Friday, 2022-06-03

*** mfo is now known as Guest104200:07
*** mfo_ is now known as mfo00:07
gibio/07:05
bauzas\o07:27
*** bhagyashris|rover is now known as bhagyashris07:37
opendevreviewBalazs Gibizer proposed openstack/nova master: Unparent PciDeviceSpec from PciAddressSpec  https://review.opendev.org/c/openstack/nova/+/84449108:02
gibielodilles, melwitt: what is the stable view on squashes? https://review.opendev.org/c/openstack/nova/+/843680/6#message-f9be680631a00541158243c55609a5804cc7777d08:33
gibistephenfin: do you have some context on https://review.opendev.org/c/openstack/hacking/+/816676 ? the hacking repo is green now with flake 4 but in nova we have a bunch of new findings. Was there any previous trial moving nova to the newer hacking?08:39
songwenpingbauzas:we donnot delete vgpu mdev when delete vm?08:42
sean-k-mooney**cough** black **cough**08:42
sean-k-mooneysongwenping: ya our down stream qe may have also found that yesterday08:43
sean-k-mooneysongwenping: we are still confirming08:43
sean-k-mooneysongwenping: its unclear if something has changed recently i.e. if this only happens if you have mig mode enabled08:44
sean-k-mooneyor if this was always an oversight08:44
songwenpingsean-k-mooney: from this line https://github.com/openstack/nova/blob/8260979b71b29ce2666d37b3adc7c256482aa16d/nova/virt/libvirt/driver.py#L6485 we can reuse the mdev08:44
sean-k-mooneysongwenping: we should not be reusing the mdev08:44
kashyapsean-k-mooney: I think you have Ubuntu box handy, can you please tell the QEMU binary path on it?  (Not the /usr/bin/qemu-kvm, but the actual binary)08:45
sean-k-mooneysure that my home server. its not in libexec like its on rhel related distros08:45
sean-k-mooneythe fact its different is waht prevent live migration form happening one sec while i check08:45
songwenpingso we should delete the mdev with vm.08:45
kashyapsean-k-mooney: I know it's not in /usr/libexec - that's a RHEL/CentOS thing08:45
sean-k-mooneykashyap: yep just sayign they used to be the same08:46
kashyapsean-k-mooney: The thing is on Fedora the binary is this: /usr/bin/qemu-system-x86_6408:46
sean-k-mooneykashyap: but when centos change to that it broke live migration between centos and ubunutu08:46
bauzassongwenping: no, we leave them08:47
kashyap(And IIRC, Ubuntu uses a different binary name and a different path)08:47
sean-k-mooney/usr/bin/qemu-system-x86_6408:47
kashyapAh, same as Fedora then.  Thanks08:47
sean-k-mooneyits qemu im not seeing qemu-kvm on 22.0408:47
sean-k-mooneylet me double check08:47
sean-k-mooneysean@cloud:~$ qemu-kvm08:47
bauzassongwenping: you can even precreate the mdevs before starting the nova-compute service, then the n-cpu service will use them instead of creating them08:47
sean-k-mooneyqemu-kvm: command not found08:47
sean-k-mooneyso its not on my path at least08:47
kashyapsean-k-mooney: I don't want that; it's an alias to the QEMU binary on Fedora08:48
kashyapsean-k-mooney: So, "/usr/bin/qemu-system-x86_64" exists on Ubuntu, right?08:48
sean-k-mooneyyes08:48
songwenpingbauzas:then there will has race problem, https://bugs.launchpad.net/nova/+bug/183620408:48
sean-k-mooneysean@cloud:~$ file /usr/bin/qemu-system-x86_6408:48
sean-k-mooney/usr/bin/qemu-system-x86_64: ELF 64-bit LSB pie executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, BuildID[sha1]=f28deae95b4b0f19a8c8be87c45d1be213cf8b6b, for GNU/Linux 3.2.0, stripped08:49
kashyapsean-k-mooney: Thanks!08:49
sean-k-mooneykashyap: it exiss and is not a symlink its the binary08:49
kashyapOkay, I see that I could've checked it here: https://packages.ubuntu.com/bionic/i386/qemu-system-x86/filelist08:49
kashyapsean-k-mooney: Yeah, I know; if that's the name, then it's the real binary08:49
kashyap'qemu-kvm' is the symlink on Fedora/CentOS08:49
sean-k-mooneykashyap: not quite08:50
sean-k-mooneywell it might be a symlink but it was a script on ubuntu08:50
kashyap$> file /usr/bin/qemu-kvm 08:50
kashyap/usr/bin/qemu-kvm: symbolic link to qemu-system-x86_6408:50
songwenpingbauzas:if we create new mdev when alloc, there will be no this problem.08:50
kashyapsean-k-mooney: Yeah, it used to be a wrapper script08:50
sean-k-mooneyyep08:50
