| *** slaweq has joined #openstack-nova | 00:16 | |
| *** slaweq has quit IRC | 00:21 | |
| *** gyee has quit IRC | 00:28 | |
| melwitt | alex_xu: hey, would appreciate to have your thoughts on the unified limits spec around the proxy API/deprecation and migration parts https://review.opendev.org/602201 | 00:40 |
|---|---|---|
| melwitt | *especially the proxy API/deprecation and migration parts | 00:40 |
| openstackgerrit | Merged openstack/nova stable/rocky: Init HostState.failed_builds https://review.opendev.org/668520 | 00:45 |
| openstackgerrit | Merged openstack/nova master: Stabilize unshelve notification sample tests https://review.opendev.org/668675 | 00:45 |
| *** brinzhang has joined #openstack-nova | 00:47 | |
| *** ricolin_ has joined #openstack-nova | 00:51 | |
| *** ricolin_ is now known as ricolin | 00:52 | |
| *** tbachman has quit IRC | 01:02 | |
| *** panda has quit IRC | 01:03 | |
| *** tbachman has joined #openstack-nova | 01:05 | |
| *** panda has joined #openstack-nova | 01:05 | |
| *** imacdonn has quit IRC | 01:11 | |
| *** imacdonn has joined #openstack-nova | 01:11 | |
| *** Kevin_Zheng has joined #openstack-nova | 01:12 | |
| openstackgerrit | Merged openstack/nova stable/queens: Fail to live migration if instance has a NUMA topology https://review.opendev.org/629597 | 01:21 |
| *** lbragstad has quit IRC | 02:04 | |
| openstackgerrit | Merged openstack/nova stable/queens: libvirt: Avoid using os-brick encryptors when device_path isn't provided https://review.opendev.org/656464 | 02:06 |
| *** dasp has quit IRC | 02:46 | |
| *** dasp has joined #openstack-nova | 03:03 | |
| *** StevenK has quit IRC | 03:08 | |
| *** psachin has joined #openstack-nova | 03:34 | |
| *** StevenK has joined #openstack-nova | 03:37 | |
| *** ricolin_ has joined #openstack-nova | 03:54 | |
| *** udesale has joined #openstack-nova | 03:56 | |
| *** ricolin has quit IRC | 03:57 | |
| *** _alastor_ has quit IRC | 04:25 | |
| *** toabctl has quit IRC | 04:31 | |
| *** whoami-rajat has joined #openstack-nova | 04:34 | |
| *** shilpasd has joined #openstack-nova | 04:34 | |
| *** brinzhang_ has joined #openstack-nova | 04:47 | |
| *** brinzhang has quit IRC | 04:50 | |
| *** tkajinam has quit IRC | 05:01 | |
| *** brinzhang_ has quit IRC | 05:05 | |
| *** brinzhang_ has joined #openstack-nova | 05:05 | |
| *** pcaruana has joined #openstack-nova | 05:06 | |
| *** ratailor has joined #openstack-nova | 05:23 | |
| *** ratailor_ has joined #openstack-nova | 05:29 | |
| *** ratailor has quit IRC | 05:32 | |
| *** shilpasd has quit IRC | 05:36 | |
| *** ricolin_ is now known as ricolin | 05:37 | |
| *** udesale has quit IRC | 05:38 | |
| *** igordc has quit IRC | 05:41 | |
| *** artom has joined #openstack-nova | 05:41 | |
| *** ratailor_ has quit IRC | 05:41 | |
| *** artom is now known as artom|gmtplus3 | 05:41 | |
| *** shilpasd has joined #openstack-nova | 05:44 | |
| *** ratailor has joined #openstack-nova | 05:47 | |
| *** Luzi has joined #openstack-nova | 05:54 | |
| *** tkajinam has joined #openstack-nova | 06:00 | |
| openstackgerrit | Artom Lifshitz proposed openstack/nova master: [WIP] Introduce live_migration_claim() https://review.opendev.org/635669 | 06:17 |
| openstackgerrit | Artom Lifshitz proposed openstack/nova master: New objects for NUMA live migration https://review.opendev.org/634827 | 06:17 |
| openstackgerrit | Artom Lifshitz proposed openstack/nova master: LM: add support for sending NUMAMigrateData to the source https://review.opendev.org/634828 | 06:17 |
| openstackgerrit | Artom Lifshitz proposed openstack/nova master: LM: add support for updating NUMA-related XML on the source https://review.opendev.org/635229 | 06:17 |
| openstackgerrit | Artom Lifshitz proposed openstack/nova master: RPC changes to prepare for NUMA live migration https://review.opendev.org/634605 | 06:17 |
| openstackgerrit | Artom Lifshitz proposed openstack/nova master: NUMA live migration support https://review.opendev.org/634606 | 06:17 |
| openstackgerrit | Artom Lifshitz proposed openstack/nova master: CONF.enable_numa_live_migration is not needed >= Train https://review.opendev.org/640021 | 06:17 |
| *** luksky has joined #openstack-nova | 06:27 | |
| *** rha has quit IRC | 06:28 | |
| *** maciejjozefczyk has joined #openstack-nova | 06:36 | |
| *** shilpasd has quit IRC | 06:38 | |
| *** belmoreira has joined #openstack-nova | 06:39 | |
| *** slaweq has joined #openstack-nova | 06:46 | |
| *** rdopiera has joined #openstack-nova | 06:48 | |
| *** luksky has quit IRC | 06:48 | |
| *** ricolin_ has joined #openstack-nova | 06:51 | |
| *** maciejjozefczyk has quit IRC | 06:53 | |
| *** ricolin has quit IRC | 06:53 | |
| *** maciejjozefczyk has joined #openstack-nova | 06:55 | |
| *** ttsiouts has joined #openstack-nova | 06:58 | |
| openstackgerrit | Yongli He proposed openstack/nova master: Add server sub-resource topology API https://review.opendev.org/621476 | 07:02 |
