*** tbachman has quit IRC | 00:00 | |
*** ociuhandu has joined #openstack-nova | 00:01 | |
*** ociuhandu has quit IRC | 00:05 | |
*** artom has joined #openstack-nova | 00:13 | |
*** TxGirlGeek has quit IRC | 00:26 | |
*** jaosorior has quit IRC | 00:28 | |
*** TxGirlGeek has joined #openstack-nova | 00:31 | |
*** TxGirlGeek has quit IRC | 00:36 | |
*** jmlowe has joined #openstack-nova | 00:39 | |
*** luksky has quit IRC | 00:47 | |
*** jmlowe has quit IRC | 00:54 | |
*** jmlowe has joined #openstack-nova | 01:12 | |
*** damien_r has joined #openstack-nova | 01:16 | |
*** damien_r has quit IRC | 01:17 | |
*** damien_r has joined #openstack-nova | 01:17 | |
*** jmlowe has quit IRC | 01:18 | |
*** tbachman has joined #openstack-nova | 01:19 | |
*** gyee has quit IRC | 01:52 | |
*** tbachman has quit IRC | 02:09 | |
*** damien_r has quit IRC | 02:19 | |
*** adriant has quit IRC | 02:41 | |
*** adriant has joined #openstack-nova | 02:43 | |
*** tbachman has joined #openstack-nova | 02:47 | |
*** artom has quit IRC | 02:50 | |
*** damien_r has joined #openstack-nova | 03:01 | |
*** damien_r has quit IRC | 03:05 | |
*** TxGirlGeek has joined #openstack-nova | 03:33 | |
*** tbachman has quit IRC | 03:35 | |
*** psachin has joined #openstack-nova | 03:36 | |
*** TxGirlGeek has quit IRC | 03:39 | |
*** udesale has joined #openstack-nova | 04:11 | |
*** TxGirlGeek has joined #openstack-nova | 04:16 | |
*** mkrai_ has joined #openstack-nova | 04:23 | |
*** TxGirlGeek has quit IRC | 04:24 | |
*** jhesketh has quit IRC | 04:54 | |
*** jhesketh has joined #openstack-nova | 04:55 | |
*** macz has joined #openstack-nova | 05:00 | |
*** macz has quit IRC | 05:05 | |
*** ratailor has joined #openstack-nova | 05:21 | |
*** evrardjp has quit IRC | 05:34 | |
*** evrardjp has joined #openstack-nova | 05:34 | |
*** TxGirlGeek has joined #openstack-nova | 05:46 | |
*** TxGirlGeek has quit IRC | 05:55 | |
*** links has joined #openstack-nova | 05:56 | |
*** tkajinam has quit IRC | 06:01 | |
*** tkajinam has joined #openstack-nova | 06:03 | |
*** ccamacho has quit IRC | 06:26 | |
*** mkrai_ has quit IRC | 06:42 | |
*** _rajat_ has joined #openstack-nova | 06:47 | |
*** mkrai_ has joined #openstack-nova | 06:47 | |
*** jaosorior has joined #openstack-nova | 06:47 | |
*** tkajinam_ has joined #openstack-nova | 07:01 | |
*** tkajinam has quit IRC | 07:03 | |
*** heso has joined #openstack-nova | 07:12 | |
*** Luzi has joined #openstack-nova | 07:21 | |
*** tkajinam__ has joined #openstack-nova | 07:47 | |
*** iurygregory has joined #openstack-nova | 07:48 | |
*** tkajinam_ has quit IRC | 07:50 | |
*** ccamacho has joined #openstack-nova | 07:50 | |
*** luksky has joined #openstack-nova | 07:53 | |
*** tesseract has joined #openstack-nova | 07:56 | |
*** aarents has quit IRC | 07:57 | |
*** f0o has quit IRC | 07:58 | |
*** rmk has quit IRC | 07:59 | |
*** fungi has quit IRC | 07:59 | |
*** fnordahl has quit IRC | 07:59 | |
*** gouthamr has quit IRC | 07:59 | |
*** owalsh has quit IRC | 07:59 | |
*** Roamer` has quit IRC | 07:59 | |
*** toanster has quit IRC | 07:59 | |
*** frickler has quit IRC | 07:59 | |
*** Jeffrey4l has quit IRC | 07:59 | |
*** cz3 has quit IRC | 07:59 | |
*** cmurphy has quit IRC | 07:59 | |
*** Anticimex has quit IRC | 07:59 | |
*** yankcrime has quit IRC | 07:59 | |
*** dtantsur|afk has quit IRC | 07:59 | |
*** johnthetubaguy has quit IRC | 08:01 | |
*** Jeffrey4l has joined #openstack-nova | 08:02 | |
*** cz3 has joined #openstack-nova | 08:02 | |
*** yankcrime has joined #openstack-nova | 08:02 | |
*** cmurphy has joined #openstack-nova | 08:02 | |
*** Anticimex has joined #openstack-nova | 08:02 | |
*** dtantsur|afk has joined #openstack-nova | 08:02 | |
*** irclogbot_3 has quit IRC | 08:04 | |
*** irclogbot_2 has joined #openstack-nova | 08:05 | |
*** bnemec has joined #openstack-nova | 08:05 | |
*** dansmith has quit IRC | 08:06 | |
*** dansmith has joined #openstack-nova | 08:08 | |
*** aarents has joined #openstack-nova | 08:10 | |
*** maciejjozefczyk has joined #openstack-nova | 08:10 | |
*** tosky has joined #openstack-nova | 08:13 | |
*** mkrai_ has quit IRC | 08:20 | |
*** tkajinam__ has quit IRC | 08:29 | |
*** tkajinam has joined #openstack-nova | 08:29 | |
openstackgerrit | Merged openstack/nova master: Func test for failed and aborted live migration https://review.opendev.org/699065 | 08:31 |
---|---|---|
openstackgerrit | Merged openstack/nova master: Remove unused code https://review.opendev.org/703408 | 08:31 |
*** udesale_ has joined #openstack-nova | 08:31 | |
*** mkrai_ has joined #openstack-nova | 08:32 | |
*** udesale_ has quit IRC | 08:33 | |
*** udesale_ has joined #openstack-nova | 08:33 | |
*** udesale has quit IRC | 08:34 | |
*** tkajinam has quit IRC | 08:35 | |
*** bnemec has quit IRC | 08:41 | |
*** rpittau|afk is now known as rpittau | 08:43 | |
*** ociuhandu has joined #openstack-nova | 08:45 | |
*** rmk has joined #openstack-nova | 08:45 | |
*** fungi has joined #openstack-nova | 08:45 | |
*** f0o has joined #openstack-nova | 08:45 | |
*** fnordahl has joined #openstack-nova | 08:45 | |
*** gouthamr has joined #openstack-nova | 08:45 | |
*** owalsh has joined #openstack-nova | 08:45 | |
*** Roamer` has joined #openstack-nova | 08:45 | |
*** toanster has joined #openstack-nova | 08:45 | |
*** frickler has joined #openstack-nova | 08:45 | |
*** openstackstatus has quit IRC | 08:48 | |
*** bnemec has joined #openstack-nova | 08:50 | |
*** martinkennelly has joined #openstack-nova | 08:51 | |
*** udesale_ has quit IRC | 08:54 | |
*** ociuhandu has quit IRC | 08:56 | |
*** mrch has joined #openstack-nova | 08:58 | |
*** udesale has joined #openstack-nova | 09:02 | |
*** bnemec has quit IRC | 09:08 | |
*** xek has joined #openstack-nova | 09:09 | |
*** bbowen has quit IRC | 09:24 | |
*** bbowen has joined #openstack-nova | 09:24 | |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove remaining Python 2.7-only dependencies https://review.opendev.org/698895 | 09:27 |
*** pcaruana has joined #openstack-nova | 09:30 | |
*** bnemec has joined #openstack-nova | 09:33 | |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Update API reference guide https://review.opendev.org/703796 | 09:33 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'nova.image.api' module https://review.opendev.org/702451 | 09:33 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: WIP: nova-net: Remove unused nova-network objects https://review.opendev.org/697156 | 09:33 |
*** dtantsur|afk is now known as dtantsur | 09:36 | |
*** iurygregory has quit IRC | 09:40 | |
*** Liang__ has joined #openstack-nova | 09:43 | |
*** ratailor has quit IRC | 09:44 | |
*** ratailor has joined #openstack-nova | 09:51 | |
*** derekh has joined #openstack-nova | 09:58 | |
*** iurygregory has joined #openstack-nova | 10:01 | |
*** jistr has quit IRC | 10:06 | |
*** jistr has joined #openstack-nova | 10:06 | |
*** jaosorior has quit IRC | 10:08 | |
*** zbr has left #openstack-nova | 10:09 | |
*** sridharg has joined #openstack-nova | 10:12 | |
*** dviroel has joined #openstack-nova | 10:42 | |
*** sridharg has quit IRC | 10:43 | |
*** udesale has quit IRC | 10:56 | |
*** bnemec has quit IRC | 11:02 | |
*** tbachman has joined #openstack-nova | 11:04 | |
*** sridharg has joined #openstack-nova | 11:06 | |
*** sridharg has quit IRC | 11:11 | |
*** rchurch has quit IRC | 11:18 | |
openstackgerrit | Merged openstack/nova master: Switch to hacking 2.x https://review.opendev.org/695733 | 11:21 |
