Thursday, 2020-01-23

*** tbachman has quit IRC00:00
*** ociuhandu has joined #openstack-nova00:01
*** ociuhandu has quit IRC00:05
*** artom has joined #openstack-nova00:13
*** TxGirlGeek has quit IRC00:26
*** jaosorior has quit IRC00:28
*** TxGirlGeek has joined #openstack-nova00:31
*** TxGirlGeek has quit IRC00:36
*** jmlowe has joined #openstack-nova00:39
*** luksky has quit IRC00:47
*** jmlowe has quit IRC00:54
*** jmlowe has joined #openstack-nova01:12
*** damien_r has joined #openstack-nova01:16
*** damien_r has quit IRC01:17
*** damien_r has joined #openstack-nova01:17
*** jmlowe has quit IRC01:18
*** tbachman has joined #openstack-nova01:19
*** gyee has quit IRC01:52
*** tbachman has quit IRC02:09
*** damien_r has quit IRC02:19
*** adriant has quit IRC02:41
*** adriant has joined #openstack-nova02:43
*** tbachman has joined #openstack-nova02:47
*** artom has quit IRC02:50
*** damien_r has joined #openstack-nova03:01
*** damien_r has quit IRC03:05
*** TxGirlGeek has joined #openstack-nova03:33
*** tbachman has quit IRC03:35
*** psachin has joined #openstack-nova03:36
*** TxGirlGeek has quit IRC03:39
*** udesale has joined #openstack-nova04:11
*** TxGirlGeek has joined #openstack-nova04:16
*** mkrai_ has joined #openstack-nova04:23
*** TxGirlGeek has quit IRC04:24
*** jhesketh has quit IRC04:54
*** jhesketh has joined #openstack-nova04:55
*** macz has joined #openstack-nova05:00
*** macz has quit IRC05:05
*** ratailor has joined #openstack-nova05:21
*** evrardjp has quit IRC05:34
*** evrardjp has joined #openstack-nova05:34
*** TxGirlGeek has joined #openstack-nova05:46
*** TxGirlGeek has quit IRC05:55
*** links has joined #openstack-nova05:56
*** tkajinam has quit IRC06:01
*** tkajinam has joined #openstack-nova06:03
*** ccamacho has quit IRC06:26
*** mkrai_ has quit IRC06:42
*** _rajat_ has joined #openstack-nova06:47
*** mkrai_ has joined #openstack-nova06:47
*** jaosorior has joined #openstack-nova06:47
*** tkajinam_ has joined #openstack-nova07:01
*** tkajinam has quit IRC07:03
*** heso has joined #openstack-nova07:12
*** Luzi has joined #openstack-nova07:21
*** tkajinam__ has joined #openstack-nova07:47
*** iurygregory has joined #openstack-nova07:48
*** tkajinam_ has quit IRC07:50
*** ccamacho has joined #openstack-nova07:50
*** luksky has joined #openstack-nova07:53
*** tesseract has joined #openstack-nova07:56
*** aarents has quit IRC07:57
*** f0o has quit IRC07:58
*** rmk has quit IRC07:59
*** fungi has quit IRC07:59
*** fnordahl has quit IRC07:59
*** gouthamr has quit IRC07:59
*** owalsh has quit IRC07:59
*** Roamer` has quit IRC07:59
*** toanster has quit IRC07:59
*** frickler has quit IRC07:59
*** Jeffrey4l has quit IRC07:59
*** cz3 has quit IRC07:59
*** cmurphy has quit IRC07:59
*** Anticimex has quit IRC07:59
*** yankcrime has quit IRC07:59
*** dtantsur|afk has quit IRC07:59
*** johnthetubaguy has quit IRC08:01
*** Jeffrey4l has joined #openstack-nova08:02
*** cz3 has joined #openstack-nova08:02
*** yankcrime has joined #openstack-nova08:02
*** cmurphy has joined #openstack-nova08:02
*** Anticimex has joined #openstack-nova08:02
*** dtantsur|afk has joined #openstack-nova08:02
*** irclogbot_3 has quit IRC08:04
*** irclogbot_2 has joined #openstack-nova08:05
*** bnemec has joined #openstack-nova08:05
*** dansmith has quit IRC08:06
*** dansmith has joined #openstack-nova08:08
*** aarents has joined #openstack-nova08:10
*** maciejjozefczyk has joined #openstack-nova08:10
*** tosky has joined #openstack-nova08:13
*** mkrai_ has quit IRC08:20
*** tkajinam__ has quit IRC08:29
*** tkajinam has joined #openstack-nova08:29
openstackgerritMerged openstack/nova master: Func test for failed and aborted live migration  https://review.opendev.org/69906508:31
openstackgerritMerged openstack/nova master: Remove unused code  https://review.opendev.org/70340808:31
*** udesale_ has joined #openstack-nova08:31
*** mkrai_ has joined #openstack-nova08:32
*** udesale_ has quit IRC08:33
*** udesale_ has joined #openstack-nova08:33
*** udesale has quit IRC08:34
*** tkajinam has quit IRC08:35
*** bnemec has quit IRC08:41
*** rpittau|afk is now known as rpittau08:43
*** ociuhandu has joined #openstack-nova08:45
*** rmk has joined #openstack-nova08:45
*** fungi has joined #openstack-nova08:45
*** f0o has joined #openstack-nova08:45
*** fnordahl has joined #openstack-nova08:45
*** gouthamr has joined #openstack-nova08:45
*** owalsh has joined #openstack-nova08:45
*** Roamer` has joined #openstack-nova08:45
*** toanster has joined #openstack-nova08:45
*** frickler has joined #openstack-nova08:45
*** openstackstatus has quit IRC08:48
*** bnemec has joined #openstack-nova08:50
*** martinkennelly has joined #openstack-nova08:51
*** udesale_ has quit IRC08:54
*** ociuhandu has quit IRC08:56
*** mrch has joined #openstack-nova08:58
*** udesale has joined #openstack-nova09:02
*** bnemec has quit IRC09:08
*** xek has joined #openstack-nova09:09
*** bbowen has quit IRC09:24
*** bbowen has joined #openstack-nova09:24
openstackgerritStephen Finucane proposed openstack/nova master: Remove remaining Python 2.7-only dependencies  https://review.opendev.org/69889509:27
*** pcaruana has joined #openstack-nova09:30
*** bnemec has joined #openstack-nova09:33
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Update API reference guide  https://review.opendev.org/70379609:33
openstackgerritStephen Finucane proposed openstack/nova master: Remove 'nova.image.api' module  https://review.opendev.org/70245109:33
openstackgerritStephen Finucane proposed openstack/nova master: WIP: nova-net: Remove unused nova-network objects  https://review.opendev.org/69715609:33
*** dtantsur|afk is now known as dtantsur09:36
*** iurygregory has quit IRC09:40
*** Liang__ has joined #openstack-nova09:43
*** ratailor has quit IRC09:44
*** ratailor has joined #openstack-nova09:51
*** derekh has joined #openstack-nova09:58
*** iurygregory has joined #openstack-nova10:01
*** jistr has quit IRC10:06
*** jistr has joined #openstack-nova10:06
*** jaosorior has quit IRC10:08
*** zbr has left #openstack-nova10:09
*** sridharg has joined #openstack-nova10:12
*** dviroel has joined #openstack-nova10:42
*** sridharg has quit IRC10:43
*** udesale has quit IRC10:56
*** bnemec has quit IRC11:02
*** tbachman has joined #openstack-nova11:04
*** sridharg has joined #openstack-nova11:06
*** sridharg has quit IRC11:11
*** rchurch has quit IRC11:18
openstackgerritMerged openstack/nova master: Switch to hacking 2.x  https://review.opendev.org/69573311:21
openstackgerritMerged openstack/nova master: requirements: Limit hacking to one minor version  https://review.opendev.org/70340511:21
*** rpittau is now known as rpittau|bbl11:21
*** sridharg has joined #openstack-nova11:24
*** Liang__ has quit IRC11:28
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove unused nova-network objects  https://review.opendev.org/69715611:32
