah I kind of wondered about that ... how helpful is the fake driver for lifecycle testing14:28
sean-k-mooneyso i woudl prefer to only test program and create with the fake driver14:28
sean-k-mooneyand test the rest with pci/mdev/nvme drivers14:28
chandankumarfor program I have awip patch https://review.opendev.org/c/openstack/cyborg-tempest-plugin/+/99458014:29
sean-k-mooneychandankumar: so we have a exsitng boot test with the fake driver14:29
sean-k-mooneyonce we merge your program tests14:29
sean-k-mooneyi woudl close https://bugs.launchpad.net/openstack-cyborg/+bug/215805814:29
jgilaberclose after the patch merges or close it now?14:29
sean-k-mooneywell i guess it depend on if we add Close-Bug to oen of them :)14:30
sean-k-mooneyor related bug14:30
chandankumarI can add closes -bug , then it will automatically closes14:30
sean-k-mooneycool those that work for others?14:30
melwittprobably want to add a comment to it about not doing lifecycle tests as part of it. or edit the description14:31
jgilaberack, so in the meantime we set it as wishlist and low/medium?14:31
chandankumarI will add a comment to the bug 14:32
chandankumarlow/medium +114:32
chandankumarlow14:32
jgilaberack set then14:33
jgilabernext one14:33
opendevreviewMerged openstack/cyborg master: Add api-ref/ to irrelevant-files in Zuul config  https://review.opendev.org/c/openstack/cyborg/+/99481614:33
jgilaber#link https://bugs.launchpad.net/openstack-cyborg/+bug/215814714:33
melwittoh, that's the one in flight patch approved from earlier14:34
melwittso it will get closed when the patch merges14:34
jgilaberok, so that seems valid and is in progress, only missing the importnace14:34
jgilabermedium?14:34
sean-k-mooneyya it has no end user impact directly14:35
melwittoh wait sorry. I misread that it's after the api-ref thing14:35
sean-k-mooneybut melwitt has already fixed some incorrect examples14:35
melwittanyway. yeah not high importance14:35
jgilaberack thanks melwitt 14:35
jgilaberthat was the last one14:35
jgilaber#topic Open discussion14:36
jgilaberany last minute topic to raise?14:36
sean-k-mooneymaybe we coudl chat abouy https://github.com/adamchainz/flake8-logging/pull/187 quickly14:36
jgilaberlooking14:36
sean-k-mooneymelwitt: so we have moved away form using flake8 in general to using ruff for linting14:37
sean-k-mooneywe coudl run that check via the hacking env14:37
sean-k-mooneysince it uses flake internally14:37
melwittok that's good to know14:37
sean-k-mooneybut woudl it be better to just have a local hacking check instead14:37
melwittI don't know a lot about it so I wasn't aware we could make local checks. if that's an option then seems like a nice idea?14:38
sean-k-mooneyso there are 2 ways we coudl do this14:39
melwittI'm skeptical my PR will go anywhere but I tried it anyway14:39
sean-k-mooneya pre-commit hook or a local hackign check 14:39
sean-k-mooneythe later is proably simpler14:39
sean-k-mooneythis might be genericlly usefully enough to put in hacking propper as well14:39
melwittok, I will look into that then, thanks14:40
sean-k-mooneyah right14:41
sean-k-mooneyhttps://github.com/openstack/cyborg/commit/8eb3a5a4b2a8fa0fc969290ceace9f574715d02714:41
sean-k-mooneyi delete our hacking plugin14:42
sean-k-mooneybut we can add one back14:42
melwitt:)14:42
sean-k-mooneyi deleted it because i didnt like the preior rules it used in general14:42
sean-k-mooneywell or the were duplicats of other checks14:43
melwittoh ok14:43
sean-k-mooneynova has a bunch of hackign plugin examples too for what its worth14:43
sean-k-mooneyanyway i think we can chat about that outside the meeting14:44
sean-k-mooneyany other topics or questions on this topic?14:44
melwittsure. I will put something together. thanks for the info14:44
sean-k-mooney... https://github.com/openstack/cyborg/blob/master/cyborg/__init__.py sorry got distracted14:45
sean-k-mooneywe need to fix  ^ 14:45
sean-k-mooneyi nocied that lookign at the old commit 14:45
sean-k-mooneybut we have been asked to remove any runtim usage of pbr by the pbr maintianers14:45
melwittok good to know14:46
sean-k-mooneyill file a bug for that later14:46
sean-k-mooneywe are emnt to use import lib metadta or soemthign instead14:46
sean-k-mooneyor just stop providing __version__14:46
sean-k-mooneythere was a mail thread about it a month ago or there abouts14:46