openstackgerrit | Merged openstack/nova master: requirements: Limit hacking to one minor version https://review.opendev.org/703405 | 11:21 |
*** rpittau is now known as rpittau|bbl | 11:21 | |
*** sridharg has joined #openstack-nova | 11:24 | |
*** Liang__ has quit IRC | 11:28 | |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unused nova-network objects https://review.opendev.org/697156 | 11:32 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unnecessary exception handling https://review.opendev.org/703973 | 11:32 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unused parameters https://review.opendev.org/703974 | 11:32 |
stephenfin | dansmith: 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 around | 11:34 |
*** bbowen has quit IRC | 11:34 | |
stephenfin | dansmith: 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 case | 11:34 |
stephenfin | sean-k-mooney: Maybe relevant, idk ^ | 11:34 |
sean-k-mooney | we should not move the network object to os-vif | 11:43 |
sean-k-mooney | we already have one | 11:43 |
sean-k-mooney | the os vif objects are constucted from info often taken form the nova ones but they are not the same objects | 11:44 |
*** tbachman has quit IRC | 11:49 | |
stephenfin | sean-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 |
gibi | kashyap: rng patch looks good to me, thanks for the update | 11:58 |
lucidguy | sean-k-mooney: The only option I see regarding hugepages is reserved_huge_pages. Is that what you were talking about? | 11:58 |
stephenfin | If 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' object | 11:58 |
sean-k-mooney | stephenfin: i know dansmith found that the namespace stuff does not seam to be working properly | 11:58 |
stephenfin | did he now? | 11:59 |
stephenfin | any notes or bugs on that? | 11:59 |
sean-k-mooney | lucidguy: no, hugepages is a afeature that is enabled in the flavor/image by setting hw:mem_page_size=large | 11:59 |
sean-k-mooney | reserved_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 vms | 12:00 |
sean-k-mooney | stephenfin: it cam up on one of his patch let me see if it merged or not | 12:01 |
lucidguy | sean-k-mooney: Understood | 12:01 |
sean-k-mooney | stephenfin: https://review.opendev.org/#/c/702049/ | 12:01 |
sean-k-mooney | the ephemeral object patch | 12:01 |
sean-k-mooney | stephenfin: line 1175 https://review.opendev.org/#/c/702049/4/nova/tests/unit/objects/test_objects.py | 12:02 |
sean-k-mooney | lucidguy: i think these are the docs on hugepage that you want https://docs.openstack.org/nova/latest/admin/huge-pages.html | 12:03 |
sean-k-mooney | lucidguy: we have summit presintaion on this too that stephenfin did | 12:03 |
lucidguy | I was just looking at that link | 12:04 |
lucidguy | Cool | 12:05 |
kashyap | gibi: Hi, thanks for looking! | 12:07 |
kashyap | gibi: Want me to respin removing the stray word "this"? | 12:09 |
* kashyap --> lunch | 12:09 | |
gibi | kashyap: if you respin quickly then I can re +2 it. but I'm fine merging the patch as is | 12:09 |
sean-k-mooney | stephenfin: i think you need to update https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_objects.py#L1225 | 12:09 |
stephenfin | sean-k-mooney: incoming :) | 12:09 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Make the security group API a module https://review.opendev.org/696746 | 12:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove remaining nova-network quotas https://review.opendev.org/697152 | 12:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove use of legacy 'FloatingIP' object https://review.opendev.org/697153 | 12:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove use of legacy 'Network' object https://review.opendev.org/697154 | 12:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove use of legacy 'SecurityGroup' object https://review.opendev.org/697155 | 12:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove now unnecessary nova-net workaround https://review.opendev.org/702440 | 12:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Update API reference guide https://review.opendev.org/703796 | 12:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'nova.image.api' module https://review.opendev.org/702451 | 12:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unnecessary exception handling https://review.opendev.org/703973 | 12:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unused parameters https://review.opendev.org/703974 | 12:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unused nova-network objects https://review.opendev.org/697156 | 12:10 |
sean-k-mooney | stephenfin: change base.NovaObjectRegistry.obj_classes() to call get_nova_objects https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_objects.py#L1159 | 12:10 |
stephenfin | you b***** | 12:10 |
stephenfin | gdi git | 12:10 |
stephenfin | :( | 12:10 |
sean-k-mooney | so it only compares the version of hte nova object not also the os-vif ones | 12:11 |
sean-k-mooney | that is why you are having issues with the network object | 12:11 |
stephenfin | Yup, makes sense. Thanks :) | 12:11 |
stephenfin | Should be fixed now | 12:11 |
sean-k-mooney | along with a rebase :P | 12:11 |
stephenfin | I 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-mooney | ah ok | 12:12 |
sean-k-mooney | well you could also delete the nova network object now right? | 12:13 |
stephenfin | yup, done https://review.opendev.org/#/c/697156/25/nova/objects/network.py | 12:13 |
sean-k-mooney | or is there still nova code using that | 12:13 |
sean-k-mooney | oh then you missed removing it form the object versions | 12:13 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unused nova-network objects https://review.opendev.org/697156 | 12:14 |
stephenfin | whoops | 12:14 |
sean-k-mooney | line 1102 https://review.opendev.org/#/c/697156/25/nova/tests/unit/objects/test_objects.py | 12:14 |
stephenfin | ta | 12:14 |
*** jaosorior has joined #openstack-nova | 12:14 | |
sean-k-mooney | the thing that uses that might also need the get_nova_object change | 12:14 |
stephenfin | If it doesn't, it's not obvious | 12:15 |
stephenfin | as in there are no tests failing fwict | 12:15 |
*** rcernin has quit IRC | 12:15 | |
sean-k-mooney | ya so the issue is all this stuff was set up befor namespaces were added to ovos | 12:16 |
sean-k-mooney | so it was really not expecting ot have two object with the same name but be in differne namespaces | 12:16 |
jroll | efried: cool, thanks, just wanted to make sure you weren't waiting on me for something in particular there :) | 12:17 |
openstackgerrit | Kashyap Chamarthy proposed openstack/nova master: libvirt: Add a default VirtIO-RNG device to guests https://review.opendev.org/616603 | 12:18 |