| *** cdent has joined #openstack-nova | 07:04 | |
| *** luksky has joined #openstack-nova | 07:06 | |
| *** rpittau|afk is now known as rpittau | 07:06 | |
| *** xek has joined #openstack-nova | 07:07 | |
| *** tssurya has joined #openstack-nova | 07:10 | |
| *** damien_r has joined #openstack-nova | 07:12 | |
| *** tesseract has joined #openstack-nova | 07:21 | |
| openstackgerrit | Balazs Gibizer proposed openstack/nova stable/stein: Stabilize unshelve notification sample tests https://review.opendev.org/668806 | 07:25 |
| *** ttsiouts has quit IRC | 07:26 | |
| *** ttsiouts has joined #openstack-nova | 07:27 | |
| *** toabctl has joined #openstack-nova | 07:30 | |
| *** ttsiouts has quit IRC | 07:31 | |
| *** bhagyashris has joined #openstack-nova | 07:32 | |
| openstackgerrit | Lee Yarwood proposed openstack/nova master: Get rid of args to RBDDriver.__init__() https://review.opendev.org/668564 | 07:34 |
| *** luksky11 has joined #openstack-nova | 07:43 | |
| *** luksky has quit IRC | 07:45 | |
| *** ttsiouts has joined #openstack-nova | 07:51 | |
| openstackgerrit | Lee Yarwood proposed openstack/nova master: libvirt: Remove unreachable native QEMU iSCSI initiator config code https://review.opendev.org/668750 | 07:52 |
| *** luksky11 has quit IRC | 07:55 | |
| *** ralonsoh has joined #openstack-nova | 07:59 | |
| *** ociuhandu has joined #openstack-nova | 08:05 | |
| *** ociuhandu has quit IRC | 08:05 | |
| *** ociuhandu has joined #openstack-nova | 08:06 | |
| *** trident has quit IRC | 08:08 | |
| *** luksky11 has joined #openstack-nova | 08:08 | |
| *** trident has joined #openstack-nova | 08:09 | |
| *** ratailor_ has joined #openstack-nova | 08:10 | |
| *** luksky11 has quit IRC | 08:11 | |
| *** ratailor has quit IRC | 08:12 | |
| *** ratailor__ has joined #openstack-nova | 08:13 | |
| *** ratailor_ has quit IRC | 08:16 | |
| *** ratailor_ has joined #openstack-nova | 08:16 | |
| *** sridharg has joined #openstack-nova | 08:18 | |
| *** ratailor__ has quit IRC | 08:19 | |
| *** priteau has joined #openstack-nova | 08:26 | |
| *** luksky11 has joined #openstack-nova | 08:27 | |
| *** luksky11 has quit IRC | 08:33 | |
| *** ivve has joined #openstack-nova | 08:34 | |
| openstackgerrit | Boxiang Zhu proposed openstack/nova master: Make evacuation respects anti-affinity rule https://review.opendev.org/649963 | 08:37 |
| *** tkajinam has quit IRC | 08:39 | |
| *** ricolin__ has joined #openstack-nova | 08:44 | |
| *** ricolin_ has quit IRC | 08:47 | |
| kashyap | johnthetubaguy: You about by any chance? If so, the Secure Boot spec requires the final +2 (& a +W), assuming you're satisfied with all the things I've addressed: https://review.opendev.org/#/c/506720/ | 08:47 |
| *** zzzeek has quit IRC | 08:48 | |
| *** zzzeek has joined #openstack-nova | 08:49 | |
| *** mdbooth has joined #openstack-nova | 08:49 | |
| *** davidsha has joined #openstack-nova | 08:49 | |
| *** ricolin__ is now known as ricolin | 08:55 | |
| *** tesseract-RH has joined #openstack-nova | 08:57 | |
| *** tesseract has quit IRC | 08:58 | |
| *** udesale has joined #openstack-nova | 09:04 | |
| *** ratailor_ has quit IRC | 09:05 | |
| *** ratailor has joined #openstack-nova | 09:07 | |
| *** luksky11 has joined #openstack-nova | 09:12 | |
| *** tesseract-RH has quit IRC | 09:29 | |
| *** tesseract has joined #openstack-nova | 09:29 | |
| *** tesseract has quit IRC | 09:33 | |
| *** tesseract has joined #openstack-nova | 09:33 | |
| *** tesseract has quit IRC | 09:38 | |
| *** tesseract has joined #openstack-nova | 09:39 | |
| *** guozijn has joined #openstack-nova | 09:50 | |
| kashyap | alex_xu: Hi, maybe you can give the final ACK (if you don't spot anything else): https://review.opendev.org/649963 | 09:57 |
| *** lei-zh has joined #openstack-nova | 09:59 | |
| *** lei-zh has quit IRC | 10:01 | |
| *** lei-zh has joined #openstack-nova | 10:02 | |
| *** bhagyashris has quit IRC | 10:07 | |
| *** guozijn has quit IRC | 10:14 | |
| *** ttsiouts has quit IRC | 10:16 | |
| *** ttsiouts has joined #openstack-nova | 10:16 | |
| *** ttsiouts has quit IRC | 10:21 | |
| *** udesale has quit IRC | 10:28 | |
| *** udesale has joined #openstack-nova | 10:29 | |
| *** brinzhang_ has quit IRC | 10:35 | |
| *** lei-zh has quit IRC | 10:36 | |
| *** udesale has quit IRC | 10:52 | |
| *** udesale has joined #openstack-nova | 10:54 | |
| *** udesale has quit IRC | 10:58 | |
| *** guozijn has joined #openstack-nova | 10:58 | |
| *** udesale has joined #openstack-nova | 10:58 | |
| *** cdent has quit IRC | 10:58 | |
| *** ttsiouts has joined #openstack-nova | 11:01 | |
| *** psachin has quit IRC | 11:06 | |
| *** guozijn has quit IRC | 11:10 | |
| *** sapd1_x has joined #openstack-nova | 11:14 | |
| *** udesale has quit IRC | 11:19 | |