kashyap(Which just pointed to the real binary.  So they removed that)08:50
sean-k-mooneyit was a wrapper on ubuntu in 20.0408:50
sean-k-mooneyin 22.04 they droped it08:51
bauzassongwenping: I don't see how it could help 08:51
sean-k-mooneykashyap: looks like ti snow a virtual package/metapackage https://packages.ubuntu.com/jammy/qemu-kvm08:51
sean-k-mooneylikely for upgrade reasons08:52
kashyapYeah, same on Fedora08:53
kashyap$> rpm -ql qemu-kvm08:53
kashyap(contains no files)08:53
sean-k-mooneykashyap: if you want to poke around i can give you access to the server just let me know08:53
elodillesgibi: for the sake of easier reviewing i'd say squash is better to avoid. some rare cases though we can have squash: e.g. in case otherwise we could not unblock the gate. or in case we can avoid some introduced bug that the follow-up fixes.08:53
kashyapsean-k-mooney: Thank you, it's not needed at this time.  I just wanted to be sure of the path - so that we could supply it to `coredumpctl` :) 08:54
sean-k-mooneyhum never had to use that08:54
sean-k-mooneydoes that parse them or cause them :)08:54
songwenpingbauzas: in this function https://github.com/openstack/nova/blob/8260979b71b29ce2666d37b3adc7c256482aa16d/nova/virt/libvirt/driver.py#L6416, we use _create_new_mediated_device for every vm instead of select a mdev from mdevs_available.08:55
sean-k-mooneysongwenping: we inteded to not reuse mdevs for new vms08:57
sean-k-mooneythe other path for intedd for agent restart if i rememebr correctly08:57
sean-k-mooneysongwenping: so we shoudl be delete the mdevs if we delete the vm08:58
songwenpingsean-k-mooney: cool08:58
sean-k-mooneysongwenping: this become more important for cards that can support more then one mdev at a time with generic mdev support08:58
sean-k-mooneyi dont think nvidia do with any of there gpus btu other vendors might in the future08:59
sean-k-mooneybauzas: precrating does not work and i dont think it was part of the sepc so if that was added its a bug09:00
songwenpingsean-k-mooney: ack, agree to delete the mdevs if we delete the vm09:00
bauzassean-k-mooney: if we make nova creating the mdevs directly, this could create other problems09:00
bauzasfor reboot, for example09:00
sean-k-mooneybauzas: we dont support using mdevctl and the current case faild our qe09:00
bauzasremember the difference between SR-IOV devices and mdevs09:01
sean-k-mooneybauzas: nova is not able to create the mdevs because it did not clean up09:01
sean-k-mooneybauzas: yes i know09:01
sean-k-mooneywe only support nova creating the mdevs today09:01
sean-k-mooneyincluding in the mig case09:01
bauzasin the mig case, nothing changes from nova pov09:02
bauzasthis is just that we precreate the pci devices09:02
bauzasnot the mdevs09:02
sean-k-mooneyyep09:02
songwenpingwhen manage vgpus by cyborg, we precreate the mdevs when discover09:02
sean-k-mooneypre creat the pci device and list the vfs instead of the pf09:02
bauzassongwenping: sean-k-mooney: anyway, I'm not against fixing this old bug09:02
bauzassongwenping: if you have time for fixing it, would be appreciated09:03
bauzasnone of this requires a spec09:03
songwenpingbauzas:i'll try.09:06
kashyapdansmith: To answer to your question in the scrollback: yeah, I'm guessing it's different - based on the trigger here (backup) vs. the older one (disk-detach)09:29
fricklergibi: hacking-integration-nova has been busted for ages09:32
fricklerlikely before we moved to zull v3 even https://zuul.opendev.org/t/openstack/builds?job_name=hacking-integration-nova&project=openstack%2Fhacking&result=SUCCESS&skip=009:32
frickler*zuul09:33
opendevreviewBalazs Gibizer proposed openstack/nova master: Fix PciAddressSpec descendants to call super.__init__  https://review.opendev.org/c/openstack/nova/+/84456509:34
gibifrickler: I think it shows that nova is not compatible with latest hacking09:34
gibifrickler: am I mistaken?09:34
fricklergibi: this is true. afaict nova also hasn't been compatible with at least 3 years of previous releases of hacking, so not much change09:35
gibiahh OK09:36
gibifrickler: so you think we can land the flake version bump after I drop py35 testing or we need to fix nova first?09:36