kashyap | gibi: --^ Done | 12:18 |
gibi | kashyap: thanks, +2 | 12:19 |
gibi | stephenfin: if you have time there is an easy patch to +2 https://review.opendev.org/616603 | 12:19 |
*** rcernin has joined #openstack-nova | 12:20 | |
*** mgariepy has joined #openstack-nova | 12:22 | |
stephenfin | kashyap: If you can you address my nit on the releasenote, I'm +2 | 12:22 |
stephenfin | gibi too ^ | 12:22 |
kashyap | stephenfin: Yeah, I already hesitated about that first bit. As the URL will take care | 12:23 |
kashyap | Let me do it real quick | 12:23 |
*** jaosorior has quit IRC | 12:24 | |
* gibi is waiting patiently for the next PS :) | 12:24 | |
openstackgerrit | Kashyap Chamarthy proposed openstack/nova master: libvirt: Add a default VirtIO-RNG device to guests https://review.opendev.org/616603 | 12:25 |
stephenfin | ta. +2 | 12:25 |
kashyap | Alright, fixed the reno. | 12:25 |
*** jaosorior has joined #openstack-nova | 12:25 | |
kashyap | gibi: Your patient wait is over :D | 12:26 |
*** rcernin has quit IRC | 12:26 | |
* gibi is waiting for gerrit to load :D | 12:27 | |
gibi | done | 12:27 |
gibi | today something is wrong with my uplink | 12:27 |
*** bnemec has joined #openstack-nova | 12:29 | |
openstackgerrit | Balazs Gibizer proposed openstack/nova stable/stein: Mask the token used to allow access to consoles https://review.opendev.org/702181 | 12:31 |
gibi | elod: fixed your comments in ^^ | 12:31 |
sean-k-mooney | stephenfin: have you seen issues with nova.tests.functional.test_nova_manage.TestDBArchiveDeletedRowsMultiCell failing out of interest? | 12:37 |
stephenfin | I haven't paid attention to it, no | 12:37 |
*** jawad_axd has joined #openstack-nova | 12:37 | |
sean-k-mooney | ok i was wondering if that was the failing db test ye were talking about yesterday | 12:38 |
sean-k-mooney | i guess not | 12:38 |
*** bbowen has joined #openstack-nova | 12:38 | |
elod | gibi: thx, looking | 12:39 |
*** ociuhandu has joined #openstack-nova | 12:46 | |
*** ociuhandu has quit IRC | 12:46 | |
*** mkrai_ has quit IRC | 12:48 | |
*** _rajat_ has quit IRC | 12:50 | |
stephenfin | sean-k-mooney: Don't know if I showed this to you before Xmas or not https://github.com/testing-cabal/subunit/pull/40 | 12:54 |
*** ociuhandu has joined #openstack-nova | 12:54 | |
stephenfin | I 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/ anyway | 12:55 |
stephenfin | whoops | 12:55 |
sean-k-mooney | you did not but ill take a look at both | 12:55 |
stephenfin | https://review.opendev.org/#/c/700522/ | 12:56 |
sean-k-mooney | why are you importing the print fucntion explcitly | 12:56 |
sean-k-mooney | are you using py26 lol | 12:56 |
sean-k-mooney | you can still do that but it was never needed in py27 | 12:57 |
stephenfin | it definitely is/was :) | 12:58 |
sean-k-mooney | but ya does that work | 12:58 |
stephenfin | to use the 'file=foo' thing, anyway | 12:58 |
*** ociuhandu has quit IRC | 12:58 | |
sean-k-mooney | oh ya that is python3 only | 12:58 |
stephenfin | 'zactly | 12:59 |
sean-k-mooney | but again nova has droped python 2 support so :P | 12:59 |
stephenfin | so that nova patch just prints a load of junk that overwhelms subunit | 12:59 |
stephenfin | without my subunit change, it craps out with the subunit parser error | 12:59 |
stephenfin | with it, it still craps out but with a proper error | 12:59 |
jawad_axd | Quick question guys: What does cpu allocation 2.0 means ? Is it 2:1? | 13:00 |
*** rpittau|bbl is now known as rpittau | 13:00 | |
stephenfin | saying the packet is > 4k (I think) | 13:00 |
*** damien_r has joined #openstack-nova | 13:01 | |
jawad_axd | and it is safe to change cpu allocation ratio in running environment? | 13:01 |
stephenfin | sorry, 4M https://github.com/testing-cabal/subunit/blob/master/python/subunit/v2.py#L202-L208 | 13:01 |
*** damien_r has quit IRC | 13:01 | |
stephenfin | jawad_axd: yup, 2:1 (20 enabled host CPUs = 40 VCPU inventory) | 13:01 |
*** damien_r has joined #openstack-nova | 13:02 | |
stephenfin | jawad_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 used | 13:02 |
*** nweinber__ has joined #openstack-nova | 13:02 | |
stephenfin | I'm not actually sure what would happen then. I assume the periodic task to update placement's inventory would start failing | 13:02 |
stephenfin | easily tested in a pre-prod environment :) | 13:03 |
sean-k-mooney | stephenfin: ok well if we have a repoducer that means we have a chance of fixing it | 13:03 |
sean-k-mooney | we dont need nova at all we can repodcuse this in stestr | 13:03 |
stephenfin | Exactly. That's pretty damn consistent | 13:04 |
stephenfin | We should be able to but I couldn't do so when I tried | 13:04 |
stephenfin | Probably didn't have the correct fixtures configured or something | 13:04 |
sean-k-mooney | hum ok well i might take a look at this later. and see if i can figure something out | 13:05 |
sean-k-mooney | if we can create a simpler repoducer that woudl be good if not this works | 13:05 |
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 |
stephenfin | I imagine Horizon is pulling that info from the os-hypevisor API which doesn't take overcommit ratios into account | 13:07 |
sean-k-mooney | i think so too | 13:07 |
sean-k-mooney | i think horizon is showing the correct value | 13:08 |
sean-k-mooney | the allocation raitio can be change per host it would not be resonable to have to compute it differnetly per hosts | 13:08 |
jawad_axd | If horizon is showing the same cpu's on host. Does it mean, overcommitment is not being used/applied ? | 13:09 |
sean-k-mooney | no | 13:10 |
sean-k-mooney | over commit is calulated in the scheduler/placment | 13:10 |
sean-k-mooney | it should not be see in horizon | 13:10 |
stephenfin | what sean-k-mooney said | 13:10 |
stephenfin | Horizon might go into negative available values (I'm not sure) but instances will still be scheduled | 13:10 |
sean-k-mooney | jawad_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 2.0 | 13:11 |
*** luksky has quit IRC | 13:11 | |
sean-k-mooney | well on my home system it currently shos 44/24 | 13:11 |
sean-k-mooney | so it does not go into negitiv but the used can exceed the available | 13:12 |
jawad_axd | I 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_axd | I am using stein | 13:12 |
sean-k-mooney | you cannot use more the 24cpu in one guest or in general | 13:12 |
jawad_axd | In general. | 13:13 |
sean-k-mooney | did you set the cpu allcoation ratio in the compute node config | 13:13 |
stephenfin | also, how are you creating the instance? Via the Horizon UI or on the CLI? | 13:13 |
jawad_axd | # Scheduler | 13:13 |
jawad_axd | cpu_allocation_ratio = 2.0 | 13:13 |
sean-k-mooney | ya that does not work in stien | 13:13 |
jawad_axd | From both, cli and horizon | 13:13 |
sean-k-mooney | you hage to set it per compute node | 13:13 |
jawad_axd | ah ok. | 13:13 |
jawad_axd | how to set per compute node? Can you give me some pointers plz? | 13:14 |