| openstackgerrit | Josephine Seifert proposed openstack/nova-specs master: Spec for the Nova part of Image Encryption https://review.opendev.org/608696 | 11:23 |
| Luzi | sean-k-mooney, I added a few sentences concerning the lifecycle | 11:26 |
| *** cdent has joined #openstack-nova | 11:28 | |
| *** sapd1_x has quit IRC | 11:30 | |
| *** bbowen has joined #openstack-nova | 11:31 | |
| *** priteau has quit IRC | 11:32 | |
| *** tssurya has quit IRC | 11:34 | |
| *** tssurya has joined #openstack-nova | 11:37 | |
| *** ratailor has quit IRC | 11:42 | |
| openstackgerrit | Artom Lifshitz proposed openstack/nova master: NUMA live migration support https://review.opendev.org/634606 | 11:56 |
| openstackgerrit | Artom Lifshitz proposed openstack/nova master: CONF.enable_numa_live_migration is not needed >= Train https://review.opendev.org/640021 | 11:56 |
| *** shilpasd has joined #openstack-nova | 11:59 | |
| *** tbachman has quit IRC | 12:05 | |
| *** cdent has quit IRC | 12:08 | |
| *** francoisp has joined #openstack-nova | 12:10 | |
| *** elod has quit IRC | 12:11 | |
| *** priteau has joined #openstack-nova | 12:18 | |
| *** tbachman has joined #openstack-nova | 12:28 | |
| *** cdent has joined #openstack-nova | 12:29 | |
| *** _erlon_ has joined #openstack-nova | 12:33 | |
| *** eharney has joined #openstack-nova | 12:37 | |
| sean-k-mooney | Luzi: thanks ill re review but overall i liked the spec | 12:37 |
| sean-k-mooney | Luzi: so all server operations like shelve,rescure, mirgrations will all work | 12:41 |
| Luzi | yes | 12:42 |
| sean-k-mooney | and for shelve if the instance is create form an encypted image teh snapshot that is created automaticly by shelve will be encrypted | 12:43 |
| *** psachin has joined #openstack-nova | 12:43 | |
| Luzi | I will have a deeper look into this, but we only wanted to create encrypted images when a user triggers the encryption | 12:46 |
| sean-k-mooney | without a change to the shelve api the user would not be able to say use encyption since for shvele and cross cell resize a snapshot is created automatically | 12:47 |
| *** elod has joined #openstack-nova | 12:47 | |
| sean-k-mooney | rather then the user calling the snapshot api directly | 12:48 |
| sean-k-mooney | so we would really want the encyption to be sticky. e.g. if the instacne was created form an encypeted image defult to encyption | 12:48 |
| Luzi | well it seems there are three options here: | 12:52 |
| Luzi | 1. leaving this alone, so there would not be any encryption during shelve | 12:53 |
| Luzi | 2. automaticly encryption - which is a huge problem for glance, btw... | 12:53 |
| Luzi | or 3. let the user trigger the encryption (and maybe deny the operation, when the original image was encrypted) | 12:54 |
| Luzi | the last one would need a metadata change, so we can check, whether the snapshot could be taken without encryption | 12:55 |
| Luzi | but it seems to be the most reasonable way | 12:55 |
| sean-k-mooney | well if we do 1 if i wanted to steal you sensitive data all i have to do is shelve you instnace grap a copy of the snapshot which is in the clear and unshelve it | 12:56 |
| Luzi | there was a misunderstanding here with the word snapshot, so thank you for the hint | 12:57 |
| Luzi | yeah, thats not nice from a security perspective :D | 12:57 |
| sean-k-mooney | 2 and 3 are the most secure | 12:57 |
| sean-k-mooney | the issue with 3 woudl be that its would be fine for shelve to be exteded with a "encypte this please" option | 12:58 |
| sean-k-mooney | resize im less sure about | 12:58 |
| *** elod_ has joined #openstack-nova | 12:58 | |
| Luzi | 2 would not be practically as automatically creating secrets for a ressource of another project is horrible | 12:58 |
| *** elod has quit IRC | 12:58 | |
| sean-k-mooney | for resize we only create a snapshot if its a cross cell resize | 12:58 |
| *** elod_ is now known as elod | 12:59 | |
| sean-k-mooney | a normal resize wont create a snapshot | 12:59 |
| Luzi | so i will add number three in the spec and will have a look into resize too | 12:59 |
| sean-k-mooney | and as an unprivaldaged user i cant tell if its cross cell or not | 12:59 |
| Luzi | thank you, that was reasonable the best review I got so far :D | 13:00 |
| sean-k-mooney | :) well hopefully we can get this landed soon. as i said the rest of the spec looks good to me | 13:01 |
| sean-k-mooney | just to undersdand the problem with option 2, what is the issue with createing keys automatically | 13:02 |
| sean-k-mooney | i had assumed we would use the same key we are currently using | 13:02 |
| *** udesale has joined #openstack-nova | 13:03 | |
| sean-k-mooney | rather then create a new one but i also have not considered it much | 13:03 |
| Luzi | who will delete a newly created key bound to an image? | 13:03 |
| Luzi | well if its not exactly the same image anymore, would you encrypt it with the same key? | 13:04 |