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove unnecessary exception handling  https://review.opendev.org/70397311:32
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove unused parameters  https://review.opendev.org/70397411:32
stephenfindansmith: I'm doing possibly bad stuff with objects in https://review.opendev.org/697156. I _think_ it's okay but would appreciate your input when you're around11:34
*** bbowen has quit IRC11:34
stephenfindansmith: Also want to figure out how to move objects between projects (nova -> os-vif) since that's seems like an obvious thing to do in that case11:34
stephenfinsean-k-mooney: Maybe relevant, idk ^11:34
sean-k-mooneywe should not move the network object to os-vif11:43
sean-k-mooneywe already have one11:43
sean-k-mooneythe os vif objects are constucted from info often taken form the nova ones but they are not the same objects11:44
*** tbachman has quit IRC11:49
stephenfinsean-k-mooney: Damn. Okay, maybe you can figure out what's going on. Fancy pulling down the earlier version of that change along with this patch for o.vo https://review.opendev.org/#/c/703970/ ?11:57
gibikashyap: rng patch looks good to me, thanks for the update11:58
lucidguysean-k-mooney:  The only option I see regarding hugepages is reserved_huge_pages.  Is that what you were talking about?11:58
stephenfinIf you install that version of the o.vo in the venv and run 'tox -e py36 -- -n nova/tests/unit/objects/test_objects.py::TestObjectVersions.test_obj_make_compatible', it'll fail saying it can't find the 'Network' object11:58
sean-k-mooneystephenfin: i know dansmith found that the namespace stuff does not seam to be working properly11:58
stephenfindid he now?11:59
stephenfinany notes or bugs on that?11:59
sean-k-mooneylucidguy: no, hugepages is a afeature that is enabled in the flavor/image by setting hw:mem_page_size=large11:59
sean-k-mooneyreserved_huge_pages is intended to allow you say that some of the hugepages on the host are reserved for host use and cannot be used by vms12:00
sean-k-mooneystephenfin: it cam up on one of his patch let me see if it merged or not12:01
lucidguysean-k-mooney:  Understood12:01
sean-k-mooneystephenfin: https://review.opendev.org/#/c/702049/12:01
sean-k-mooneythe ephemeral object patch12:01
sean-k-mooneystephenfin: line 1175 https://review.opendev.org/#/c/702049/4/nova/tests/unit/objects/test_objects.py12:02
sean-k-mooneylucidguy: i think these are the docs on hugepage that you want https://docs.openstack.org/nova/latest/admin/huge-pages.html12:03
sean-k-mooneylucidguy: we have summit presintaion on this too that stephenfin did12:03
lucidguyI was just looking at that link12:04
lucidguyCool12:05
kashyapgibi: Hi, thanks for looking!12:07
kashyapgibi: Want me to respin removing the stray word "this"?12:09
* kashyap --> lunch12:09
gibikashyap: if you respin quickly then I can re +2 it. but I'm fine merging the patch as is12:09
sean-k-mooneystephenfin: i think you need to update https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_objects.py#L122512:09
stephenfinsean-k-mooney: incoming :)12:09
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Make the security group API a module  https://review.opendev.org/69674612:10
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove remaining nova-network quotas  https://review.opendev.org/69715212:10
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove use of legacy 'FloatingIP' object  https://review.opendev.org/69715312:10
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove use of legacy 'Network' object  https://review.opendev.org/69715412:10
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove use of legacy 'SecurityGroup' object  https://review.opendev.org/69715512:10
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove now unnecessary nova-net workaround  https://review.opendev.org/70244012:10
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Update API reference guide  https://review.opendev.org/70379612:10
openstackgerritStephen Finucane proposed openstack/nova master: Remove 'nova.image.api' module  https://review.opendev.org/70245112:10
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove unnecessary exception handling  https://review.opendev.org/70397312:10
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove unused parameters  https://review.opendev.org/70397412:10
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove unused nova-network objects  https://review.opendev.org/69715612:10
sean-k-mooneystephenfin: change  base.NovaObjectRegistry.obj_classes() to  call get_nova_objects https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_objects.py#L115912:10
stephenfinyou b*****12:10
stephenfingdi git12:10
stephenfin:(12:10
sean-k-mooneyso it only compares the version of hte nova object not also the os-vif ones12:11
sean-k-mooneythat is why you are having issues with the network object12:11
stephenfinYup, makes sense. Thanks :)12:11
stephenfinShould be fixed now12:11
sean-k-mooneyalong with a rebase :P12:11
stephenfinI tab-completed 'git rebase --committer-date-is-author-date' instead of 'git rebase --continue' :(12:12
stephenfin(had already finished the rebase)12:12
sean-k-mooneyah ok12:12
sean-k-mooneywell you could also delete the nova network object now right?12:13
stephenfinyup, done https://review.opendev.org/#/c/697156/25/nova/objects/network.py12:13
sean-k-mooneyor is there still nova code using that12:13
sean-k-mooneyoh then you missed removing it form the object versions12:13
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove unused nova-network objects  https://review.opendev.org/69715612:14
stephenfinwhoops12:14
sean-k-mooneyline 1102 https://review.opendev.org/#/c/697156/25/nova/tests/unit/objects/test_objects.py12:14
stephenfinta12:14
*** jaosorior has joined #openstack-nova12:14
sean-k-mooneythe thing that uses that might also need the get_nova_object change12:14
stephenfinIf it doesn't, it's not obvious12:15
stephenfinas in there are no tests failing fwict12:15
*** rcernin has quit IRC12:15
sean-k-mooneyya so the issue is all this stuff was set up befor namespaces were added to ovos12:16
sean-k-mooneyso it was really not expecting ot have two object with the same name but be in differne namespaces12:16
jrollefried: cool, thanks, just wanted to make sure you weren't waiting on me for something in particular there :)12:17
openstackgerritKashyap Chamarthy proposed openstack/nova master: libvirt: Add a default VirtIO-RNG device to guests  https://review.opendev.org/61660312:18
kashyapgibi: --^ Done12:18
gibikashyap: thanks, +212:19
gibistephenfin: if you have time there is an easy patch  to +2  https://review.opendev.org/61660312:19
*** rcernin has joined #openstack-nova12:20
*** mgariepy has joined #openstack-nova12:22
stephenfinkashyap: If you can you address my nit on the releasenote, I'm +212:22
stephenfingibi too ^12:22
kashyapstephenfin: Yeah, I already hesitated about that first bit.  As the URL will take care12:23