sean-k-mooney | https://docs.openstack.org/nova/latest/configuration/config.html#DEFAULT.cpu_allocation_ratio | 13:14 |
sean-k-mooney | you just set cpu_allocation_ratio in the default section | 13:15 |
sean-k-mooney | in the nova.conf | 13:15 |
sean-k-mooney | alternitivly if you want to manage it via the placement api you can use initial_cpu_allocation_ratio | 13:15 |
jawad_axd | So cpu_allocation_ratio in nova.conf at each compute node will work for that compute node? | 13:18 |
sean-k-mooney | correct it should | 13:18 |
sean-k-mooney | what i belive is happening is that the allcoation raition in the placement inventory is set to 1 | 13:19 |
sean-k-mooney | and that is causeign the host to be eliminated before you get to the schduler | 13:19 |
sean-k-mooney | where you set it to 2 | 13:19 |
sean-k-mooney | you could check that | 13:19 |
sean-k-mooney | you can do "openstack resource provider list" | 13:20 |
*** tbachman has joined #openstack-nova | 13:20 | |
sean-k-mooney | then do "openstack resource provider inventory show <RP uuid>" | 13:20 |
rouk | cant use 24 vcore or more per guest? when was this? i have 32vcore guests. | 13:20 |
sean-k-mooney | you can never have more cores in the guest then are on the host but you can have multple guests sharing cores on the host | 13:21 |
*** belmoreira has joined #openstack-nova | 13:22 | |
sean-k-mooney | so on my home server i currently have 44 guest cpus spread over 20 host cpus | 13:22 |
sean-k-mooney | but i could not boot a 44 core vm | 13:22 |
*** udesale has joined #openstack-nova | 13:23 | |
jawad_axd | @sean-k-mooney ok | 13:23 |
*** mkrai_ has joined #openstack-nova | 13:24 | |
*** ociuhandu has joined #openstack-nova | 13:24 | |
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 IRC | 13:29 | |
*** yiipi has joined #openstack-nova | 13:30 | |
sean-k-mooney | jawad_axd: there are some other constrats as well | 13:33 |
sean-k-mooney | if you are using cpu pinning you cannot over commit | 13:33 |
sean-k-mooney | and if you are using a virtual numa toplgoy then within each numa node the cpus cant be over commited | 13:33 |
jawad_axd | ok my hosts have mostly this architecture, Can you have a look at it plz. http://paste.openstack.org/show/788715/ | 13:35 |
sean-k-mooney | ya so you have 4 numa nodes and 12 cores per numa node | 13:36 |
sean-k-mooney | well 12 threads | 13:36 |
jawad_axd | Does it allow overcommitment ? | 13:36 |
sean-k-mooney | sure | 13:37 |
sean-k-mooney | but it depends on the vm | 13:37 |
sean-k-mooney | if the vm has a numa toplogy it impacts how the vm can be placed | 13:37 |
jawad_axd | If vms are using cpu host-model, then? | 13:37 |
sean-k-mooney | no | 13:37 |
rouk | its a flag, since numa is not at all free. | 13:37 |
sean-k-mooney | if the vm flavor has hw:cpu_polciy=dedeicated | 13:37 |
sean-k-mooney | hw:mem_page_size or hw:numa_nodes set or is using vPMEM | 13:38 |
jawad_axd | I am not using numa virtually. | 13:38 |
sean-k-mooney | then it will have a numa toplogy | 13:38 |
*** rouk has quit IRC | 13:38 | |
sean-k-mooney | in that case only the gloabl host core count and allcoation ration for that host matter | 13:38 |
jawad_axd | So I guess, it should be fine for me then. | 13:39 |
*** tbachman has quit IRC | 13:39 | |
sean-k-mooney | yep | 13:39 |
jawad_axd | Thanks !!! | 13:39 |
*** yiipi has quit IRC | 13:39 | |
sean-k-mooney | if 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=1 | 13:40 |
*** rouk has joined #openstack-nova | 13:40 | |
sean-k-mooney | so with that toploy you would only be able to have at most 12 cores in a vm | 13:40 |
sean-k-mooney | but since you are not using numa for the vms it does not matter in your case | 13:41 |
jawad_axd | Okay.. | 13:41 |
sean-k-mooney | the numa feature are just for performance tuning | 13:42 |
sean-k-mooney | if you dont need them you can ignore them | 13:42 |
*** tbachman has joined #openstack-nova | 13:42 | |
jawad_axd | I dont need them now, maybe in future. | 13:42 |
*** ratailor has quit IRC | 13:43 | |
*** shilpasd has joined #openstack-nova | 13:47 | |
*** jaosorior has quit IRC | 13:47 | |
*** mkrai_ has quit IRC | 13:50 | |
efried | nova meeting now-ish | 13:59 |
*** luksky has joined #openstack-nova | 13:59 | |
shilpasd | efried: hi | 13:59 |
efried | hello | 13:59 |
*** ociuhandu has joined #openstack-nova | 14:00 | |
shilpasd | efried: dansmith: here added one question, https://review.opendev.org/#/c/694462/8/nova/db/sqlalchemy/api.py@2008 | 14:03 |
shilpasd | help me to understand 'any issues in setting 'volume_backed' field on the existing instances in the upgrade db script?' | 14:04 |
*** eharney has quit IRC | 14:04 | |
efried | shilpasd: 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 |
efried | IOW this may have to wait until next week to close on. | 14:05 |
shilpasd | efried: ok, thanks | 14:05 |
*** rchurch has joined #openstack-nova | 14:05 | |
shilpasd | will wait for them | 14:06 |
*** sridharg has quit IRC | 14:07 | |
*** jmlowe has joined #openstack-nova | 14:09 | |
*** mgariepy has quit IRC | 14:10 | |
*** rchurch has quit IRC | 14:10 | |
*** mgariepy has joined #openstack-nova | 14:12 | |
sean-k-mooney | well that was quick :) | 14:12 |
shilpasd | :] (y) | 14:12 |
sean-k-mooney | looking at dansmith's suggestion i tened to agree that offloadign this to sql or adding a new field would be better | 14:13 |
sean-k-mooney | you could perhaps store the "is_volume_backed" info in the system_metadata table | 14:16 |
sean-k-mooney | that woudl avoid the need for a schema change but i think option 1 makes the most sense too | 14:17 |
sean-k-mooney | the only thing is we would have to keep the logic to handel old instance for a few release for FFU | 14:17 |
sean-k-mooney | but eventurlaly we could drop it | 14:18 |
shilpasd | sean-k-mooney: thanks for your opinion | 14:18 |
shilpasd | please clear me last point 'keep the logic to handel old instance for a few release for FFU' | 14:18 |
*** rchurch has joined #openstack-nova | 14:18 | |
*** jmlowe has quit IRC | 14:18 | |
sean-k-mooney | we "support" skip level upgrade via Fast Forward upgrdes at lest downstream in OSP/tripleo | 14:19 |
sean-k-mooney | so if we add the code handel loading old instances and calulating the value | 14:20 |
sean-k-mooney | we need to keep it for a cople of release before we can remove it | 14:20 |
sean-k-mooney | and safly assume it will always be set | 14:20 |
sean-k-mooney | eventually that code for old instance can be removed but not for 2-3 releases | 14:20 |
*** alistarle has joined #openstack-nova | 14:21 | |
shilpasd | for old instances will update 'volume_backed' based on bdms.volume and bdms.boot_index | 14:22 |
*** rchurch has quit IRC | 14:23 | |
*** jmlowe has joined #openstack-nova | 14:23 | |
shilpasd | so still not clear why we need to keep code for couple of release, please elaborate further | 14:23 |
sean-k-mooney | its technically also stored in the request spec for instance created after rocky | 14:23 |
sean-k-mooney | FFU do not require the compute agents to start during the upgrade | 14:24 |