| sean-k-mooney | well for shelve we delete the image we created when unshelving | 13:04 |
| Luzi | so there would be a point for nova to create and delete a key, that would be okay | 13:05 |
| sean-k-mooney | and for cross cell resize we auto delete teh image snapthot when we create the instnace on the dest node | 13:05 |
| sean-k-mooney | so nova | 13:05 |
| sean-k-mooney | yes in both cases nova could create a temperoy key and it could delete it at the correct time | 13:06 |
| Luzi | i know that discussion from cinder and glance, there cinder creates a key which glance has to delete - that is not a good practise | 13:06 |
| *** elod has quit IRC | 13:06 | |
| sean-k-mooney | right that is messy | 13:06 |
| *** elod has joined #openstack-nova | 13:07 | |
| Luzi | sean-k-mooney, that seems good from my perspective, so automatic key creation for shelve and cross-cellresize, given a metadata parameter indicating a formerly encrypted image | 13:07 |
| sean-k-mooney | yep and if that metadata paramter is not set just do what we do today | 13:08 |
| Luzi | yes | 13:08 |
| sean-k-mooney | so you can make it complete transparent to the end user and no api changes needed | 13:08 |
| Luzi | that sounds good, I will add this to the spec :) | 13:09 |
| *** nicolasbock has joined #openstack-nova | 13:11 | |
| *** owalsh has quit IRC | 13:13 | |
| *** owalsh has joined #openstack-nova | 13:14 | |
| *** owalsh has quit IRC | 13:15 | |
| *** priteau has quit IRC | 13:15 | |
| *** owalsh has joined #openstack-nova | 13:15 | |
| lyarwood | mdbooth: https://review.opendev.org/#/c/668750/ - thoughts on this? Came across it last night while looking at upstream bugs and suddenly noticed this codepath is dead. We could just keep it but it hasn't been tested in ~3 years so it might be easier to start over with a clean volume driver. | 13:16 |
| *** priteau has joined #openstack-nova | 13:18 | |
| mdbooth | lyarwood: Hmm. It doesn't sound like we're going to implement native iscsi any time soon, tbh, due to lacking multipath and no upstream motivation to implement it. | 13:18 |
| mdbooth | Making this just dead code which isn't like to be used any time soon. | 13:18 |
| mdbooth | And which probably doesn't work anyway. | 13:18 |
| mdbooth | lyarwood: Is os-brick based iscsi problematic? | 13:19 |
| *** owalsh_ has joined #openstack-nova | 13:20 | |
| lyarwood | mdbooth: it has some rough edges wrt to multipath etc but the basic single path stuff also covered by this native support is QEMU is pretty solid | 13:20 |
| *** BjoernT has joined #openstack-nova | 13:20 | |
| lyarwood | s/is/in/g | 13:20 |
| lyarwood | so yeah I'm fine nuking this for now and if anyone cares enough we can reintroduce it later in a broken out volume driver | 13:21 |
| lyarwood | and have it actually tested in the gate etc | 13:21 |
| mdbooth | lyarwood: +1 from me for nuking unusable code | 13:21 |
| lyarwood | ack thanks | 13:22 |
| *** owalsh has quit IRC | 13:23 | |
| *** owalsh_ is now known as owalsh | 13:23 | |
| *** brinzhang has joined #openstack-nova | 13:24 | |
| shilpasd | efried: Hi, need discussion on https://review.opendev.org/#/c/667952/1/nova/conf/scheduler.py@188 | 13:25 |
| efried | Sure shilpasd | 13:26 |
| efried | I'm sure we must have discussed this during the spec review | 13:26 |
| shilpasd | here saying 'enable_forbidden_aggregates_filter' conf option bydefault enabled | 13:26 |
| efried | It must be about the performance penalty. | 13:26 |
| shilpasd | efried: but what we observed for other request filters, its not so | 13:26 |
| efried | Well, this one is doing a pretty expensive db lookup every time. The other filters are (I think) exclusively local processing. | 13:27 |
| *** ttsiouts has quit IRC | 13:27 | |
| efried | Have you done any kind of performance analysis of this filter? | 13:27 |
| efried | especially on a large database with lots of aggregates | 13:27 |
| shilpasd | efried: not on that large scale | 13:28 |
| efried | dansmith: you were also concerned about the performance of this filter. | 13:29 |
| efried | Having thought about it some, we're better off being conservative unless we can demonstrate that the db lookup is super cheap. | 13:30 |
| efried | So - leave the conf option as is shilpasd :) | 13:30 |
| efried | I'll update the review. | 13:30 |
| sean-k-mooney | efried: i think all the prefilters default to false so for consitency i would keep it as false too | 13:31 |
| *** avolkov has quit IRC | 13:31 | |
| shilpasd | efried: great, thank you, we are keeping it as False | 13:31 |
| sean-k-mooney | at least until we have info from deploymetns that its cheap and works well | 13:31 |
| efried | sean-k-mooney: In this case I was suggesting that we remove the toggle entirely. You have to do enough other stuff to opt in, the only benefit you get by having it switched off is that you don't run the code. | 13:31 |