jgilaberack thanks sean-k-mooney we can look at the bug in the next meeting14:47
jgilaberI think I missed this email14:47
sean-k-mooneyhttps://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/V7JFTBRA3CHHIWCZM5CM2LE27ESDJHO7/#V7JFTBRA3CHHIWCZM5CM2LE27ESDJHO714:47
sean-k-mooneyhttps://review.opendev.org/c/openstack/python-openstackclient/+/987782/2/openstackclient/__init__.py14:48
sean-k-mooneyso that the exampel fo what we shoudl do instead14:48
jgilaberseems simple enough14:48
sean-k-mooneyok ill file a bug for that and create a patch later14:48
sean-k-mooneyya pretty much14:49
sean-k-mooneythey just would prefer to keep pbr as a build time only tool going forward to limit the maintance14:49
jgilaberthanks for bringing it up sean-k-mooney, anything else on this topic or any other?14:51
jgilaberlast thing for today14:53
jgilaber#topic Volunteers to chair next meeting14:53
jgilaberany volunteer?14:53
rlandyI can do it14:53
jgilaberthanks rlandy 14:54
jgilaberthat is all for today, thanks everyone!14:54
jgilaber#endmeeting14:54
opendevmeetMeeting ended Tue Jun 30 14:54:19 2026 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)14:54
opendevmeetMinutes:        https://meetings.opendev.org/meetings/cyborg/2026/cyborg.2026-06-30-14.00.html14:54
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/cyborg/2026/cyborg.2026-06-30-14.00.txt14:54
opendevmeetLog:            https://meetings.opendev.org/meetings/cyborg/2026/cyborg.2026-06-30-14.00.log.html14:54
morenodthanks!!14:54
rlandyjgilaber++ thank you14:54
melwittsean-k-mooney: a primer for the API samples testing patch ... I used claude and gave a goal for something much lighter weight than the usual .tpl model. if you hate it let me know and I can try something else15:11
sean-k-mooneyi hate our normal model so im sure ill like it :)15:12
melwitthaha well, I have left the hate door open if you need it :)15:13
sean-k-mooneyso long term (in like 6-12 months time) i woudl love to not need this because we fully impleitne the OPENAPI schema goal for cybrog15:13
melwittnext in the "plan" is gabbi API tests15:14
melwittyeah, good point, that takes some of the pressure off15:14
sean-k-mooneyi.e. it woudl be nice if the schmea was validate speratly form the sampels which jsut woudl help with docuemtion15:14
sean-k-mooneyi need ot start a wider converation on evovling the api15:15
melwittso openapi does both request and response schemas right?15:15
sean-k-mooneyyes15:15
sean-k-mooneyif you implemnt both of course15:15
melwittI mean, should I instead be starting with that?15:15
sean-k-mooneyno i think this is low haning fruit15:16
sean-k-mooneyand its good to get this in place to have any coverage15:16
melwittok cool15:16
melwittyeah. it found some minor errors15:16
sean-k-mooneybut if you wanted to start on that after sure. we should ahve that converstaion about the other api issues that i was hopign to look at next cycle15:16
melwittwell, I guess define "minor" but some parameters that should not have been in the samples and some missing15:16
melwittnah just wanted to make sure I stay aligned with the wanted plans15:17
sean-k-mooneynope your totally aligned now worries on that front15:17
sean-k-mooneyso tempst ahas only ever had scemeas for nova and cindier i belive15:17
sean-k-mooneyi woudl perfer to avodi addig them for cybrog and jsut do the openapi feature isntead15:17
melwittthe tempest schemas are so painful in my experience15:18
sean-k-mooneyespcially since cybnorg api is small15:18
sean-k-mooneyby the way i think one fo the api routes is missing form the api ref15:18
melwittoh hm15:19
sean-k-mooneyhttps://docs.openstack.org/api-ref/accelerator/#deployables or https://docs.openstack.org/api-ref/accelerator/#devices15:19
sean-k-mooneyhas a update/program endpoint15:19
sean-k-mooneyfor programign fpgas15:19
sean-k-mooneyi belive it sdeployables via a patch or put15:19
melwittok I'll check on it15:20
sean-k-mooneylooking at my srbac serise its deployables https://review.opendev.org/c/openstack/cyborg/+/992295/5/cyborg/policies/deployables.py#4115:21
sean-k-mooneyah and yes PATCH15:21
sean-k-mooneychandan has a chagne up to test that with the fake driver via tempest 15:22
sean-k-mooneylong term im not conviced that api shoudl exsit but that a sperate matter15:23
melwittso speaking of the tempest test stuff ... is your pci kernel module thing what we can use to do proper lifecycle coverage?15:23