kashyapLet me do it real quick12:23
*** jaosorior has quit IRC12:24
* gibi is waiting patiently for the next PS :)12:24
openstackgerritKashyap Chamarthy proposed openstack/nova master: libvirt: Add a default VirtIO-RNG device to guests  https://review.opendev.org/61660312:25
stephenfinta. +212:25
kashyapAlright, fixed the reno.12:25
*** jaosorior has joined #openstack-nova12:25
kashyapgibi: Your patient wait is over :D12:26
*** rcernin has quit IRC12:26
* gibi is waiting for gerrit to load :D12:27
gibidone12:27
gibitoday something is wrong with my uplink12:27
*** bnemec has joined #openstack-nova12:29
openstackgerritBalazs Gibizer proposed openstack/nova stable/stein: Mask the token used to allow access to consoles  https://review.opendev.org/70218112:31
gibielod: fixed your comments in ^^12:31
sean-k-mooneystephenfin: have you seen issues with nova.tests.functional.test_nova_manage.TestDBArchiveDeletedRowsMultiCell failing out of interest?12:37
stephenfinI haven't paid attention to it, no12:37
*** jawad_axd has joined #openstack-nova12:37
sean-k-mooneyok i was wondering if that was the failing db test ye were talking about yesterday12:38
sean-k-mooneyi guess not12:38
*** bbowen has joined #openstack-nova12:38
elodgibi: thx, looking12:39
*** ociuhandu has joined #openstack-nova12:46
*** ociuhandu has quit IRC12:46
*** mkrai_ has quit IRC12:48
*** _rajat_ has quit IRC12:50
stephenfinsean-k-mooney: Don't know if I showed this to you before Xmas or not https://github.com/testing-cabal/subunit/pull/4012:54
*** ociuhandu has joined #openstack-nova12:54
stephenfinI don't understand Python's IO model well enough to come up with better, but that fixed things for me for https://review.opendev.org/#/c/682111/ anyway12:55
stephenfinwhoops12:55
sean-k-mooney you did not but ill take a look at both12:55
stephenfinhttps://review.opendev.org/#/c/700522/12:56
sean-k-mooneywhy are you importing the print fucntion explcitly12:56
sean-k-mooneyare you using py26 lol12:56
sean-k-mooneyyou can still do that but it was never needed in py2712:57
stephenfinit definitely is/was :)12:58
sean-k-mooneybut ya does that work12:58
stephenfinto use the 'file=foo' thing, anyway12:58
*** ociuhandu has quit IRC12:58
sean-k-mooneyoh ya that is python3 only12:58
stephenfin'zactly12:59
sean-k-mooneybut again nova has droped python 2 support so :P12:59
stephenfinso that nova patch just prints a load of junk that overwhelms subunit12:59
stephenfinwithout my subunit change, it craps out with the subunit parser error12:59
stephenfinwith it, it still craps out but with a proper error12:59
jawad_axdQuick question guys: What does cpu allocation 2.0 means ? Is it 2:1?13:00
*** rpittau|bbl is now known as rpittau13:00
stephenfinsaying the packet is > 4k (I think)13:00
*** damien_r has joined #openstack-nova13:01
jawad_axdand it is safe to change cpu allocation ratio in running environment?13:01
stephenfinsorry, 4M https://github.com/testing-cabal/subunit/blob/master/python/subunit/v2.py#L202-L20813:01
*** damien_r has quit IRC13:01
stephenfinjawad_axd: yup, 2:1 (20 enabled host CPUs = 40 VCPU inventory)13:01
*** damien_r has joined #openstack-nova13:02
stephenfinjawad_axd: It should be safe to change so long as you don't lower it to the point that there's less inventory available than you have used13:02
*** nweinber__ has joined #openstack-nova13:02
stephenfinI'm not actually sure what would happen then. I assume the periodic task to update placement's inventory would start failing13:02
stephenfineasily tested in a pre-prod environment :)13:03
sean-k-mooneystephenfin: ok well if we have a repoducer that means we have a chance of fixing it13:03
sean-k-mooneywe dont need nova at all we can repodcuse this in stestr13:03
stephenfinExactly. That's pretty damn consistent13:04
stephenfinWe should be able to but I couldn't do so when I tried13:04
stephenfinProbably didn't have the correct fixtures configured or something13:04
sean-k-mooneyhum ok well i might take a look at this later. and see if i can figure something out13:05
sean-k-mooneyif we can create a simpler repoducer that woudl be good if not this works13:05
jawad_axd@stephenfin Thanks. One more thing, In horizon I see cpu's under compute(hypervisor ) tab as 24cpu. So with 2.0 cpu allocation ratio, I should be able to use 24x2=48 vcpus, right? But I am only able to use 24 cpus.What you say about it?13:06
stephenfinI imagine Horizon is pulling that info from the os-hypevisor API which doesn't take overcommit ratios into account13:07
sean-k-mooneyi think so too13:07
sean-k-mooneyi think horizon is showing the correct value13:08
sean-k-mooneythe allocation raitio can be change per host it would not be resonable to have to compute it differnetly per hosts13:08
jawad_axdIf horizon is showing the same cpu's on host. Does it mean, overcommitment is not being used/applied ?13:09
sean-k-mooneyno13:10
sean-k-mooneyover commit is calulated in the scheduler/placment13:10
sean-k-mooneyit should not be see in horizon13:10
stephenfinwhat sean-k-mooney said13:10
stephenfinHorizon might go into negative available values (I'm not sure) but instances will still be scheduled13:10
sean-k-mooneyjawad_axd: if you are using a recent version of openstack and look at the placment RP for the host you will see an cpu inventory where the total = the number of cores on the host and the allocation ratio will be 2.013:11
*** luksky has quit IRC13:11
sean-k-mooneywell on my home system it currently shos 44/2413:11
sean-k-mooneyso it does not go into negitiv but the used can exceed the available13:12
jawad_axdI am just stuck because I can not use more than 24 cpus, while  with overcommitment I should be able to use 48.13:12
jawad_axdI am using stein13:12
sean-k-mooneyyou cannot use more the 24cpu in one guest or in general13:12
jawad_axdIn general.13:13
sean-k-mooneydid you set the cpu allcoation ratio in the compute node config13:13
stephenfinalso, how are you creating the instance? Via the Horizon UI or on the CLI?13:13
jawad_axd# Scheduler13:13
jawad_axdcpu_allocation_ratio = 2.013:13
sean-k-mooneyya that does not work in stien13:13
jawad_axdFrom both, cli and horizon13:13
sean-k-mooneyyou hage to set it per compute node13:13
jawad_axdah ok.13:13
jawad_axdhow to set  per compute node? Can you give me some pointers plz?13:14
sean-k-mooneyhttps://docs.openstack.org/nova/latest/configuration/config.html#DEFAULT.cpu_allocation_ratio13:14
sean-k-mooneyyou just set cpu_allocation_ratio in the default section13:15
sean-k-mooneyin the nova.conf13:15
sean-k-mooneyalternitivly if you want to manage it via the placement api you can use initial_cpu_allocation_ratio13:15
jawad_axdSo  cpu_allocation_ratio in nova.conf at each compute node will work for that compute node?13:18
sean-k-mooneycorrect it should13:18
sean-k-mooneywhat i belive is happening is that the allcoation raition in the  placement inventory is set to 113:19