| shilpasd | sean-k-mooney: thank you for opinion | 13:32 |
| sean-k-mooney | ah right ok | 13:32 |
| shilpasd | efried: thanks for discussion | 13:33 |
| efried | yw | 13:33 |
| *** lpetrut has joined #openstack-nova | 13:34 | |
| *** francoisp has quit IRC | 13:39 | |
| *** _alastor_ has joined #openstack-nova | 13:42 | |
| gibi | stephenfin, efried: I'm +2 on the extra_spec validation spec https://review.opendev.org/#/c/638734 . alex_xu and bauzas had comments earlier. do we want to wait for them to chime in? | 13:44 |
| stephenfin | gibi: alex_xu, sure, but I think bauzas is MIA now since it's summer and he's French | 13:44 |
| *** ttsiouts has joined #openstack-nova | 13:44 | |
| *** francoisp has joined #openstack-nova | 13:44 | |
| gibi | stephenfin: OK, cool with me | 13:44 |
| efried | probably want to get sean-k-mooney to re-ack as well | 13:46 |
| sean-k-mooney | which reivew? | 13:46 |
| efried | he has the medal for most comments in a single review on that one. | 13:46 |
| efried | sean-k-mooney: https://review.opendev.org/#/c/638734 | 13:46 |
| efried | extra spec validation | 13:46 |
| sean-k-mooney | ah ok | 13:46 |
| sean-k-mooney | am ya that is a medal i get a little too often | 13:47 |
| *** psachin has quit IRC | 13:47 | |
| sean-k-mooney | ill quickly re review it now | 13:47 |
| *** zzzeek has quit IRC | 13:49 | |
| efried | stephenfin: any merit in changing the topic of that spec and its patch (https://review.opendev.org/640733) to ...-extended to match the new bp? | 13:49 |
| efried | I could understand if you want to keep them grouped with the old ones, as currently. | 13:49 |
| efried | either way | 13:49 |
| stephenfin | I'm happy with either too | 13:49 |
| stephenfin | so I'll leave it but won't object if you want to change it :) | 13:50 |
| efried | I only glanced at the number of results https://review.opendev.org/#/q/topic:bp/flavor-extra-spec-image-property-validation+(status:open+OR+status:merged) -- no idea how related/unrelated they actually are. | 13:50 |
| efried | is your work really a continuation of that? | 13:50 |
| stephenfin | To be honest, not really now. It's the same idea but on a broader scale | 13:50 |
| *** zzzeek has joined #openstack-nova | 13:51 | |
| stephenfin | Perhaps I should have a totally different name. Open to suggestions | 13:51 |
| stephenfin | Given that I'm not doing image metadata validation now, a new name would definitely be a good idea | 13:51 |
| efried | Oh, let's not go through the paperwork of changing the blueprint and stuff. | 13:51 |
| efried | but yeah, that's a point | 13:51 |
| *** zzzeek has quit IRC | 13:53 | |
| *** zzzeek has joined #openstack-nova | 13:54 | |
| sean-k-mooney | cool your going with my stevador suggestion for cyborge extra spec valdaditon or other services. | 13:55 |
| sean-k-mooney | almsot done | 13:55 |
| *** zzzeek has quit IRC | 13:55 | |
| stephenfin | yeah, it was a good one | 13:55 |
| *** zzzeek has joined #openstack-nova | 13:56 | |
| efried | sean-k-mooney: I saw (but did not read) you discussing image encryption with Luzi. Where are you on that spec? | 13:57 |
| efried | I noticed she added a sentence on (no) impact on lifecycle operations; is that sufficient to address your concern from the previous PS? | 13:57 |
| efried | also lyarwood: are you +1 on that? Your words imply so, but there's no vote. | 13:58 |
| efried | talking about https://review.opendev.org/#/c/608696/ btw | 13:58 |
| sean-k-mooney | efried: nova should create a tempory key on shelve and encypte the snapshot if the instance image was encypted and delete the key when unshelving. same for cross cell resize | 13:58 |
| Luzi | efried, i have to reconsider shelve and cross-cell resize | 13:58 |
| lyarwood | efried: yeah sorry the comments were against an older PS so I couldn't vote | 13:59 |
| * lyarwood votes now | 13:59 | |
| efried | okay, cool, thanks all. I'll hold off and wait for a new PS | 13:59 |
| *** BjoernT_ has joined #openstack-nova | 13:59 | |
| sean-k-mooney | either that or we need to expsoe it on the shelve/resize api endpoint for the end user to decide | 13:59 |
| *** mriedem has joined #openstack-nova | 13:59 | |
| openstackgerrit | Dakshina Ilangovan proposed openstack/nova-specs master: Spec: Provider config YAML file https://review.opendev.org/612497 | 14:00 |
| efried | sean-k-mooney, Luzi: But that changes everything. | 14:00 |
| openstackgerrit | Matt Riedemann proposed openstack/nova stable/queens: Init HostState.failed_builds https://review.opendev.org/668911 | 14:01 |
| sean-k-mooney | efried: what changes everything? auto encypte or api endpoint or both | 14:01 |
| efried | Auto encrypt: As previously written, you didn't keep track of whether a server was originated from an encrypted image. Now you would have to. | 14:01 |
| efried | API endpoint: now you need a microversion (right?) | 14:01 |