sean-k-mooneyfrickler: gibi  if we use black hacking become less importnat. still nice for extra non style related checks like dont improt prev sep stuff with from09:37
sean-k-mooneyor any other rules we want to enforece but the systle related ones are less requried09:38
gibisean-k-mooney: the current failures are thinks that are not covered by black / flake09:38
fricklergibi: nova caps to 3.1.0, there are 3.2.0, 4.0.0 and 4.1.0 already. so adapting nova to recent hacking should not block updating hacking IMO09:38
sean-k-mooneyack i was just hoping it would reduce the number of things we have to port09:38
gibifrickler: thanks09:38
gibiI will update the hacking patch to drop py35 then09:39
sean-k-mooneyfrickler: yep it should not09:39
frickleralso I must reduce my "3 years" claim to 2 years, 3.1.0 was released in May 2020. no idea how far back our history of zuul builds actually goes09:40
sean-k-mooneyfrickler: i think there was a change in hackign that broke some fo our checks which is why we pinned09:54
sean-k-mooneyfrickler: but then stephenfin who was lookign at fixing it moved to not work on nova as part of there day job09:55
sean-k-mooneyand we had others move too09:55
sean-k-mooneyso we just never got aroudn to adressint the issues09:55
sean-k-mooneywe might still have patches form stephenfin for some of the issues09:55
sean-k-mooneynot that i can see on this topic09:57
sean-k-mooneyfrickler: we did recently move to 3.109:58
sean-k-mooneyhttps://review.opendev.org/c/openstack/nova/+/836639/1/.pre-commit-config.yaml09:58
sean-k-mooneyactully i guess the bump was done in victoria https://github.com/openstack/nova/commit/61b99a1295c5208deef806b69ba74a7d031ad85110:03
fricklersean-k-mooney: yes, that was right after the release of 3.1.0, but it seems after that, nova lost track10:04
sean-k-mooneyyep it was not broken so i guess we just never updated it10:05
sean-k-mooneyproably also just resouce constrianed/pandmic related10:05
* gibi hates the pci whitelist parsing code 12:18
sean-k-mooney:)12:18
sean-k-mooneydont you love how we only support bash globs if the adress is a sting and only support regexs if its a dictionary12:19
gibinah, the self.is_physical_function handling is worst in my eyes12:20
gibiwe write that flag twice based on two different utility function reading the same sysfs location12:21
sean-k-mooneyheh of course we do12:21
gibibut yes, that string / dict duality is close second12:21
sean-k-mooneywe did it because of the conflict between * in glob and regex meendnin ins * is .* in regex land12:22
sean-k-mooneyi just wish we used regex form from the start12:22
gibiyepp I figured that regex was added later and blow up the picture12:22
gibiand of course the whole devname special case is a pain12:22
sean-k-mooneyi would not mind devname if it worked for all devices12:23
sean-k-mooneybut the fact that its unreliable and only works for nic12:23
sean-k-mooneysometimes12:23
sean-k-mooneyis why i hate it12:23
sean-k-mooneyonce you are done with the placment work i do still think its worth revisigint our config format in a differnt discussion12:24
gibiI agree12:25
gibiI will try to untangle as much of the current parsing code as possible before I add the new things to it 12:25
sean-k-mooneylike in general can oslo config supprot yaml or can we add a resouces.yaml for all the complext host level resouces12:25
sean-k-mooneyok cool12:26
sean-k-mooneythe other thing to be aware of is how we parse physical_network12:26
gibiI'm not there yet :) 12:26
sean-k-mooneyspeicifcliy we are relying on physical_network=null being converted to python None12:27
sean-k-mooneyso we are using the json parsing to differnceat between not set, physical_network=null and pyhsical_network="null" or physical_network="None"12:28
gibinice12:28
sean-k-mooneyall 4 of those have differnt meanings12:28
sean-k-mooneythe last 2 are just strings that are the name of a phsyical network in neutron12:28
sean-k-mooneypyhical_network=null without quotes is converted to python None and that is used for hardware offloaded ovs with tunneld networks12:29
sean-k-mooneyand unset means this is not a nic12:29
gibiI would like to give prizes to deployments naming there physnets  None and null :D12:29