sean-k-mooneyand that is causeign the host to be eliminated before you get to the schduler13:19
sean-k-mooneywhere you set it to 213:19
sean-k-mooneyyou could check that13:19
sean-k-mooneyyou can do "openstack resource provider list"13:20
*** tbachman has joined #openstack-nova13:20
sean-k-mooneythen do "openstack resource provider inventory show <RP uuid>"13:20
roukcant use 24 vcore or more per guest? when was this? i have 32vcore guests.13:20
sean-k-mooneyyou can never have more cores in the guest then are on the host but you can have multple guests sharing cores on the host13:21
*** belmoreira has joined #openstack-nova13:22
sean-k-mooneyso on my home server i currently have 44 guest cpus spread over 20 host cpus13:22
sean-k-mooneybut i could not boot a 44 core vm13:22
*** udesale has joined #openstack-nova13:23
jawad_axd@sean-k-mooney ok13:23
*** mkrai_ has joined #openstack-nova13:24
*** ociuhandu has joined #openstack-nova13:24
jawad_axd@sean-k-mooney Yeah "openstack resource provider inventory show 04b5aba0-ca20-4e15-881b-eaca3ce73db7 VCPU" show overcommitment 24x2=48. So probably, I was using more than 24cpus for one guest which I cant do as you said.13:29
*** ociuhandu has quit IRC13:29
*** yiipi has joined #openstack-nova13:30
sean-k-mooneyjawad_axd: there are some other constrats as well13:33
sean-k-mooneyif you are using cpu pinning you cannot over commit13:33
sean-k-mooneyand if you are using a virtual numa toplgoy then within each numa node the cpus cant be over commited13:33
jawad_axdok my hosts have mostly this architecture, Can you have a look at it plz. http://paste.openstack.org/show/788715/13:35
sean-k-mooneyya so you have 4 numa nodes and 12 cores per numa node13:36
sean-k-mooneywell 12 threads13:36
jawad_axdDoes it allow overcommitment ?13:36
sean-k-mooneysure13:37
sean-k-mooneybut it depends on the vm13:37
sean-k-mooneyif the vm has a numa toplogy it impacts how the vm can be placed13:37
jawad_axdIf vms are using cpu host-model, then?13:37
sean-k-mooneyno13:37
roukits a flag, since numa is not at all free.13:37
sean-k-mooneyif the vm flavor has hw:cpu_polciy=dedeicated13:37
sean-k-mooneyhw:mem_page_size or hw:numa_nodes set or is using vPMEM13:38
jawad_axdI am not using numa virtually.13:38
sean-k-mooneythen it will have a numa toplogy13:38
*** rouk has quit IRC13:38
sean-k-mooneyin that case only the gloabl host core count and allcoation ration for that host matter13:38
jawad_axdSo I guess, it should be fine for me then.13:39
*** tbachman has quit IRC13:39
sean-k-mooneyyep13:39
jawad_axdThanks !!!13:39
*** yiipi has quit IRC13:39
sean-k-mooneyif you set hw:numa_nodes=1 you are basicaly saying i want a guest with 1 numa node and map it to a single host numa node. if you enbale hugepage or cpu pinning it imples hw:numa_nodes=113:40
*** rouk has joined #openstack-nova13:40
sean-k-mooneyso with that toploy you would only be able to have at most 12 cores in a vm13:40
sean-k-mooneybut since you are not using numa for the vms it does not matter in your case13:41
jawad_axdOkay..13:41
sean-k-mooneythe numa feature are just for performance tuning13:42
sean-k-mooneyif you dont need them you can ignore them13:42
*** tbachman has joined #openstack-nova13:42
jawad_axdI dont need them now, maybe in future.13:42
*** ratailor has quit IRC13:43
*** shilpasd has joined #openstack-nova13:47
*** jaosorior has quit IRC13:47
*** mkrai_ has quit IRC13:50
efriednova meeting now-ish13:59
*** luksky has joined #openstack-nova13:59
shilpasdefried: hi13:59
efriedhello13:59
*** ociuhandu has joined #openstack-nova14:00
shilpasdefried: dansmith: here added one question, https://review.opendev.org/#/c/694462/8/nova/db/sqlalchemy/api.py@200814:03
shilpasdhelp me to understand 'any issues in setting 'volume_backed' field on the existing instances in the upgrade db script?'14:04
*** eharney has quit IRC14:04
efriedshilpasd: dansmith is on vacation this week, and I'm not really qualified to discuss this one. It's Chinese New Year around now, so I'm not sure if alex_xu is around either.14:05
efriedIOW this may have to wait until next week to close on.14:05
shilpasdefried: ok, thanks14:05
*** rchurch has joined #openstack-nova14:05
shilpasdwill wait for them14:06
*** sridharg has quit IRC14:07
*** jmlowe has joined #openstack-nova14:09
*** mgariepy has quit IRC14:10
*** rchurch has quit IRC14:10
*** mgariepy has joined #openstack-nova14:12
sean-k-mooneywell that was quick :)14:12
shilpasd:] (y)14:12
sean-k-mooneylooking at dansmith's suggestion i tened to agree that offloadign this to sql or adding a new field would be better14:13
sean-k-mooneyyou could perhaps store the "is_volume_backed" info in the system_metadata table14:16
sean-k-mooneythat woudl avoid the need for a schema change but i think option 1 makes the most sense too14:17
sean-k-mooneythe only thing is we would have to keep the logic to handel old instance for a few release for FFU14:17
sean-k-mooneybut eventurlaly we could drop it14:18
shilpasdsean-k-mooney: thanks for your opinion14:18
shilpasdplease clear me last point 'keep the logic to handel old instance for a few release for FFU'14:18
*** rchurch has joined #openstack-nova14:18
*** jmlowe has quit IRC14:18
sean-k-mooneywe "support" skip level upgrade via Fast Forward upgrdes at lest downstream in OSP/tripleo14:19
sean-k-mooneyso if we add the code handel loading old instances and calulating the value14:20
sean-k-mooneywe need to keep it for a cople of release before we can remove it14:20
sean-k-mooneyand safly assume it will always be set14:20
sean-k-mooneyeventually that code for old instance can be removed but not for 2-3 releases14:20
*** alistarle has joined #openstack-nova14:21
shilpasdfor old instances will update 'volume_backed' based on bdms.volume and bdms.boot_index14:22
*** rchurch has quit IRC14:23
*** jmlowe has joined #openstack-nova14:23
shilpasdso still not clear why we need to keep code for couple of release, please elaborate further14:23
sean-k-mooneyits technically also stored in the request spec for instance created after rocky14:23
sean-k-mooneyFFU do not require the compute agents to start during the upgrade14:24
sean-k-mooneyso you can go form qeens to train without ever starting the agent untill you get to train14:24
sean-k-mooneyso unless you are groin to write an online data migration for this14:25
sean-k-mooneywhich is not what dansmith was suggesting14:25
sean-k-mooneyif you take the approch of fixing it when you load the objects you need to keep that code for a few releases to not force peopel that are doing FFU to stop in ussuri14:25
sean-k-mooneystart the agent and have it fix the instnaces14:25
sean-k-mooneyso all i was really saying is eventaully we can remove that compatiablity code for old instnace but we will have to wait a releast or 314:27
*** rchurch has joined #openstack-nova14:27
*** nweinber__ has quit IRC14:28
*** mkrai_ has joined #openstack-nova14:28
shilpasdok, got it14:29