| sean-k-mooney | yes | 14:02 |
| sean-k-mooney | too ^ we would need a microversion | 14:02 |
| *** BjoernT has quit IRC | 14:02 | |
| sean-k-mooney | not sure about keeping track of if it was encypted | 14:02 |
| sean-k-mooney | i think Luzi said there was a metadta key on the image/instance | 14:03 |
| efried | I thought the spec said we forget about that. | 14:03 |
| sean-k-mooney | i assume it would be in the embeded copy of hte image meta we store in the request spec | 14:03 |
| efried | "...the encryption-related | 14:04 |
| efried | additional metadata properties become irrelevant and thus may be discarded | 14:04 |
| efried | instead of being carried over into the server's metadata." | 14:04 |
| Luzi | efried, i thought so until sean-k-mooney asked me, we misinterpreted the word snapshot | 14:04 |
| efried | I'm on the fence about whether it's better to encrypt automatically or on-demand when doing things involving snapshot. | 14:06 |
| efried | but it seemed reasonable as written (on-demand) | 14:06 |
| efried | and simpler | 14:06 |
| sean-k-mooney | either would be ok. what i dont think is ok would be that shelve and cross cell resize alway result in unencypted snap shots with no way to encyrpte them | 14:06 |
| efried | (cross-cell resize creates a snapshot?) | 14:07 |
| sean-k-mooney | so if we do the microversion bump or auto encypt i think both are fine | 14:07 |
| sean-k-mooney | efried: yes | 14:07 |
| efried | okay | 14:07 |
| sean-k-mooney | efried: for corss cell resize we can rely on direct connectivity to the host | 14:07 |
| *** BjoernT_ is now known as BjoernT | 14:08 | |
| efried | if we auto-encrypt, then we should do the same for the already-considered snapshot case (createImage) | 14:08 |
| efried | i.e. it should be symmetrical. | 14:08 |
| Luzi | efried, and that cannot happen | 14:08 |
| efried | but that would require a key... | 14:08 |
| *** mrch_ has quit IRC | 14:09 | |
| Luzi | because the new key would be generated in nova and is then used in glance, which doesn't delete the key in the case a user deletes the image | 14:09 |
| openstackgerrit | Lee Yarwood proposed openstack/nova master: block_device: Optionally recreate attachments when refreshing connection_info https://review.opendev.org/579004 | 14:09 |
| efried | Luzi: I assume saving the *original* key with the server metadata is a nonstarter. | 14:10 |
| efried | like taping your car keys to the fender | 14:10 |
| Luzi | well the key could have been deleted at the point | 14:11 |
| sean-k-mooney | Luzi: how? | 14:11 |
| Luzi | because it was just tied to the original image | 14:11 |
| stephenfin | Newb question: when I'm told to delay string interpolation by doing "LOG.warning('foo %s', 'bar')" instead of "LOG.warning('foo %s' % 'bar')", what actually resolves that? | 14:11 |
| sean-k-mooney | dose the instance not have a key associated to it | 14:11 |
| stephenfin | Is it a feature of the logging library | 14:12 |
| Luzi | if the original image is deleted, the key can be deleted too | 14:12 |
| sean-k-mooney | fo use in its encyped ephmeral disk | 14:12 |
| stephenfin | or the translation framework? | 14:12 |
| Luzi | but that would only be the case if we use a backend which supports ephemeral storage encryption | 14:12 |
| sean-k-mooney | ah right | 14:13 |
| Luzi | ceph right now doesn't - maybe a nice thing to look for in the future :D | 14:13 |
| stephenfin | I ask because we appear to be delaying interpolation here but I didn't that was supported for exceptions https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L5842-L5847 | 14:13 |
| stephenfin | So I'm guessing it's a minor bug | 14:13 |
| sean-k-mooney | yes native encyption fo ceph rbds would be nice at somepoint | 14:14 |
| efried | stephenfin: yes that looks like a bug | 14:14 |
| stephenfin | okay, phew. I'll fix | 14:14 |
| *** ivve has quit IRC | 14:14 | |
| *** tbachman has quit IRC | 14:16 | |
| sean-k-mooney | Luzi: efried so if we cant make auto encyption symetric for normal snapshoting those that make it a non starter in your minds | 14:16 |
| *** ricolin has quit IRC | 14:16 | |
| sean-k-mooney | i would personlally prefer teh asymatry to give a better user experince and be secure by defualt | 14:17 |
| efried | stephenfin: http://paste.openstack.org/show/753821/ | 14:17 |
| efried | your exception message winds up being a %s-ification of a tuple of a string and a dict. | 14:17 |
| stephenfin | efried: I did the exact same thing :) However, I wasn't sure if the i18n marker changed things | 14:18 |
| efried | oh, definitely not, it returns a string | 14:18 |
| *** munimeha1 has joined #openstack-nova | 14:18 | |
| efried | The marker can't see the stuff after the comma. | 14:18 |
| stephenfin | Sweet, so it is logging that does the delayed interpolation. Good to know I hadn't been mistaken all these years, heh | 14:18 |