sean-k-mooneyyep15:23
melwittok great15:23
sean-k-mooneythe emulated device can be passed to the qemu isntance and actully work15:23
melwittsuper15:23
sean-k-mooneyso the uart is a simple loop back15:23
sean-k-mooneychandan has extened it to emulate nvme ssds15:24
sean-k-mooneyi plan ot add nic emualtion evnutlly adn supprot for mdevs15:24
sean-k-mooneyso we will be able to emulate all of the devices we manage via the 3 generici driver and in the future htat will allow us to do sriov testing in nova as well if/whewn we want too15:25
melwittawesome15:25
sean-k-mooneymelwitt: i have been holding of adding that to the nova jobs until we are happy with the current feature set and the stableity 15:25
melwitt++ I think that is wise15:25
sean-k-mooneybut im thinking of adding it to the new nova-whitebox job15:25
melwittspeaking of that, we need your re-review on https://review.opendev.org/c/openstack/nova/+/898100 to land it15:27
sean-k-mooneyits pretty stable form what i can tell and compiling it is fast but i was goign to give it a month or two before goign town that route15:27
sean-k-mooneyoh yes cool ill look at that shortly15:27
melwittyeah ++ for adequate bake time15:27
sean-k-mooneyit also currently need debian 13 or ubuntu 26.04 which i dont think that whiteboxc job uses15:28
melwittoh right15:29
sean-k-mooneyso i want to get the whitebox tests running on the cybrog job first15:29
sean-k-mooneyand only look at updatign the nova oen after we have all the kinks worked out15:29
melwitt++15:29
sean-k-mooneyone of the optimisation we need to do for whitebox is use the serial decorator15:29
sean-k-mooneyinstead of --concurancy=115:29
sean-k-mooneyi think gmann was consideritng doign that refactor but if not i woudl like to do that before integrating pci-sim into it15:30
melwittok so about that15:30
melwittthere is one test in whitebox, a vtpm test, that restarts libvirtd. this is what messes up all ability to use parallel. we need to separate out that test for it to run on its own15:31
melwittI could work on that actually, I just didn't think about it at the time15:32
melwittI guess there might be other similar tests in whitebox too. if so, just all of those need to go in their own serial/ directory and run after everything else like normal tempest does15:32
gmaani have that in my list but I think I will not be able to start that for next 2 weeks, feel free to do it if you can before that15:32
melwittah cool :)15:33
gmaansean-k-mooney: melwitt I was thinking to add it in meeting but missed. as you both are here, please check/test the enforce_scope removal for cyborg, I have proposed the example changes for other projects, melwitt has reviewed the nova/placement one already. https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/436HHMRVNUY3OSXE4ULPQ4P4DSHQEMES/15:35
melwittok cool yes I will look15:36
gmaanthanks15:36
gmaani can see only a few reference of that but you can add testing change with oslo.policy change and test if anything else other than those few test changes is needed https://github.com/search?q=repo%3Aopenstack%2Fcyborg%20enforce_scope&type=code15:38
melwittoh yikes the deadline is july 315:38
opendevreviewTakashi Kajinami proposed openstack/cyborg master: Remove unused context properties  https://review.opendev.org/c/openstack/cyborg/+/99520515:44
opendevreviewTakashi Kajinami proposed openstack/cyborg master: Remove unused context properties  https://review.opendev.org/c/openstack/cyborg/+/99520515:46
opendevreviewmelanie witt proposed openstack/cyborg master: Fix api-ref inaccuracies and add missing program endpoint  https://review.opendev.org/c/openstack/cyborg/+/99553316:02
opendevreviewTakashi Kajinami proposed openstack/cyborg master: Remove unused _extract_attributes_v2  https://review.opendev.org/c/openstack/cyborg/+/99553416:02
sean-k-mooneygmaan: yes i have been folowing that form afar16:02
sean-k-mooneygmaan: basiclly waitign for you and stephen to figure out the way forward16:03
gmaansean-k-mooney: ack, thanks16:03
opendevreviewTakashi Kajinami proposed openstack/cyborg master: Remove unused method for image v1 API  https://review.opendev.org/c/openstack/cyborg/+/99553416:03
sean-k-mooneygmaan: so i need to remove my overried where i explicly default that to true16:03