*** mgariepy has quit IRC14:30
*** Sundar has joined #openstack-nova14:32
*** Luzi has quit IRC14:32
*** links has quit IRC14:35
*** Sundar has quit IRC14:38
*** ociuhandu has quit IRC14:39
*** ociuhandu has joined #openstack-nova14:40
*** mgariepy has joined #openstack-nova14:43
*** jawad_axd has quit IRC14:45
*** ociuhandu has quit IRC14:46
*** alistarle has quit IRC14:47
*** iurygregory has quit IRC14:49
*** jawad_axd has joined #openstack-nova14:49
*** bnemec has quit IRC14:51
*** jawad_axd has quit IRC14:54
*** shilpasd has quit IRC15:01
*** tbachman has quit IRC15:01
*** mriedem has joined #openstack-nova15:02
*** pcaruana has quit IRC15:04
*** damien_r has quit IRC15:06
*** damien_r has joined #openstack-nova15:06
sean-k-mooneyby the way we can recheck stuff currently right15:09
sean-k-mooneythere is not specific gate blockers at the moment out side of the intermitenly failing jobs15:09
*** eharney has joined #openstack-nova15:10
*** efried has quit IRC15:12
*** lpetrut has quit IRC15:12
*** priteau has joined #openstack-nova15:12
*** nweinber__ has joined #openstack-nova15:12
*** psachin has quit IRC15:13
*** efried has joined #openstack-nova15:13
*** ociuhandu has joined #openstack-nova15:23
*** nweinber__ has quit IRC15:24
*** nweinber__ has joined #openstack-nova15:27
*** nweinber_ has joined #openstack-nova15:29
*** nweinber__ has quit IRC15:31
*** tbachman has joined #openstack-nova15:35
*** priteau has quit IRC15:51
*** artom has joined #openstack-nova15:52
*** lbragstad has quit IRC15:54
*** mlavalle has joined #openstack-nova16:00
*** lbragstad has joined #openstack-nova16:00
stephenfinsean-k-mooney: not that I'm aware of, no16:03
*** alex_xu has quit IRC16:03
*** jmlowe has quit IRC16:08
*** gyee has joined #openstack-nova16:14
*** mkrai_ has quit IRC16:18
*** jmlowe has joined #openstack-nova16:21
*** eharney has quit IRC16:22
*** mrch has quit IRC16:23
*** jawad_axd has joined #openstack-nova16:24
KeithMnemonicgood morning melwitt: should i keep playing recheck roulette for that patch on stable/pike or wait to see if the issue with migration might be found?16:25
*** ccamacho has quit IRC16:26
melwittKeithMnemonic: I dunno what the pass rate for that job is on stable/pike at this point but it seems pretty terrible. sean-k-mooney and I looked at it at bit yesterday and no clues yet what's going wrong16:28
*** tosky has quit IRC16:28
KeithMnemonicare you ok with me continuing rechecks to see if it passes eventually16:29
*** TxGirlGeek has joined #openstack-nova16:29
openstackgerritMerged openstack/nova master: nova-net: Make the security group API a module  https://review.opendev.org/69674616:30
*** lbragstad has quit IRC16:30
melwittKeithMnemonic: yeah, you can16:32
KeithMnemonicthanks16:32
*** jmlowe has quit IRC16:35
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove unnecessary exception handling  https://review.opendev.org/70397316:36
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove unused parameters  https://review.opendev.org/70397416:36
openstackgerritStephen Finucane proposed openstack/nova master: nova-net: Remove unused nova-network objects  https://review.opendev.org/69715616:36
stephenfinefried: You saw https://zuul.opendev.org/t/openstack/build/471f2388615e4ee5806a85e19b79569d ?16:38
stephenfinHave you a patch up to address that yet?16:38
* efried clicks16:38
stephenfinSeeing as you were attacking them in a piecemeal fashion?16:38
efriedstephenfin: I hadn't seen that. I can throw out a patch quick...16:39
sean-k-mooneymelwitt: i got side tracked with backports so i did not look at it properly so ill try to do it today16:39
stephenfinefried: if you could, that'd be good, yup :)16:39
melwittthanks sean-k-mooney16:39
*** tesseract has quit IRC16:41
*** maciejjozefczyk has quit IRC16:45
*** TxGirlGeek has quit IRC16:55
*** artom has quit IRC16:56
*** derekh has quit IRC16:57
*** jawad_axd has quit IRC16:57
*** tbachman has quit IRC16:58
*** tbachman has joined #openstack-nova17:00
*** TxGirlGeek has joined #openstack-nova17:01
*** mkrai has joined #openstack-nova17:02
*** jmlowe has joined #openstack-nova17:05
*** udesale has quit IRC17:05
*** lbragstad has joined #openstack-nova17:10
openstackgerritMerged openstack/nova master: Remove remaining Python 2.7-only dependencies  https://review.opendev.org/69889517:12
* efried lunches17:12
*** deke997 has joined #openstack-nova17:15
deke997Hey guys17:15
deke997I need to change how the guest pci addresses are generated. I have been trying to edit the config.py and driver.py files to add this functionality, but I cannot find the relevant code in the files17:15
*** tbachman has quit IRC17:19
*** rchurch has quit IRC17:20
*** rpittau is now known as rpittau|afk17:21
*** dtantsur is now known as dtantsur|afk17:24
*** jmlowe has quit IRC17:25
*** luksky has quit IRC17:26
*** artom has joined #openstack-nova17:31
*** mkrai has quit IRC17:31
*** ociuhandu_ has joined #openstack-nova17:32
*** evrardjp has quit IRC17:34
*** evrardjp has joined #openstack-nova17:34
stephenfindeke997: What are you trying to achieve?17:35
*** ociuhandu has quit IRC17:36
*** rchurch has joined #openstack-nova17:36
*** ociuhandu_ has quit IRC17:37
deke997So I have a unique multi-function pci device that only works properly in the VM if the both devices are in the same slot but different functions. Like 0000:00:05.0 and 0000:00:05.117:38
deke997by manually editing the libvirt.xml file, I verified that it works properly in the VM when set up like that17:38
deke997However, the libvirt.xml file is now generated automatically by nova, and by default it puts the devices in separate slots like: 0000:00:05.0 and 0000:00:06.017:39
deke997So I want to change that behavior so that when I have a pci device that is passed through, it will pay attention to whether 2 devices were in the same slot with different functions on the host machine, and if so, it will repeat that behavior in the VM17:40
*** jmlowe has joined #openstack-nova17:41
deke997stephenfin my understanding is that the xml file generation is handled by driver.py and config.py, but I cannot find the specific areas that deal with generating the guest pci address17:47
stephenfinI'm trying to root that out at the moment17:47
yoctozeptodeke997: btw, is the device in question a secret or can you shed some light on it?17:48
sean-k-mooneydeke997: we dont do all the generation in nova17:49
sean-k-mooneywe actully leave it up to libvirt to assign the guesst pci adress and as a result define the toplogy17:50
sean-k-mooneydeke997: i have a backloag item to modify how that works for other reasons in the futre as i want o model 1 pci root complex per numa node to optimize how devices are exposed to vms17:51
*** KeithMnemonic has quit IRC17:52
sean-k-mooneybut it sould like you need to detect that the device is a multi fucntion device and then we would need to explcitly stat generating the toplogy and assing pci address in the guest to enable use to expose the device as a multi function device to the guest17:52