sean-k-mooneyhehe ya12:30
sean-k-mooneybut since the null vaule was not planned to be supported12:30
sean-k-mooneyand was a bug that was abused for hardware offloed ovs12:30
sean-k-mooneyits posible12:30
sean-k-mooneyhttps://bugs.launchpad.net/nova/+bug/191528212:32
dansmithkashyap: so it seems like this heisenbug has receded into the background, even though only one package has changed in fedora since last week (that we install)13:35
dansmithkashyap: so I'm going to formalize my devstack patch so we'll just always capture qemu coredmps so we're ready for next time13:35
kashyapdansmith: Huh, the bugs from hell.  Same story w/ that other detach thing.  Although we sussed out a latent libvirt bug from it13:36
kashyapdansmith: Yeah, thank you; I also saw that the same path works for Ubuntu as well - I checked, and by extension, Debian too13:36
dansmithscary if that means users can hit them, but yeah :/13:36
dansmithkashyap: ah cool13:36
kashyapdansmith: I commented w/ a link to evidence your DNM patch13:36
dansmithcool thanks13:37
opendevreviewArtom Lifshitz proposed openstack/nova stable/wallaby: fake: Ensure need_legacy_block_device_info returns False  https://review.opendev.org/c/openstack/nova/+/84367814:03
opendevreviewArtom Lifshitz proposed openstack/nova stable/wallaby: Add a regression test for bug 1939545  https://review.opendev.org/c/openstack/nova/+/84370214:03
opendevreviewArtom Lifshitz proposed openstack/nova stable/wallaby: compute: Ensure updates to bdms during pre_live_migration are saved  https://review.opendev.org/c/openstack/nova/+/84368014:03
opendevreviewArtom Lifshitz proposed openstack/nova stable/wallaby: fup: Make connection_info returned by CinderFixture unique per attachment  https://review.opendev.org/c/openstack/nova/+/84459414:03
opendevreviewArtom Lifshitz proposed openstack/nova stable/wallaby: fup: Assert state of connection_info during LM rollback in func tests  https://review.opendev.org/c/openstack/nova/+/84459514:03
opendevreviewArtom Lifshitz proposed openstack/nova stable/victoria: compute: Ensure updates to bdms during pre_live_migration are saved  https://review.opendev.org/c/openstack/nova/+/84394914:15
opendevreviewArtom Lifshitz proposed openstack/nova stable/victoria: fup: Make connection_info returned by CinderFixture unique per attachment  https://review.opendev.org/c/openstack/nova/+/84459814:15
opendevreviewArtom Lifshitz proposed openstack/nova stable/victoria: fup: Assert state of connection_info during LM rollback in func tests  https://review.opendev.org/c/openstack/nova/+/84459914:15
opendevreviewArtom Lifshitz proposed openstack/nova stable/ussuri: fake: Ensure need_legacy_block_device_info returns False  https://review.opendev.org/c/openstack/nova/+/84395015:19
opendevreviewArtom Lifshitz proposed openstack/nova stable/ussuri: Add a regression test for bug 1939545  https://review.opendev.org/c/openstack/nova/+/84395115:19
opendevreviewArtom Lifshitz proposed openstack/nova stable/ussuri: compute: Ensure updates to bdms during pre_live_migration are saved  https://review.opendev.org/c/openstack/nova/+/84395215:19
opendevreviewArtom Lifshitz proposed openstack/nova stable/ussuri: fup: Make connection_info returned by CinderFixture unique per attachment  https://review.opendev.org/c/openstack/nova/+/84460615:19
opendevreviewArtom Lifshitz proposed openstack/nova stable/ussuri: fup: Assert state of connection_info during LM rollback in func tests  https://review.opendev.org/c/openstack/nova/+/84460715:19
gibisean-k-mooney: I just realized that there is antoher edge case in the PCI parsing, the remove managed feature allows PF address and VF product ID in the same device spec test_remote_managed_pf_raises15:53
gibihttps://github.com/openstack/nova/blob/ffb810e2ba2fdec9b2a881a88fa6d65cd32f8fa3/nova/tests/unit/pci/test_devspec.py#L501-L51315:53
*** xek__ is now known as xek15:56
sean-k-mooneygibi: that was a prexisting feature15:57
sean-k-mooneythey did not add it15:57
sean-k-mooneythat is how you whitelisted the VFs of a pf before we added teh glob and regex support15:57
sean-k-mooneyso thats been there since like icehouse15:57