| efried | yes, logging does delayed, because it's conditional. The exception is always happening (if you get to that point), so there's no reason to delay anything. | 14:19 |
| *** jaosorior has quit IRC | 14:20 | |
| *** tbachman has joined #openstack-nova | 14:34 | |
| openstackgerrit | Kashyap Chamarthy proposed openstack/nova master: libvirt: Remove MIN_{QEMU,LIBVIRT}_LUKS_VERSION https://review.opendev.org/668924 | 14:35 |
| openstackgerrit | Dakshina Ilangovan proposed openstack/nova-specs master: Spec: Provider config YAML file https://review.opendev.org/612497 | 14:36 |
| openstackgerrit | Balazs Gibizer proposed openstack/nova master: Translatable output strings in heal allocation https://review.opendev.org/668925 | 14:39 |
| *** tbachman has quit IRC | 14:39 | |
| *** tbachman has joined #openstack-nova | 14:49 | |
| *** igordc has joined #openstack-nova | 14:50 | |
| *** bnemec has quit IRC | 14:50 | |
| adrianc | stephenfin, efried, thanks for taking the time to review | 14:51 |
| adrianc | [FUP] Follow-up patch for SR-IOV live migration | 14:51 |
| efried | yw adrianc | 14:51 |
| efried | hope the minor updates were acceptable | 14:51 |
| *** factor has quit IRC | 14:51 | |
| adrianc | sure thx for uploading the new PS | 14:52 |
| *** factor has joined #openstack-nova | 14:52 | |
| *** brinzhang has quit IRC | 14:55 | |
| *** Luzi has quit IRC | 14:58 | |
| *** icarusfactor has joined #openstack-nova | 15:00 | |
| *** brinzhang has joined #openstack-nova | 15:00 | |
| *** factor has quit IRC | 15:00 | |
| *** mriedem has left #openstack-nova | 15:05 | |
| *** mriedem has joined #openstack-nova | 15:05 | |
| cdent | efried: I've been waiting for you +1/2 on https://review.opendev.org/#/c/612497/ before reviewing it, assuming until it meets your expectations it wasn't worth the effort. So I'll probably get to it tomorrow morning... | 15:08 |
| efried | cdent: I'm working through it (again) now. | 15:08 |
| efried | your input would be valuable at this stage anyway, but I understand wanting to wait until the dust settles a bit. | 15:08 |
| cdent | i've also already crested my 8 hour limit aleady today (was up early to work with some folk in china), so unless I get an afflatus this evening, it'll be manana | 15:10 |
| *** tssurya has quit IRC | 15:11 | |
| *** cdent has quit IRC | 15:11 | |
| kashyap | johnthetubaguy: Ping^2, maybe now you're about :-) -- Want to get the Secure Boot spec across the line? -- https://review.openstack.org/#/c/506720/ ) | 15:12 |
| *** tbachman has quit IRC | 15:13 | |
| *** priteau has quit IRC | 15:14 | |
| efried | apparently cdent should stock bean-o | 15:14 |
| dansmith | mriedem: I think I could probably slam your experimental job add patch amirite? | 15:18 |
| *** hoonetorg has quit IRC | 15:20 | |
| mriedem | the ovs hybrid plug one? yeah. | 15:20 |
| *** brinzhang has quit IRC | 15:21 | |
| *** liuyulong has joined #openstack-nova | 15:22 | |
| dansmith | mriedem: yeah | 15:22 |
| dansmith | so slammed. | 15:23 |
| *** tbachman has joined #openstack-nova | 15:23 | |
| bauzas | efried: gibi: johnthetubaguy: any reason why https://review.opendev.org/#/c/638734/8 wasn't +W yet ? I'm also cool with it, so I wonder why we hold it | 15:28 |
| efried | bauzas: We're waiting for alex_xu to re-ack at this point. | 15:28 |
| bauzas | ack, thanks | 15:28 |
| *** priteau has joined #openstack-nova | 15:28 | |
| bauzas | we basically already agreed on it at the PTG | 15:28 |
| bauzas | so I'm not sure why we need to wait more but okay :) | 15:29 |
| *** bnemec has joined #openstack-nova | 15:29 | |
| bauzas | the key part of the validation would be more controversial | 15:29 |
| bauzas | but just checking the value looks just cool and simple | 15:29 |
| bauzas | efried: ^ | 15:29 |
| bauzas | I mean, plain simple | 15:29 |
| *** sridharg has quit IRC | 15:29 | |
| bauzas | from a spec perspective, we can nitpick on the implementation tho | 15:30 |
| *** liuyulong has quit IRC | 15:30 | |
| *** ttsiouts has quit IRC | 15:30 | |
| efried | I'll go ahead and merge it. | 15:30 |
| *** ttsiouts has joined #openstack-nova | 15:31 | |
| bauzas | cool | 15:31 |
| *** gyee has joined #openstack-nova | 15:31 | |
| efried | stephenfin: assuming you didn't want to change the bp name? | 15:31 |
| efried | ...which I just realized isn't in the commit message, ima fix that quick. | 15:31 |
| openstackgerrit | Eric Fried proposed openstack/nova-specs master: Add 'flavor-extra-spec-image-property-validation-extended' spec https://review.opendev.org/638734 | 15:32 |
| efried | f | 15:32 |
| efried | now ima fix it right | 15:32 |
| openstackgerrit | Eric Fried proposed openstack/nova-specs master: Add 'flavor-extra-spec-image-property-validation-extended' spec https://review.opendev.org/638734 | 15:33 |
| *** hoonetorg has joined #openstack-nova | 15:33 | |
| *** icarusfactor has quit IRC | 15:34 | |