sean-k-mooneydeke997: this would be well into the catagory of a spec not a bugfix in my view17:52
*** jmlowe has quit IRC17:53
sean-k-mooneywe might be able to skip the adress assignmentand still delegate that to libvirt but indicate that the device are both part of a multi function device but we would have to explore that to confirm17:53
deke997sean-k-mooney I see. This sounds like a fun project to work on. In the meantime, however, I have to find a solution that will let me get my VMs up17:53
deke997libvirt has a multifunction=on setting too17:54
sean-k-mooneyhow to you request and whitelist the device at present17:55
sean-k-mooneyi assume you are using the pci whitelist and a flavor alias17:55
deke997yes17:55
deke997but I have to use a separate alias for each function17:55
sean-k-mooneyand you have 2 aliase right17:55
sean-k-mooneyone for each funciton17:55
sean-k-mooneyya17:55
deke997We could add a multifunction alias type17:55
sean-k-mooneyso how are you ensuring the two VF/PF come form the same device on the hsot17:56
sean-k-mooneydeke997: we could add a tage but unless we can adress ^17:56
sean-k-mooneywe have bigger problems17:56
deke997right now, I am just using all of them, so it's not an issue haha17:56
sean-k-mooneyso you have only 1 device on the compute node and your booting 1 vm that use all the fucntion form that device17:57
deke997yes that's what I have set up for testing17:57
sean-k-mooneyya in that specific case you whave worked around that issue17:57
deke997But in the libvirt XML it is very easy to see which devices from the host were passed through17:57
deke997and match them17:57
sean-k-mooneythis has come up in the context of gpus in the past17:58
sean-k-mooneythe GPU and audio contoler  are usally 2 differnt pci endpoint on the same device17:58
sean-k-mooneyat pressent we dont have a way to passthough both form the same card and ensure they are form the same device17:58
*** psachin has joined #openstack-nova17:59
sean-k-mooneydeke997: form a nova point of view you would basically want ot have an alsi that can refernce a second alias and request both as a pair17:59
deke997yea exactly18:00
*** TxGirlGeek has quit IRC18:00
deke997That doesn't seem super difficult18:00
sean-k-mooneyand then in nova/libvirt have it select the pair form the same device an make them a multifunction device in the xml if they are expose that way on the host18:00
sean-k-mooneydeke997: you say that now18:01
deke997haha18:01
deke997Is there a way to make nova play nice with edited libvirt.xmls?18:01
sean-k-mooneyso this all live in the pci manger code https://github.com/openstack/nova/tree/master/nova/pci18:01
sean-k-mooneyno18:01
sean-k-mooneynova regenerates the xml based on seveal api actions18:02
sean-k-mooneylike hard reboot18:02
sean-k-mooneyso your eddit would last until someone stops and start the vm18:02
*** derekh has joined #openstack-nova18:02
deke997Is there a way to disable that regeneration?18:02
sean-k-mooneyno18:03
*** rchurch_ has joined #openstack-nova18:03
openstackgerritMerged openstack/nova master: libvirt: Add a default VirtIO-RNG device to guests  https://review.opendev.org/61660318:03
*** jaosorior has joined #openstack-nova18:03
sean-k-mooneyso there are three parts. the pci mager woudl have to be enhanced to pars and track paired deivces, the pci pasthough filter would have to be update and the libvirt driver xml generateion would  have to be updated18:04
sean-k-mooneythe alias schema coudl certenly be extended to referen a child device https://github.com/openstack/nova/blob/master/nova/pci/request.py#L66-L9918:05
deke997How would we go about updating the xml generation?18:05
*** rchurch_ has quit IRC18:06
sean-k-mooneyam ill see if i can find it. but likely i would see if we can still leave libvirt choose the target pci adders in the guest but mark a specicifc pci requiest as multi funcion18:06
*** rchurch_ has joined #openstack-nova18:06
*** rchurch has quit IRC18:06
*** rchurch_ has quit IRC18:08
*** rchurch has joined #openstack-nova18:08
*** TxGirlGeek has joined #openstack-nova18:08
deke997I need to know a bit more about how libvirt chooses those addresses18:08
sean-k-mooneyso you would have to modify https://github.com/openstack/nova/blob/6db486e9fd4f6b8dd02371b043e66808cdd1e0cd/nova/virt/libvirt/driver.py#L587118:09
*** rchurch has quit IRC18:09
sean-k-mooneythe general docs for thi in libvirt are https://libvirt.org/formatdomain.html#elementsHostDevSubsys18:10
*** rchurch has joined #openstack-nova18:10
sean-k-mooneynormally we just generate a hostdev element like this18:11
sean-k-mooney<hostdev mode='subsystem' type='pci' managed='yes'>18:11
sean-k-mooney    <source>18:11
sean-k-mooney      <address domain='0x0000' bus='0x06' slot='0x02' function='0x0'/>18:11
sean-k-mooney    </source>18:11
sean-k-mooney    <boot order='1'/>18:11
sean-k-mooney    <rom bar='on' file='/etc/fake/boot.bin'/>18:11
sean-k-mooney  </hostdev>18:11
sean-k-mooneyso we specify the source element and the host pci adress and libvirt chooes a target adress18:11
sean-k-mooneythat i assume is what you were changing with your manual edit18:12
deke997Yes18:12
deke997but I could only start that instance manually through libvirt after the edit, because nova just tries to overwrite it18:12
*** amodi has quit IRC18:12
sean-k-mooneyhttps://libvirt.org/formatdomain.html#elementsAddress this is the docs for the device addressing18:13
deke997Also, adding multifunction = 'on' to the source address may be necessary. Libvirt docs say to use it, but I was able to have success either way18:13
deke997which would have to be done in nova18:14
deke997would be easy to just add multifunction = yes iff alias_type = multifunction or something18:15
sean-k-mooneyya the multifunction bit is not what im worried about18:16
sean-k-mooneyi think the only way to make this work si to have nova set the target Adress on both devices18:17
sean-k-mooneybut since nova does not track the target adress today that is a prbalem18:17
sean-k-mooneyalso you would have to take into account both q35 and the i440x machine types18:17
sean-k-mooneythe old i440x or pc machine type is pci only18:18
sean-k-mooneyand q35 is pcie by default but you can add pci busses too18:18
sean-k-mooneyso that makes it more complicated18:18
deke997definitely18:19
sean-k-mooneywith q35 if we wanted to set <address domain='0x0000' bus='0x06' slot='0x02' function='0x0'/>18:20
sean-k-mooneyfor the target we would need to first create a pcie root port for it18:20
sean-k-mooneyif we leave out the target element libvirt does it for us18:21
sean-k-mooneyso baseicly when  you specify it you also have to do the stuff libvirt does for you automatically18:21
sean-k-mooneybasically18:22
sean-k-mooneydevices>18:22
sean-k-mooney  <controller type='pci' index='0' model='pci-root'/>18:22
sean-k-mooney  <controller type='pci' index='1' model='pci-bridge'>18:22