sean-k-mooney | so you can go form qeens to train without ever starting the agent untill you get to train | 14:24 |
sean-k-mooney | so unless you are groin to write an online data migration for this | 14:25 |
sean-k-mooney | which is not what dansmith was suggesting | 14:25 |
sean-k-mooney | if 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 ussuri | 14:25 |
sean-k-mooney | start the agent and have it fix the instnaces | 14:25 |
sean-k-mooney | so 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 3 | 14:27 |
*** rchurch has joined #openstack-nova | 14:27 | |
*** nweinber__ has quit IRC | 14:28 | |
*** mkrai_ has joined #openstack-nova | 14:28 | |
shilpasd | ok, got it | 14:29 |
*** mgariepy has quit IRC | 14:30 | |
*** Sundar has joined #openstack-nova | 14:32 | |
*** Luzi has quit IRC | 14:32 | |
*** links has quit IRC | 14:35 | |
*** Sundar has quit IRC | 14:38 | |
*** ociuhandu has quit IRC | 14:39 | |
*** ociuhandu has joined #openstack-nova | 14:40 | |
*** mgariepy has joined #openstack-nova | 14:43 | |
*** jawad_axd has quit IRC | 14:45 | |
*** ociuhandu has quit IRC | 14:46 | |
*** alistarle has quit IRC | 14:47 | |
*** iurygregory has quit IRC | 14:49 | |
*** jawad_axd has joined #openstack-nova | 14:49 | |
*** bnemec has quit IRC | 14:51 | |
*** jawad_axd has quit IRC | 14:54 | |
*** shilpasd has quit IRC | 15:01 | |
*** tbachman has quit IRC | 15:01 | |
*** mriedem has joined #openstack-nova | 15:02 | |
*** pcaruana has quit IRC | 15:04 | |
*** damien_r has quit IRC | 15:06 | |
*** damien_r has joined #openstack-nova | 15:06 | |
sean-k-mooney | by the way we can recheck stuff currently right | 15:09 |
sean-k-mooney | there is not specific gate blockers at the moment out side of the intermitenly failing jobs | 15:09 |
*** eharney has joined #openstack-nova | 15:10 | |
*** efried has quit IRC | 15:12 | |
*** lpetrut has quit IRC | 15:12 | |
*** priteau has joined #openstack-nova | 15:12 | |
*** nweinber__ has joined #openstack-nova | 15:12 | |
*** psachin has quit IRC | 15:13 | |
*** efried has joined #openstack-nova | 15:13 | |
*** ociuhandu has joined #openstack-nova | 15:23 | |
*** nweinber__ has quit IRC | 15:24 | |
*** nweinber__ has joined #openstack-nova | 15:27 | |
*** nweinber_ has joined #openstack-nova | 15:29 | |
*** nweinber__ has quit IRC | 15:31 | |
*** tbachman has joined #openstack-nova | 15:35 | |
*** priteau has quit IRC | 15:51 | |
*** artom has joined #openstack-nova | 15:52 | |
*** lbragstad has quit IRC | 15:54 | |
*** mlavalle has joined #openstack-nova | 16:00 | |
*** lbragstad has joined #openstack-nova | 16:00 | |
stephenfin | sean-k-mooney: not that I'm aware of, no | 16:03 |
*** alex_xu has quit IRC | 16:03 | |
*** jmlowe has quit IRC | 16:08 | |
*** gyee has joined #openstack-nova | 16:14 | |
*** mkrai_ has quit IRC | 16:18 | |
*** jmlowe has joined #openstack-nova | 16:21 | |
*** eharney has quit IRC | 16:22 | |
*** mrch has quit IRC | 16:23 | |
*** jawad_axd has joined #openstack-nova | 16:24 | |
KeithMnemonic | good 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 IRC | 16:26 | |
melwitt | KeithMnemonic: 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 wrong | 16:28 |
*** tosky has quit IRC | 16:28 | |
KeithMnemonic | are you ok with me continuing rechecks to see if it passes eventually | 16:29 |
*** TxGirlGeek has joined #openstack-nova | 16:29 | |
openstackgerrit | Merged openstack/nova master: nova-net: Make the security group API a module https://review.opendev.org/696746 | 16:30 |
*** lbragstad has quit IRC | 16:30 | |
melwitt | KeithMnemonic: yeah, you can | 16:32 |
KeithMnemonic | thanks | 16:32 |
*** jmlowe has quit IRC | 16:35 | |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unnecessary exception handling https://review.opendev.org/703973 | 16:36 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unused parameters https://review.opendev.org/703974 | 16:36 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-net: Remove unused nova-network objects https://review.opendev.org/697156 | 16:36 |
stephenfin | efried: You saw https://zuul.opendev.org/t/openstack/build/471f2388615e4ee5806a85e19b79569d ? | 16:38 |
stephenfin | Have you a patch up to address that yet? | 16:38 |
* efried clicks | 16:38 | |
stephenfin | Seeing as you were attacking them in a piecemeal fashion? | 16:38 |
efried | stephenfin: I hadn't seen that. I can throw out a patch quick... | 16:39 |
sean-k-mooney | melwitt: i got side tracked with backports so i did not look at it properly so ill try to do it today | 16:39 |
stephenfin | efried: if you could, that'd be good, yup :) | 16:39 |
melwitt | thanks sean-k-mooney | 16:39 |
*** tesseract has quit IRC | 16:41 | |
*** maciejjozefczyk has quit IRC | 16:45 | |
*** TxGirlGeek has quit IRC | 16:55 | |
*** artom has quit IRC | 16:56 | |
*** derekh has quit IRC | 16:57 | |
*** jawad_axd has quit IRC | 16:57 | |
*** tbachman has quit IRC | 16:58 | |
*** tbachman has joined #openstack-nova | 17:00 | |
*** TxGirlGeek has joined #openstack-nova | 17:01 | |
*** mkrai has joined #openstack-nova | 17:02 | |
*** jmlowe has joined #openstack-nova | 17:05 | |
*** udesale has quit IRC | 17:05 | |
*** lbragstad has joined #openstack-nova | 17:10 | |
openstackgerrit | Merged openstack/nova master: Remove remaining Python 2.7-only dependencies https://review.opendev.org/698895 | 17:12 |
* efried lunches | 17:12 | |
*** deke997 has joined #openstack-nova | 17:15 | |
deke997 | Hey guys | 17:15 |
deke997 | I 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 files | 17:15 |
*** tbachman has quit IRC | 17:19 | |
*** rchurch has quit IRC | 17:20 | |
*** rpittau is now known as rpittau|afk | 17:21 | |
*** dtantsur is now known as dtantsur|afk | 17:24 | |
*** jmlowe has quit IRC | 17:25 | |
*** luksky has quit IRC | 17:26 | |
*** artom has joined #openstack-nova | 17:31 | |
*** mkrai has quit IRC | 17:31 | |
*** ociuhandu_ has joined #openstack-nova | 17:32 | |
*** evrardjp has quit IRC | 17:34 | |
*** evrardjp has joined #openstack-nova | 17:34 | |
stephenfin | deke997: What are you trying to achieve? | 17:35 |
*** ociuhandu has quit IRC | 17:36 | |
*** rchurch has joined #openstack-nova | 17:36 | |
*** ociuhandu_ has quit IRC | 17:37 | |
deke997 | So 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 0000:00:05.1 | 17:38 |
deke997 | by manually editing the libvirt.xml file, I verified that it works properly in the VM when set up like that | 17:38 |
deke997 | However, 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 0000:00:06.0 | 17:39 |
deke997 | So 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 VM | 17:40 |
*** jmlowe has joined #openstack-nova | 17:41 | |
deke997 | stephenfin 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 address | 17:47 |
stephenfin | I'm trying to root that out at the moment | 17:47 |
yoctozepto | deke997: btw, is the device in question a secret or can you shed some light on it? | 17:48 |
sean-k-mooney | deke997: we dont do all the generation in nova | 17:49 |
sean-k-mooney | we actully leave it up to libvirt to assign the guesst pci adress and as a result define the toplogy | 17:50 |