gibiit is pretty remote managed specific https://github.com/openstack/nova/blob/d86916360858daa06164ebc0d012b78d19ae6497/nova/pci/devspec.py#L315-L33415:58
sean-k-mooney no this work for any sriov device15:58
sean-k-mooneythey just added a test case for it15:58
gibiI think what you are referring to is the ability to list PF and match VF, but the remote managed extends this to list PF with VF's product_id15:58
sean-k-mooneyso you used to be able to use the pf address and vf product id15:59
gibiit has an if self._remote_managed: on top so this does not run for the rest of the caseas15:59
sean-k-mooneyand that would allow all the VFs of that pf to be used15:59
sean-k-mooneybut not the pf itself15:59
sean-k-mooneyjust looking at the code you linked now15:59
gibiin the generic case I only see PF address matching for VF devs, but no such logic for vendor / product matching16:00
sean-k-mooneyi think its this https://github.com/openstack/nova/blob/d86916360858daa06164ebc0d012b78d19ae6497/nova/pci/devspec.py#L254-L265=16:02
sean-k-mooneybut i would have tro look at this carfully16:02
sean-k-mooneywe do supprot suign adress=<pf addres> product_id=<vf id>16:02
gibinope, that is only matching addresses 16:03
gibithe PhysicalPciAddress has no info about the vendor / product16:03
sean-k-mooneyits match the adress object16:03
gibiyepp 16:03
gibionly address matching happens there16:03
gibivendor / product happens independently https://github.com/openstack/nova/blob/d86916360858daa06164ebc0d012b78d19ae6497/nova/pci/devspec.py#L377-L37816:04
sean-k-mooneyright but that is combiend with the vendro id and prodcut id filter elsewhere16:04
sean-k-mooneyyes it does so the way it works is we mach on vendor and product id first16:06
sean-k-mooneyand then we match on the adrees or the partent adress16:06
sean-k-mooneyhttps://github.com/openstack/nova/blob/d86916360858daa06164ebc0d012b78d19ae6497/nova/pci/devspec.py#L379-L380=16:06
melwittsean-k-mooney: I was wondering if you might be able to help see what's wrong with this failure, I think it might be related to ansible. I have written about it on L57 https://etherpad.opendev.org/p/nova-stable-branch-ci tl;dr is we're passing executable=/bin/bash with the ansible command args but it's running with /bin/sh on the remote side. I don't understand it16:07
melwitthave you seen something like that before?16:07
sean-k-mooneymelwitt: i was talking to fungi about that yesterday on #opentack-infra16:07
gibisean-k-mooney: hm, OK, I see it now. But they why on earth we needed to re-implement the same (?) logic specifically for remote_managed at https://github.com/openstack/nova/blob/d86916360858daa06164ebc0d012b78d19ae6497/nova/pci/devspec.py#L322-L334 ?16:08
melwittsean-k-mooney: oh cool. I'll go check the channel log16:08
sean-k-mooneygibi: remote managed does not otherwise allow the PF to be listed16:09
sean-k-mooneyif i recall correctly16:09
sean-k-mooneymelwitt: they did not come to a conclution either but were speculating it was related to hoave devstrack get invoke it and the #!/bin/bash lines in the hook script16:10
sean-k-mooneybut those point to bash too16:10
sean-k-mooneymelwitt: i think the fix is to stop usinging the raw module and actully use shell16:10
sean-k-mooneyi think the raw moduel is perhaps using a shle script to run your comamdn via the executabel you specify16:11
melwittsean-k-mooney: I wondered that exact same thing. the raw module seemed the only potentially suspicious thing to me16:11
sean-k-mooneyso raw is not ment to be used in normal code16:11
gibisean-k-mooney: thanks. I think I have not enough brain power for this right now16:11
sean-k-mooneyraw is intended to install the deps that are required for command and shell so that you can bootstrap an ansible target when its mising the basic required pacakages for ansibel to work16:12
melwittsean-k-mooney: yeah. supposedly saying executable=/bin/bash will make it run under bash and it must have used to but for some reason recently the behavior has changed16:12
sean-k-mooneyyes so looking at the ansible docs the args you would normally pass in ansibel you set as varibles in the string when you are doing an adhock command16:13
melwittI had been thinking about trying using the shell module instead to see if it helps16:13