sean-k-mooney    <address type='pci' domain='0' bus='0' slot='5' function='0' multifunction='off'/>18:22
sean-k-mooney  </controller>18:22
sean-k-mooney</devices>18:22
sean-k-mooneywell it would be multifucniton=one in this case18:22
sean-k-mooneybut once we have the multi funtion bridge device you can add the enpoint beneth it18:22
sean-k-mooneyi guess the only real difference that mater for q35 vs pc is pcie-bridge vs pci-bridge in the contoler element18:24
deke997Okay so, haha, what if we let libvirt handle the target elements, but then before we boot, we have nova edit the slot and function numbers so that pairs match18:24
deke997:')18:24
*** jmlowe has joined #openstack-nova18:24
sean-k-mooneyam first reaction i dont like that but thinking it though.18:25
sean-k-mooneywe could defien the domain without starting it18:25
deke997yes18:25
sean-k-mooneythat would cause libvirt to fill in the details i think18:25
sean-k-mooneythen we could read it back and ed the pair. maybe18:25
deke997all we'd have to do is check for two source lines that match and then make the targets match too18:26
sean-k-mooneybut it kind fo feels like a hack and it like wont scale well if there are multile deivecs18:26
deke997it is definitely a hack18:26
deke997why won't it scale well?18:26
deke997never gonna have more than ~8 devices in a server18:27
sean-k-mooneyif we had multile host deves we woudl get weired gaps in the pci adress space and if we had more then 8 virtual funciton it woudl get complcated18:27
sean-k-mooneywell the bit after the . is in octal so we can on ly have 8 fucntion per slot18:27
sean-k-mooneyso if you needed more then 8 virtual funtion it would not work right18:28
deke997I don't think such a device exists18:28
deke997haven't ever seen more than like 418:28
deke997my use case is only 218:28
sean-k-mooneyya i have seen up to 8 well no i have seen way more18:28
sean-k-mooneyintel nics when using sriov can have 64 - 128 vf per pf18:29
sean-k-mooneybut they do the adressing slightly differently18:29
deke997sriov is totally different tho18:29
sean-k-mooneyya for there case teh assing a full bus per card18:29
deke997this would be specifically for non-sriov multi-function devices18:29
efriedstephenfin: https://zuul.opendev.org/t/openstack/build/cc8e185b95cf4d5cbe973c4b81493084 is on rackcdn so I can't get to it. Can you see what the fail is? (ec2)18:30
sean-k-mooneyso you have mulitple slots and funtion to support up to 256 vf totall18:30
sean-k-mooneydeke997: ack18:30
sean-k-mooneydeke997: so you can see why i think this would need a spec18:30
sean-k-mooneyto design correctly18:30
deke997yes, I'm just hoping for an intermediate band aid while we work on a spec18:31
sean-k-mooneyi do see a use case for both multi funtion deivce passthough and the more generif i have a pair of device tha tmus be allocated together18:31
sean-k-mooneye.g. the gpu and the audio contole on the gpu18:31
*** derekh has quit IRC18:32
sean-k-mooneynviad might actully be exposing it as a multi function device so that might actully be the same usecase18:32
*** martinkennelly has quit IRC18:32
sean-k-mooneyya they are18:35
sean-k-mooney81:00.0 VGA compatible controller: NVIDIA Corporation GK104GL [Quadro K5000] (rev a1)18:35
sean-k-mooney81:00.1 Audio device: NVIDIA Corporation GK104 HDMI Audio Controller (rev a1)18:35
sean-k-mooneyso if we we made ^ work it would work for you too right18:35
*** eharney has joined #openstack-nova18:36
deke997yes18:38
deke997That looks pretty much identical to my use case18:38
sean-k-mooneyok what i would basicaly suggest is as follows. extent the alias with a child tag that can reference other alisas.18:42
sean-k-mooneywhen you add an alias to a flaovr which has a child tag we woudl claim the deivce and all its childern18:42
*** psachin has quit IRC18:43
sean-k-mooneyand expose them as a multi funcitn device18:43
sean-k-mooneythe pci pasthogh through filter would have to be enhanced to ensure it only passes host where the parent and child alias can be allocated form the same slot18:44
sean-k-mooneythe pci manger would also need to updated so that when we do the pci claim it claims the pair form the same slot18:44
deke997wouldn't that always be the case?18:44
deke997I don't see why the passthrough filter needs to be updated18:45
sean-k-mooneywell if you put two of those card in the same server it would break18:45
sean-k-mooneyas we coudl get 1 fucntion of type a form one card and one fucntion of type b from another18:46
sean-k-mooneyso the filter and assignment code would have to be hardened to prevent that18:46
*** TxGirlGeek has quit IRC18:48
deke997The assignment code, yes, but the filter, I don't think so18:48
deke997because before assignment, it would always be true that both the parent and child could be allocated together18:48
sean-k-mooneyits true if you alwasy need to request both18:49
sean-k-mooneywhich is true in your case18:49
deke997right18:49
sean-k-mooneynot in the gpu case18:49
deke997I see18:49
deke997in that case, then yea the filter too18:50
sean-k-mooneyanyway the assignment code and filter share the same code18:50
sean-k-mooneythe filter basicaly passes if it would be able to assign18:50
*** openstackgerrit has quit IRC18:51
sean-k-mooneydeke997: https://github.com/openstack/nova/blob/master/nova/scheduler/filters/pci_passthrough_filter.py#L5118:51
sean-k-mooneysee we just loop over the request and ask can it suppor the request18:52
sean-k-mooneywell the loopin is don internally in the suport_request funtion18:52
sean-k-mooneywe just make a copy of the avliable devices and try to assigntem on hte copy https://github.com/openstack/nova/blob/6db486e9fd4f6b8dd02371b043e66808cdd1e0cd/nova/pci/stats.py#L374-L37518:53
*** jaosorior has quit IRC18:54
sean-k-mooneyso if you update _apply_request to handel the assignment constriatt the filter gets updated for free18:54
deke997got it18:55
deke997are we proposing to let libvirt handle initial target device config and then have nova take over, or are we proposing to have nova handle target device config?18:56
deke997These are the only two options so far, yea?18:56
sean-k-mooneythe only two i am aware of. if we were to do this updatream i would prefer to do it properly18:56
sean-k-mooneyand have nova do it. if you need to do it downstream quick well that is up to you18:57
sean-k-mooneydeke997: what is the eta you had in mind for this18:57
deke9972 days ago haha18:57
deke997I won't be sleeping much till I get this working18:58
sean-k-mooneyya fair warning to land this upstream its proably going to be next cycle18:58
deke997yea so I think we'll do both18:58
deke997downstream quick, and then work on upstream18:58
sean-k-mooneyi could be ussuri if it was really pusshed for but this code is complex and a pain to test18:58
deke997when is the next cycle?18:58
sean-k-mooneyit starts in 3 monts os like september/october18:59