sean-k-mooney | deke997: 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 vms | 17:51 |
*** KeithMnemonic has quit IRC | 17:52 | |
sean-k-mooney | but 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 guest | 17:52 |
sean-k-mooney | deke997: this would be well into the catagory of a spec not a bugfix in my view | 17:52 |
*** jmlowe has quit IRC | 17:53 | |
sean-k-mooney | we 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 confirm | 17:53 |
deke997 | sean-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 up | 17:53 |
deke997 | libvirt has a multifunction=on setting too | 17:54 |
sean-k-mooney | how to you request and whitelist the device at present | 17:55 |
sean-k-mooney | i assume you are using the pci whitelist and a flavor alias | 17:55 |
deke997 | yes | 17:55 |
deke997 | but I have to use a separate alias for each function | 17:55 |
sean-k-mooney | and you have 2 aliase right | 17:55 |
sean-k-mooney | one for each funciton | 17:55 |
sean-k-mooney | ya | 17:55 |
deke997 | We could add a multifunction alias type | 17:55 |
sean-k-mooney | so how are you ensuring the two VF/PF come form the same device on the hsot | 17:56 |
sean-k-mooney | deke997: we could add a tage but unless we can adress ^ | 17:56 |
sean-k-mooney | we have bigger problems | 17:56 |
deke997 | right now, I am just using all of them, so it's not an issue haha | 17:56 |
sean-k-mooney | so you have only 1 device on the compute node and your booting 1 vm that use all the fucntion form that device | 17:57 |
deke997 | yes that's what I have set up for testing | 17:57 |
sean-k-mooney | ya in that specific case you whave worked around that issue | 17:57 |
deke997 | But in the libvirt XML it is very easy to see which devices from the host were passed through | 17:57 |
deke997 | and match them | 17:57 |
sean-k-mooney | this has come up in the context of gpus in the past | 17:58 |
sean-k-mooney | the GPU and audio contoler are usally 2 differnt pci endpoint on the same device | 17:58 |
sean-k-mooney | at pressent we dont have a way to passthough both form the same card and ensure they are form the same device | 17:58 |
*** psachin has joined #openstack-nova | 17:59 | |
sean-k-mooney | deke997: 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 pair | 17:59 |
deke997 | yea exactly | 18:00 |
*** TxGirlGeek has quit IRC | 18:00 | |
deke997 | That doesn't seem super difficult | 18:00 |
sean-k-mooney | and 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 host | 18:00 |
sean-k-mooney | deke997: you say that now | 18:01 |
deke997 | haha | 18:01 |
deke997 | Is there a way to make nova play nice with edited libvirt.xmls? | 18:01 |
sean-k-mooney | so this all live in the pci manger code https://github.com/openstack/nova/tree/master/nova/pci | 18:01 |
sean-k-mooney | no | 18:01 |
sean-k-mooney | nova regenerates the xml based on seveal api actions | 18:02 |
sean-k-mooney | like hard reboot | 18:02 |
sean-k-mooney | so your eddit would last until someone stops and start the vm | 18:02 |
*** derekh has joined #openstack-nova | 18:02 | |
deke997 | Is there a way to disable that regeneration? | 18:02 |
sean-k-mooney | no | 18:03 |
*** rchurch_ has joined #openstack-nova | 18:03 | |
openstackgerrit | Merged openstack/nova master: libvirt: Add a default VirtIO-RNG device to guests https://review.opendev.org/616603 | 18:03 |
*** jaosorior has joined #openstack-nova | 18:03 | |
sean-k-mooney | so 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 updated | 18:04 |
sean-k-mooney | the alias schema coudl certenly be extended to referen a child device https://github.com/openstack/nova/blob/master/nova/pci/request.py#L66-L99 | 18:05 |
deke997 | How would we go about updating the xml generation? | 18:05 |
*** rchurch_ has quit IRC | 18:06 | |
sean-k-mooney | am 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 funcion | 18:06 |
*** rchurch_ has joined #openstack-nova | 18:06 | |
*** rchurch has quit IRC | 18:06 | |
*** rchurch_ has quit IRC | 18:08 | |
*** rchurch has joined #openstack-nova | 18:08 | |
*** TxGirlGeek has joined #openstack-nova | 18:08 | |
deke997 | I need to know a bit more about how libvirt chooses those addresses | 18:08 |
sean-k-mooney | so you would have to modify https://github.com/openstack/nova/blob/6db486e9fd4f6b8dd02371b043e66808cdd1e0cd/nova/virt/libvirt/driver.py#L5871 | 18:09 |
*** rchurch has quit IRC | 18:09 | |
sean-k-mooney | the general docs for thi in libvirt are https://libvirt.org/formatdomain.html#elementsHostDevSubsys | 18:10 |
*** rchurch has joined #openstack-nova | 18:10 | |
sean-k-mooney | normally we just generate a hostdev element like this | 18:11 |
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-mooney | so we specify the source element and the host pci adress and libvirt chooes a target adress | 18:11 |
sean-k-mooney | that i assume is what you were changing with your manual edit | 18:12 |
deke997 | Yes | 18:12 |
deke997 | but I could only start that instance manually through libvirt after the edit, because nova just tries to overwrite it | 18:12 |
*** amodi has quit IRC | 18:12 | |
sean-k-mooney | https://libvirt.org/formatdomain.html#elementsAddress this is the docs for the device addressing | 18:13 |
deke997 | Also, adding multifunction = 'on' to the source address may be necessary. Libvirt docs say to use it, but I was able to have success either way | 18:13 |
deke997 | which would have to be done in nova | 18:14 |
deke997 | would be easy to just add multifunction = yes iff alias_type = multifunction or something | 18:15 |
sean-k-mooney | ya the multifunction bit is not what im worried about | 18:16 |
sean-k-mooney | i think the only way to make this work si to have nova set the target Adress on both devices | 18:17 |
sean-k-mooney | but since nova does not track the target adress today that is a prbalem | 18:17 |
sean-k-mooney | also you would have to take into account both q35 and the i440x machine types | 18:17 |
sean-k-mooney | the old i440x or pc machine type is pci only | 18:18 |
sean-k-mooney | and q35 is pcie by default but you can add pci busses too | 18:18 |
sean-k-mooney | so that makes it more complicated | 18:18 |
deke997 | definitely | 18:19 |
sean-k-mooney | with q35 if we wanted to set <address domain='0x0000' bus='0x06' slot='0x02' function='0x0'/> | 18:20 |
sean-k-mooney | for the target we would need to first create a pcie root port for it | 18:20 |
sean-k-mooney | if we leave out the target element libvirt does it for us | 18:21 |
sean-k-mooney | so baseicly when you specify it you also have to do the stuff libvirt does for you automatically | 18:21 |
sean-k-mooney | basically | 18:22 |
sean-k-mooney | devices> | 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-mooney | well it would be multifucniton=one in this case | 18:22 |
sean-k-mooney | but once we have the multi funtion bridge device you can add the enpoint beneth it | 18:22 |
sean-k-mooney | i guess the only real difference that mater for q35 vs pc is pcie-bridge vs pci-bridge in the contoler element | 18:24 |