melwittright16:13
sean-k-mooneyi kid of hate that we are invokeing ansibel form bash form ansible by the way16:13
melwitthaha yeah ...16:14
sean-k-mooneyit feels like it woudl better just to use ssh and be done with it or a script that we copy and be doen with it16:14
sean-k-mooneybut its been working this way for years16:14
sean-k-mooneyand now just sudenlly stopped16:15
sean-k-mooneygibi: one thing that is true is that we dont have gret test for all the valid way that you can list thing for the pci module16:15
melwittyeah. puzzling16:15
sean-k-mooneylike not clear human readble ones with comments at anyrate16:16
sean-k-mooneyso a lot fo the reasoning why x works is buried in my brain next to memroy of swaring at a terminal 16:17
* sean-k-mooney notes the pci whitelist code predates me working on nova :)16:18
* gibi is swearing too16:19
* sean-k-mooney i just work here dont look behind the curtain of how it works if you want to be happy on a firday16:19
sean-k-mooneygibi: all set for the sumit?16:19
gibiyepp, I'm well set. 16:20
sean-k-mooneymelwitt: are you goign to push a wip patch by the way to try using shell?16:20
sean-k-mooneygibi: do you have a presination ?16:20
gibiyepp16:21
gibiabout the qos work16:21
sean-k-mooneyah nice16:21
melwittsean-k-mooney: I will unless you wanted to. I have pretty much never used ansible and learning about it now16:21
sean-k-mooneyyou had one on the bandwith work before with migule right  so this is the evolution i guess16:21
gibisean-k-mooney: yes, it was in berlin too :)16:22
gibiso a nice cycle16:22
sean-k-mooneymelwitt: i never use it for adhoc comands its kind of like python that way. sure you can pass a string of python code to python and it will execute it but i only ever use it with a script so im really not familar iwth the adhock syntax16:23
gibiat that time we only had WIP code for booting with bandwidth qos ports. No we have a lot more16:23
melwittsean-k-mooney: yeah I was just reading the docs yesterday. thought it might be fun to try16:24
sean-k-mooneythis one https://docs.ansible.com/ansible/latest/user_guide/intro_adhoc.html16:24
chateaulavgibi: im excited to see it16:25
sean-k-mooneygibi: yep its amazing what can hapne over the course of a pandemic...16:25
chateaulavits my first time to berlin16:25
melwittsean-k-mooney: thanks16:25
gibisean-k-mooney: I honestly want to fix the pci spec parsing code before I start adding to it, but it is soo full of edge cases that I start feeling the I cannot clean it up without potentially breaking couple of cases that works somehow today16:26
gibichateaulav: o/16:26
sean-k-mooneygibi: ya am if i get the vdpa stuff landed perhaps i can help add some more testcsaes to prove out some of those edgecases16:27
sean-k-mooneyliek we can test the parsign with the libvirt fuctional tests16:27
sean-k-mooneywe shoudl not need that level of testing to parse stings but that would be one way to validate teh edge  cacses16:28
sean-k-mooneygibi: in your case you are not really chanign the filtering 16:29
sean-k-mooneygibi: you are just addign new tags for resocue class and traits16:29
sean-k-mooneyso if you avoid refactoring and just focus on that it shoud eb pretty safe16:30
sean-k-mooneythe new tags you are adding will not filter in/out any devices16:30
sean-k-mooneyso you can mostly just operate on the objects that are constructed after the parsing is done16:31
sean-k-mooneymelwitt: by the way on later branches there are zuul rules to do this16:33
melwittsean-k-mooney: oh nice16:33
sean-k-mooneyhttps://github.com/SeanMooney/ansible_role_devstack/blob/master/ansible/deploy_multinode_devstack.yaml#L133-L136=16:33
sean-k-mooneysync-controller-ceph-conf-and-keys,16:33
sean-k-mooneyhttps://github.com/openstack/devstack/tree/master/roles/sync-controller-ceph-conf-and-keys16:34
sean-k-mooneyactully its in devstack16:35
sean-k-mooneyim kind of confused why we are donign this in hte gate hook 16:35
gibisean-k-mooney: I promised to reject the devname case in the new codepath, for that I need to spearate out the devname handling codepath from the current parsing16:36
sean-k-mooneyam not really you jus need to check if the whitelist containds devname16:36