sean-k-mooneythere is still about 2-3 weeks to propose specs for this cycle18:59
sean-k-mooneyyou would need to figure out what need to chagne and write up a propals and get it review by then18:59
sean-k-mooneythen there is able 2 months left to land the feature19:00
sean-k-mooneybut beign realisting if you need this quick19:00
sean-k-mooneythen likely createing a poc and upstreaming it next cycle is what your going to have to do19:00
deke997One thing to check on:19:01
sean-k-mooneyfor the poc you could do the hack where you fix up the adresses19:01
deke997I think cyborg is doing some stuff with multi function too19:01
deke997I saw it while doing research19:01
sean-k-mooneyya so cyborg would be another option19:02
sean-k-mooneyif the integration is done this cycle19:02
sean-k-mooneythen you could write a device dirver for you custom multifunion device19:02
sean-k-mooneybut the xml change would still need to be done19:02
sean-k-mooneyyou would not have to touch the filer or pci manager in that case19:03
sean-k-mooneybut you would need to do all the cyborg work19:03
sean-k-mooneydeke997: cyborg support in nova is planned for this cycle but we said that last cycle too19:04
sean-k-mooneydeke997: its much closer to being read this time however19:05
deke997Good to know19:06
deke997I need to look into how they're implementing multi function a bit more19:07
deke997but even so, I can't wait till the next cycle to get a beta working here19:08
sean-k-mooneyat present without the libvirt support for multifuntion devices in nova cybporg cant support what you need19:09
deke997I think the options in order of increasing time, complexity, and correctness would probably be19:09
deke9971. hack19:09
*** TxGirlGeek has joined #openstack-nova19:09
deke9972. nova19:09
deke9973. cyborg19:09
sean-k-mooney0 heiring an intern to manually update every vm as it spawns19:09
deke997hahahah19:10
deke997but I can't even do that19:10
deke997the xml changes don't go into effect till you reboot19:10
sean-k-mooneyactully there is one other hack you cloud do19:10
deke997and when you reboot, it gets overwritten19:10
deke997is there a way to make the xml changes before boot? or to make the changes go live while the instance is running?19:11
sean-k-mooneyfor reasons in the past i did have need to do horible things19:11
sean-k-mooneyhttps://opendev.org/x/networking-ovs-dpdk/src/branch/master/devstack/libs/ovs-dpdk#L79-L13019:11
sean-k-mooneythat is an example of a script that replace the qemu executable and rewrites the qemu commandline on the fly19:12
*** tbachman has joined #openstack-nova19:12
*** heso has quit IRC19:13
sean-k-mooneywhat that sciprt is acatlly doing is making sure nic offload that did not work with ancnet version of qemu and ovs-dpdk are disabeld and it makes ure the guest memory has the  mmap shared flag set19:13
sean-k-mooneyyou could install a qemu wrapper script as a hack to match on parts of the qemu command line and alter it19:14
sean-k-mooneyit would have 0 support form anyone but it is a -1 on your scale19:14
deke997haha i'll have to look into that too i guess19:17
sean-k-mooneyadding flags to existing optins is nasty but actully kind of simple19:17
sean-k-mooneydoing what you need to do would be harder19:17
sean-k-mooneybut at the end of the day the commandline is fairly deterministic so it would be doable19:18
sean-k-mooneydefinelty not my first choise but we dont always get to do our first choice19:18
deke997I'll see what I can figure out tonight and get back to you soon with the progress19:19
deke997lmk if you have any other ideas19:20
*** gmann is now known as gmann_lunch19:26
*** jmlowe has quit IRC19:27
*** jamesdenton has quit IRC19:29
*** jamesdenton has joined #openstack-nova19:29
*** rchurch has quit IRC19:46
*** gmann_lunch is now known as gmann19:52
*** damien_r has quit IRC19:52
*** rchurch has joined #openstack-nova19:53
*** rchurch has quit IRC19:54
*** rchurch has joined #openstack-nova19:54
*** eharney has quit IRC19:55
*** jmlowe has joined #openstack-nova20:08
*** openstackstatus has joined #openstack-nova20:08
*** ChanServ sets mode: +v openstackstatus20:08
*** jaosorior has joined #openstack-nova20:10
*** jmlowe has quit IRC20:13
*** luksky has joined #openstack-nova20:36
*** bbowen has quit IRC20:47
*** nweinber_ has quit IRC20:52
*** artom has quit IRC20:58
*** eharney has joined #openstack-nova21:01
*** eharney has quit IRC21:02
*** eharney has joined #openstack-nova21:02
*** rcernin has joined #openstack-nova21:04
*** johanssone_ has joined #openstack-nova21:05
*** jdillaman has joined #openstack-nova21:08
*** dansmith has quit IRC21:18
*** irclogbot_2 has quit IRC21:18
*** jhesketh has quit IRC21:18
*** adriant has quit IRC21:18
*** panda has quit IRC21:18
*** jkulik has quit IRC21:18
*** dklyle has quit IRC21:19
*** johanssone has quit IRC21:19
*** stephenfin has quit IRC21:19
*** mgoddard has quit IRC21:19
*** haleyb has quit IRC21:19
*** adrianc has quit IRC21:19
*** admin0 has quit IRC21:19
*** tonyb has quit IRC21:19
*** lchabert has quit IRC21:19
*** gryf has quit IRC21:19
*** hemna has quit IRC21:19
*** openstackstatus has quit IRC21:20
*** dklyle has joined #openstack-nova21:21
*** jaosorior has quit IRC21:25
*** irclogbot_3 has joined #openstack-nova21:26
*** tonyb has joined #openstack-nova21:26
*** eharney has quit IRC21:29
*** jawad_axd has joined #openstack-nova21:38
*** jawad_axd has quit IRC21:42
*** mgariepy has quit IRC21:43
*** efried has quit IRC21:44
*** efried has joined #openstack-nova21:44
*** bbowen has joined #openstack-nova21:50
*** jawad_axd has joined #openstack-nova21:59
*** dansmith has joined #openstack-nova22:01
*** jhesketh has joined #openstack-nova22:01
*** adriant has joined #openstack-nova22:01
*** panda has joined #openstack-nova22:01
*** jkulik has joined #openstack-nova22:01
*** stephenfin has joined #openstack-nova22:01
*** mgoddard has joined #openstack-nova22:01
*** haleyb has joined #openstack-nova22:01
*** admin0 has joined #openstack-nova22:01
*** adrianc has joined #openstack-nova22:01
*** lchabert has joined #openstack-nova22:01
*** gryf has joined #openstack-nova22:01
*** hemna has joined #openstack-nova22:01
*** jawad_axd has quit IRC22:03
*** tosky has joined #openstack-nova22:16
*** xek has quit IRC22:22
*** artom has joined #openstack-nova22:24
*** slaweq_ has joined #openstack-nova22:45
*** jmlowe has joined #openstack-nova22:53
*** TxGirlGeek has quit IRC23:01
*** slaweq_ has quit IRC23:01
*** TxGirlGeek has joined #openstack-nova23:02
*** tkajinam has joined #openstack-nova23:07
*** jmlowe has quit IRC23:07
*** mlavalle has quit IRC23:12
*** rchurch has quit IRC23:32
*** rchurch has joined #openstack-nova23:35
*** dviroel has quit IRC23:39
*** tbachman has quit IRC23:48
*** rchurch has quit IRC23:50
*** macz has joined #openstack-nova23:50
*** rchurch has joined #openstack-nova23:52
*** macz has quit IRC23:54

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!