deke997 | Okay 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 match | 18:24 |
deke997 | :') | 18:24 |
*** jmlowe has joined #openstack-nova | 18:24 | |
sean-k-mooney | am first reaction i dont like that but thinking it though. | 18:25 |
sean-k-mooney | we could defien the domain without starting it | 18:25 |
deke997 | yes | 18:25 |
sean-k-mooney | that would cause libvirt to fill in the details i think | 18:25 |
sean-k-mooney | then we could read it back and ed the pair. maybe | 18:25 |
deke997 | all we'd have to do is check for two source lines that match and then make the targets match too | 18:26 |
sean-k-mooney | but it kind fo feels like a hack and it like wont scale well if there are multile deivecs | 18:26 |
deke997 | it is definitely a hack | 18:26 |
deke997 | why won't it scale well? | 18:26 |
deke997 | never gonna have more than ~8 devices in a server | 18:27 |
sean-k-mooney | if 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 complcated | 18:27 |
sean-k-mooney | well the bit after the . is in octal so we can on ly have 8 fucntion per slot | 18:27 |
sean-k-mooney | so if you needed more then 8 virtual funtion it would not work right | 18:28 |
deke997 | I don't think such a device exists | 18:28 |
deke997 | haven't ever seen more than like 4 | 18:28 |
deke997 | my use case is only 2 | 18:28 |
sean-k-mooney | ya i have seen up to 8 well no i have seen way more | 18:28 |
sean-k-mooney | intel nics when using sriov can have 64 - 128 vf per pf | 18:29 |
sean-k-mooney | but they do the adressing slightly differently | 18:29 |
deke997 | sriov is totally different tho | 18:29 |
sean-k-mooney | ya for there case teh assing a full bus per card | 18:29 |
deke997 | this would be specifically for non-sriov multi-function devices | 18:29 |
efried | stephenfin: 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-mooney | so you have mulitple slots and funtion to support up to 256 vf totall | 18:30 |
sean-k-mooney | deke997: ack | 18:30 |
sean-k-mooney | deke997: so you can see why i think this would need a spec | 18:30 |
sean-k-mooney | to design correctly | 18:30 |
deke997 | yes, I'm just hoping for an intermediate band aid while we work on a spec | 18:31 |
sean-k-mooney | i 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 together | 18:31 |
sean-k-mooney | e.g. the gpu and the audio contole on the gpu | 18:31 |
*** derekh has quit IRC | 18:32 | |
sean-k-mooney | nviad might actully be exposing it as a multi function device so that might actully be the same usecase | 18:32 |
*** martinkennelly has quit IRC | 18:32 | |
sean-k-mooney | ya they are | 18:35 |
sean-k-mooney | 81:00.0 VGA compatible controller: NVIDIA Corporation GK104GL [Quadro K5000] (rev a1) | 18:35 |
sean-k-mooney | 81:00.1 Audio device: NVIDIA Corporation GK104 HDMI Audio Controller (rev a1) | 18:35 |
sean-k-mooney | so if we we made ^ work it would work for you too right | 18:35 |
*** eharney has joined #openstack-nova | 18:36 | |
deke997 | yes | 18:38 |
deke997 | That looks pretty much identical to my use case | 18:38 |
sean-k-mooney | ok what i would basicaly suggest is as follows. extent the alias with a child tag that can reference other alisas. | 18:42 |
sean-k-mooney | when you add an alias to a flaovr which has a child tag we woudl claim the deivce and all its childern | 18:42 |
*** psachin has quit IRC | 18:43 | |
sean-k-mooney | and expose them as a multi funcitn device | 18:43 |
sean-k-mooney | the 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 slot | 18:44 |
sean-k-mooney | the pci manger would also need to updated so that when we do the pci claim it claims the pair form the same slot | 18:44 |
deke997 | wouldn't that always be the case? | 18:44 |
deke997 | I don't see why the passthrough filter needs to be updated | 18:45 |
sean-k-mooney | well if you put two of those card in the same server it would break | 18:45 |
sean-k-mooney | as we coudl get 1 fucntion of type a form one card and one fucntion of type b from another | 18:46 |
sean-k-mooney | so the filter and assignment code would have to be hardened to prevent that | 18:46 |
*** TxGirlGeek has quit IRC | 18:48 | |
deke997 | The assignment code, yes, but the filter, I don't think so | 18:48 |
deke997 | because before assignment, it would always be true that both the parent and child could be allocated together | 18:48 |
sean-k-mooney | its true if you alwasy need to request both | 18:49 |
sean-k-mooney | which is true in your case | 18:49 |
deke997 | right | 18:49 |
sean-k-mooney | not in the gpu case | 18:49 |
deke997 | I see | 18:49 |
deke997 | in that case, then yea the filter too | 18:50 |
sean-k-mooney | anyway the assignment code and filter share the same code | 18:50 |
sean-k-mooney | the filter basicaly passes if it would be able to assign | 18:50 |
*** openstackgerrit has quit IRC | 18:51 | |
sean-k-mooney | deke997: https://github.com/openstack/nova/blob/master/nova/scheduler/filters/pci_passthrough_filter.py#L51 | 18:51 |
sean-k-mooney | see we just loop over the request and ask can it suppor the request | 18:52 |
sean-k-mooney | well the loopin is don internally in the suport_request funtion | 18:52 |
sean-k-mooney | we 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-L375 | 18:53 |
*** jaosorior has quit IRC | 18:54 | |
sean-k-mooney | so if you update _apply_request to handel the assignment constriatt the filter gets updated for free | 18:54 |
deke997 | got it | 18:55 |
deke997 | are 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 |
deke997 | These are the only two options so far, yea? | 18:56 |
sean-k-mooney | the only two i am aware of. if we were to do this updatream i would prefer to do it properly | 18:56 |
sean-k-mooney | and have nova do it. if you need to do it downstream quick well that is up to you | 18:57 |
sean-k-mooney | deke997: what is the eta you had in mind for this | 18:57 |
deke997 | 2 days ago haha | 18:57 |
deke997 | I won't be sleeping much till I get this working | 18:58 |
sean-k-mooney | ya fair warning to land this upstream its proably going to be next cycle | 18:58 |
deke997 | yea so I think we'll do both | 18:58 |
deke997 | downstream quick, and then work on upstream | 18:58 |
sean-k-mooney | i could be ussuri if it was really pusshed for but this code is complex and a pain to test | 18:58 |
deke997 | when is the next cycle? | 18:58 |
sean-k-mooney | it starts in 3 monts os like september/october | 18:59 |
sean-k-mooney | there is still about 2-3 weeks to propose specs for this cycle | 18:59 |
sean-k-mooney | you would need to figure out what need to chagne and write up a propals and get it review by then | 18:59 |
sean-k-mooney | then there is able 2 months left to land the feature | 19:00 |
sean-k-mooney | but beign realisting if you need this quick | 19:00 |
sean-k-mooney | then likely createing a poc and upstreaming it next cycle is what your going to have to do | 19:00 |
deke997 | One thing to check on: | 19:01 |
sean-k-mooney | for the poc you could do the hack where you fix up the adresses | 19:01 |
deke997 | I think cyborg is doing some stuff with multi function too | 19:01 |