sean-k-mooneygibi: if you realy wanted too you could also punt on that for this cycle16:37
gibiyepp that does not allows removing that devname codepath from the complexity picture16:37
sean-k-mooneywell devname does not realy change anything for the rest of the spec or code16:37
gibiI just add the guard condition, but the code below will still depend on devname16:37
gibiso the code below will as today, complex, uggly, :/16:38
sean-k-mooneyyep but that wont affect tracking the dvices in placment16:38
gibiand I just pile on that with the guard16:38
gibiI don't like piling on hard to comprehend code16:38
sean-k-mooneywhat im saying is if you dont feel like you will have time to adress the devname part16:38
sean-k-mooneywe could put that out os scope fo the spec an come back to it16:38
sean-k-mooneyafter16:38
sean-k-mooneyi.e. we can deprecate and remove it later16:39
gibiyeah I got you. But I feel that this is a good change to clean it up. I don't know when will be the next time we come back to this16:40
sean-k-mooneyyep it certenly is16:41
sean-k-mooneynext time would proably be either when we make this on by default or add support for neutron prots16:42
sean-k-mooneyyou are right that we likely wont just go clean tthis up if its not in aide of something else16:42
gibiyepp16:43
sean-k-mooneymelwitt: by the way im really confused why the hook works the way it does16:46
sean-k-mooneylike we are runnign devstack on all the noes anyway so im not sure why se need the gate hook to reach out to all the other subnodes and execute command on them to deploy ceph16:47
sean-k-mooneyyou know instead of just runign the ceph plugin as part of the devstack run on thsoe nodes16:47
sean-k-mooneymelwitt: we are using the shell module elsewhere in the file by the way16:48
sean-k-mooneyhttps://github.com/openstack/nova/blob/stable/train/gate/live_migration/hooks/ceph.sh#L6216:48
sean-k-mooneymelwitt: just in case your wondering the fist thing after $ansible is the inventory group16:49
sean-k-mooney  $ANSIBLE subnodes --become -f 5 -i "$WORKSPACE/inventory" -m raw -a "executable=/bin/bash16:49
sean-k-mooneyso this is runnign the scrip on all the hosts in the subnodes group wich is jut the compute node16:49
sean-k-mooneyso ANSIBLE subnodes --become -f 5 -i "$WORKSPACE/inventory" -m shell -a "... would be the same16:50
sean-k-mooneybecause this is using "" not '' $BASE is evaulatedin the context fo prepare_ceph too16:51
sean-k-mooneyhttps://github.com/openstack/nova/blob/stable/train/gate/live_migration/hooks/ceph.sh#L10-L17=16:52
sean-k-mooneywhich i dont know if that was inteded or not16:52
sean-k-mooneyi guess so as it woud not otherwise be defiend16:52
sean-k-mooneyanyway i think im going to call it a day if you want me to take a look at this on monday then i can16:53
melwittsean-k-mooney: thank you for all the hints16:54
opendevreviewmelanie witt proposed openstack/nova stable/train: DNM Testing for ceph setup gate fail  https://review.opendev.org/c/openstack/nova/+/84453017:21
opendevreviewBalazs Gibizer proposed openstack/nova master: Unparent PciDeviceSpec from PciAddressSpec  https://review.opendev.org/c/openstack/nova/+/84449117:33
opendevreviewBalazs Gibizer proposed openstack/nova master: Fix PciAddressSpec descendants to call super.__init__  https://review.opendev.org/c/openstack/nova/+/84456517:33
opendevreviewBalazs Gibizer proposed openstack/nova master: Add more test coverage for devname base dev spec  https://review.opendev.org/c/openstack/nova/+/84462517:33
opendevreviewBalazs Gibizer proposed openstack/nova master: test remote managed dev spec with wildcard address  https://review.opendev.org/c/openstack/nova/+/84462617:33
opendevreviewBalazs Gibizer proposed openstack/nova master: Poison /sys access in test  https://review.opendev.org/c/openstack/nova/+/84462717:33
opendevreviewBalazs Gibizer proposed openstack/nova master: More comment in the code  https://review.opendev.org/c/openstack/nova/+/84462817:33
opendevreviewMerged openstack/nova stable/yoga: Add missing condition  https://review.opendev.org/c/openstack/nova/+/84382019:14

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!