sean-k-mooneybut i wrote teh securigy bug fixes assuming the enfoce scope16:04
gmaanyes, true or false, basically all usage of enforce_scope16:04
sean-k-mooneyand that whaat my srbac serise is built around as well16:04
gmaanenforce_new_defaults stay same to toggle the old/new defaults but not the scope16:04
sean-k-mooneyyep16:04
sean-k-mooneyill tack a patch on the end of my srbac serise to drop our usage16:05
sean-k-mooneygmaan: im hopign to merge that this week as well16:05
gmaan++16:06
sean-k-mooneymelwitt: sorry was on a call16:06
sean-k-mooneymelwitt: so there are a couple of whitebox tests that restart the agent16:06
sean-k-mooneymelwitt: we used ot have test that reconfigre the nova.conf setting to test spefic feature that were mutally exclivie16:07
sean-k-mooneyso for vtpm ya there is one that need special handleing but the test for offliniing cpus and some fo the pining test used to restart agents too16:07
melwittyeah16:08
sean-k-mooneyso if you have time to work on making whitebox work with the serial decorator that woudl be cool16:08
sean-k-mooneybut no pressure either way16:08
sean-k-mooneymy current plan is wrap up srbac this week. then finish the pci-mangege mode work in then 1-2 weeks and then loook at addign whitebox tests16:09
sean-k-mooneyso if you get to it before the end fo the month the cool16:09
sean-k-mooneyif not ill look at it in august16:10
sean-k-mooney*end of the month being july16:10
sean-k-mooneyim happy with the simple lifecyle test i have that ssh into the vm and run lspci are enouch for the pci driver testing im using right now. but i want to add a whitebox test that will check the xml actuly has manged=no later16:11
sean-k-mooneyi also verifyed it on real hardware. turns out if managed mode is not set with an l4 booting a vm will hard reboot the host which was fun to debug16:12
sean-k-mooneyso nvidia still have not fixed that kernel bug :)16:12
sean-k-mooneythe pci-sim wont crash the same way16:13
melwittoh wow hah16:17
sean-k-mooneywe have a note about that in the nova docs16:25
sean-k-mooneyit really really shoudl not happen16:26
sean-k-mooneybut ya i was like why is my ssh conenction droping when i boot a vm16:26
sean-k-mooneyhost reboot will do that :)16:26
melwittcrazy haha16:38
opendevreviewmelanie witt proposed openstack/cyborg master: Fix api-ref inaccuracies and add missing program endpoint  https://review.opendev.org/c/openstack/cyborg/+/99553316:46
opendevreviewTakashi Kajinami proposed openstack/cyborg master: Remove unused context properties  https://review.opendev.org/c/openstack/cyborg/+/99520517:19
opendevreviewJoan Gilabert proposed openstack/cyborg-specs master: Add generic mdev driver spec for 2026.2  https://review.opendev.org/c/openstack/cyborg-specs/+/98227617:28
opendevreviewmelanie witt proposed openstack/cyborg master: Add C300 hacking check for set literal in logging calls  https://review.opendev.org/c/openstack/cyborg/+/99556118:15
opendevreviewmelanie witt proposed openstack/cyborg master: Add C300 hacking check for set literal in logging calls  https://review.opendev.org/c/openstack/cyborg/+/99556118:38
sean-k-mooneymelwitt: i goign to cal lit a day but i jsut re reviewd the hackign check and i was incorrect about removig pycodestyle20:52
sean-k-mooneyjust pushed updated comemnts20:52
melwittsean-k-mooney: ok cool thanks. I was just having the same confusion :)20:54
sean-k-mooneyi didnt realsie you were using it in the unit test20:55
sean-k-mooneyso its a driect depency of the unit tests so you were right20:56
sean-k-mooneyand ya i have not seen teim recommend addign teh liscen here to the __init__.py20:56
sean-k-mooneyim not sure if that is becaue  i just updated the model20:56
sean-k-mooneyor if its a chagne i hacking20:56
sean-k-mooneybut ya with the current hackign setting that is an error ill correct that20:56
sean-k-mooneyteim-ci will read HACKING.rst or AGENTS.md if they are present20:57
melwittah ok20:57
sean-k-mooneyso i can add a note in HACKING.rst about that20:58
sean-k-mooneyor tweak its prompt for this edgecase20:58
sean-k-mooneyok chat to you tomorrow o/20:58
melwittcool thanks, seeya tomorrow o/20:59
opendevreviewmelanie witt proposed openstack/cyborg master: Add C300 hacking check for set literal in logging calls  https://review.opendev.org/c/openstack/cyborg/+/99556121:42

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