deke997 | I saw it while doing research | 19:01 |
sean-k-mooney | ya so cyborg would be another option | 19:02 |
sean-k-mooney | if the integration is done this cycle | 19:02 |
sean-k-mooney | then you could write a device dirver for you custom multifunion device | 19:02 |
sean-k-mooney | but the xml change would still need to be done | 19:02 |
sean-k-mooney | you would not have to touch the filer or pci manager in that case | 19:03 |
sean-k-mooney | but you would need to do all the cyborg work | 19:03 |
sean-k-mooney | deke997: cyborg support in nova is planned for this cycle but we said that last cycle too | 19:04 |
sean-k-mooney | deke997: its much closer to being read this time however | 19:05 |
deke997 | Good to know | 19:06 |
deke997 | I need to look into how they're implementing multi function a bit more | 19:07 |
deke997 | but even so, I can't wait till the next cycle to get a beta working here | 19:08 |
sean-k-mooney | at present without the libvirt support for multifuntion devices in nova cybporg cant support what you need | 19:09 |
deke997 | I think the options in order of increasing time, complexity, and correctness would probably be | 19:09 |
deke997 | 1. hack | 19:09 |
*** TxGirlGeek has joined #openstack-nova | 19:09 | |
deke997 | 2. nova | 19:09 |
deke997 | 3. cyborg | 19:09 |
sean-k-mooney | 0 heiring an intern to manually update every vm as it spawns | 19:09 |
deke997 | hahahah | 19:10 |
deke997 | but I can't even do that | 19:10 |
deke997 | the xml changes don't go into effect till you reboot | 19:10 |
sean-k-mooney | actully there is one other hack you cloud do | 19:10 |
deke997 | and when you reboot, it gets overwritten | 19:10 |
deke997 | is 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-mooney | for reasons in the past i did have need to do horible things | 19:11 |
sean-k-mooney | https://opendev.org/x/networking-ovs-dpdk/src/branch/master/devstack/libs/ovs-dpdk#L79-L130 | 19:11 |
sean-k-mooney | that is an example of a script that replace the qemu executable and rewrites the qemu commandline on the fly | 19:12 |
*** tbachman has joined #openstack-nova | 19:12 | |
*** heso has quit IRC | 19:13 | |
sean-k-mooney | what 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 set | 19:13 |
sean-k-mooney | you could install a qemu wrapper script as a hack to match on parts of the qemu command line and alter it | 19:14 |
sean-k-mooney | it would have 0 support form anyone but it is a -1 on your scale | 19:14 |
deke997 | haha i'll have to look into that too i guess | 19:17 |
sean-k-mooney | adding flags to existing optins is nasty but actully kind of simple | 19:17 |
sean-k-mooney | doing what you need to do would be harder | 19:17 |
sean-k-mooney | but at the end of the day the commandline is fairly deterministic so it would be doable | 19:18 |
sean-k-mooney | definelty not my first choise but we dont always get to do our first choice | 19:18 |
deke997 | I'll see what I can figure out tonight and get back to you soon with the progress | 19:19 |
deke997 | lmk if you have any other ideas | 19:20 |
*** gmann is now known as gmann_lunch | 19:26 | |
*** jmlowe has quit IRC | 19:27 | |
*** jamesdenton has quit IRC | 19:29 | |
*** jamesdenton has joined #openstack-nova | 19:29 | |
*** rchurch has quit IRC | 19:46 | |
*** gmann_lunch is now known as gmann | 19:52 | |
*** damien_r has quit IRC | 19:52 | |
*** rchurch has joined #openstack-nova | 19:53 | |
*** rchurch has quit IRC | 19:54 | |
*** rchurch has joined #openstack-nova | 19:54 | |
*** eharney has quit IRC | 19:55 | |
*** jmlowe has joined #openstack-nova | 20:08 | |
*** openstackstatus has joined #openstack-nova | 20:08 | |
*** ChanServ sets mode: +v openstackstatus | 20:08 | |
*** jaosorior has joined #openstack-nova | 20:10 | |
*** jmlowe has quit IRC | 20:13 | |
*** luksky has joined #openstack-nova | 20:36 | |
*** bbowen has quit IRC | 20:47 | |
*** nweinber_ has quit IRC | 20:52 | |
*** artom has quit IRC | 20:58 | |
*** eharney has joined #openstack-nova | 21:01 | |
*** eharney has quit IRC | 21:02 | |
*** eharney has joined #openstack-nova | 21:02 | |
*** rcernin has joined #openstack-nova | 21:04 | |
*** johanssone_ has joined #openstack-nova | 21:05 | |
*** jdillaman has joined #openstack-nova | 21:08 | |
*** dansmith has quit IRC | 21:18 | |
*** irclogbot_2 has quit IRC | 21:18 | |
*** jhesketh has quit IRC | 21:18 | |
*** adriant has quit IRC | 21:18 | |
*** panda has quit IRC | 21:18 | |
*** jkulik has quit IRC | 21:18 | |
*** dklyle has quit IRC | 21:19 | |
*** johanssone has quit IRC | 21:19 | |
*** stephenfin has quit IRC | 21:19 | |
*** mgoddard has quit IRC | 21:19 | |
*** haleyb has quit IRC | 21:19 | |
*** adrianc has quit IRC | 21:19 | |
*** admin0 has quit IRC | 21:19 | |
*** tonyb has quit IRC | 21:19 | |
*** lchabert has quit IRC | 21:19 | |
*** gryf has quit IRC | 21:19 | |
*** hemna has quit IRC | 21:19 | |
*** openstackstatus has quit IRC | 21:20 | |
*** dklyle has joined #openstack-nova | 21:21 | |
*** jaosorior has quit IRC | 21:25 | |
*** irclogbot_3 has joined #openstack-nova | 21:26 | |
*** tonyb has joined #openstack-nova | 21:26 | |
*** eharney has quit IRC | 21:29 | |
*** jawad_axd has joined #openstack-nova | 21:38 | |
*** jawad_axd has quit IRC | 21:42 | |
*** mgariepy has quit IRC | 21:43 | |
*** efried has quit IRC | 21:44 | |
*** efried has joined #openstack-nova | 21:44 | |
*** bbowen has joined #openstack-nova | 21:50 | |
*** jawad_axd has joined #openstack-nova | 21:59 | |
*** dansmith has joined #openstack-nova | 22:01 | |
*** jhesketh has joined #openstack-nova | 22:01 | |
*** adriant has joined #openstack-nova | 22:01 | |
*** panda has joined #openstack-nova | 22:01 | |
*** jkulik has joined #openstack-nova | 22:01 | |
*** stephenfin has joined #openstack-nova | 22:01 | |
*** mgoddard has joined #openstack-nova | 22:01 | |
*** haleyb has joined #openstack-nova | 22:01 | |
*** admin0 has joined #openstack-nova | 22:01 | |
*** adrianc has joined #openstack-nova | 22:01 | |
*** lchabert has joined #openstack-nova | 22:01 | |
*** gryf has joined #openstack-nova | 22:01 | |
*** hemna has joined #openstack-nova | 22:01 | |
*** jawad_axd has quit IRC | 22:03 | |
*** tosky has joined #openstack-nova | 22:16 | |
*** xek has quit IRC | 22:22 | |
*** artom has joined #openstack-nova | 22:24 | |
*** slaweq_ has joined #openstack-nova | 22:45 | |
*** jmlowe has joined #openstack-nova | 22:53 | |
*** TxGirlGeek has quit IRC | 23:01 | |
*** slaweq_ has quit IRC | 23:01 | |
*** TxGirlGeek has joined #openstack-nova | 23:02 | |
*** tkajinam has joined #openstack-nova | 23:07 | |
*** jmlowe has quit IRC | 23:07 | |
*** mlavalle has quit IRC | 23:12 | |
*** rchurch has quit IRC | 23:32 | |
*** rchurch has joined #openstack-nova | 23:35 | |
*** dviroel has quit IRC | 23:39 | |
*** tbachman has quit IRC | 23:48 | |
*** rchurch has quit IRC | 23:50 | |
*** macz has joined #openstack-nova | 23:50 | |
*** rchurch has joined #openstack-nova | 23:52 | |
*** macz has quit IRC | 23:54 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!