| *** factor__ has joined #openstack-nova | 15:34 | |
| efried | stephenfin: I went ahead and made those topic changes https://review.opendev.org/#/q/topic:bp/flavor-extra-spec-image-property-validation-extended | 15:35 |
| *** ttsiouts has quit IRC | 15:36 | |
| stephenfin | efried++ thanks :) | 15:36 |
| stephenfin | That WIP probably needs to be abandoned since we've gone a totally different direction, but that's a job for later this week/next week | 15:37 |
| *** factor has joined #openstack-nova | 15:38 | |
| *** factor__ has quit IRC | 15:39 | |
| *** jangutter has quit IRC | 15:42 | |
| openstackgerrit | Eric Fried proposed openstack/python-novaclient master: Modify the url of upper_constraints_file https://review.opendev.org/665934 | 15:43 |
| *** udesale has quit IRC | 15:45 | |
| efried | mriedem: cross-cell resize got no love this runway slot? :( | 15:45 |
| efried | move back into queue, yah? | 15:46 |
| mriedem | efried: it got some love from gibi and dansmith but yes i have some comments to address from gibi and need to re-queue | 15:47 |
| efried | ight | 15:47 |
| mriedem | are you going to update the etherpad? | 15:47 |
| mriedem | nvm i'll do it | 15:51 |
| *** tbachman has quit IRC | 15:51 | |
| *** tbachman has joined #openstack-nova | 15:52 | |
| openstackgerrit | Merged openstack/nova-specs master: Add 'flavor-extra-spec-image-property-validation-extended' spec https://review.opendev.org/638734 | 15:55 |
| mriedem | gerrit is being restarted | 15:56 |
| efried | thanks | 15:57 |
| openstackgerrit | Balazs Gibizer proposed openstack/nova master: Mova consts from neutronv2/api to constants module https://review.opendev.org/668945 | 15:58 |
| openstackgerrit | Balazs Gibizer proposed openstack/nova master: Use neutron contants in cmd/manage.py https://review.opendev.org/668946 | 15:58 |
| openstackgerrit | Balazs Gibizer proposed openstack/nova master: Add 'resource_request' to neutronv2/constants https://review.opendev.org/668947 | 15:58 |
| *** cfriesen has joined #openstack-nova | 15:58 | |
| mriedem | gibi: fyi i'm going through the heal allocations change now | 15:59 |
| mriedem | and thanks for hitting the disabled compute series so fast | 15:59 |
| *** markguz_ has joined #openstack-nova | 15:59 | |
| gibi | mriedem: thanks. Above is the fup I promised to move to use contants instead of string for 'binding:profile' and the rest | 16:00 |
| gibi | mriedem: disabled compute seemed an easy hit, and also I was vocal in the spec so I felt responsible | 16:01 |
| gibi | mriedem: I need to drop soon so I will get back to your comments (if any) on the heal port allocation tomorrow | 16:04 |
| mriedem | yup np | 16:07 |
| *** hemna has quit IRC | 16:08 | |
| *** hemna has joined #openstack-nova | 16:09 | |
| openstackgerrit | Sylvain Bauza proposed openstack/nova-specs master: Proposes NUMA affinity for vGPUs https://review.opendev.org/650963 | 16:14 |
| *** rpittau is now known as rpittau|afk | 16:14 | |
| bauzas | dansmith, efried: thanks for reviewing ^, I just provided some new revision with more explanation about upgrades | 16:14 |
| bauzas | efried: I understand your concerns about this being a legacy mechanism and the fact that we approved the placement nested magic spec, so it could be seen as competitive | 16:15 |
| bauzas | efried: but what I'm trying to explain is that my spec isn't competing with NUMA be modeled in placement, as it just provides a soft affinity check that can still be done once NUMA is modeled with placement | 16:16 |
| bauzas | even the object change | 16:16 |
| bauzas | so we can keep it, and drop it when we want | 16:16 |
| openstackgerrit | Martin Midolesov proposed openstack/nova master: Implementing graceful shutdown. https://review.opendev.org/666245 | 16:25 |
| *** priteau has quit IRC | 16:28 | |
| *** qqmber has joined #openstack-nova | 16:30 | |
| *** damien_r has quit IRC | 16:30 | |
| *** munimeha1 has quit IRC | 16:31 | |
| qqmber | hi... since today when I attach a volume (cinder volume), horizon shows that it is attached to the instance (Centos 7). If I run "nova show ID", shows the new volume's ID, but nothing happens on the instance... if it supposed to be in /dev/vdd, I run "sudo dmesg | grep vdd" nothing is shown (inside the instance). Any ideas where should I look for the problem? | 16:32 |
| *** bnemec has quit IRC | 16:45 | |
| *** mvkr has quit IRC | 16:47 | |
| *** mvkr has joined #openstack-nova | 16:50 | |
| sean-k-mooney | qqmber: what does lsblk show | 16:59 |
| sean-k-mooney | qqmber: the name you specified is not garunteeded to be the one it gets | 16:59 |
| qqmber | in Nova host or the instance? | 16:59 |
| sean-k-mooney | in the instance | 16:59 |
| qqmber | ok | 16:59 |
| qqmber | sean-k-mooney: not vdd there.. only vda and vdb (vdc is another story) | 17:00 |
| sean-k-mooney | if the attach succeeded i woudl expect tehre to be a new block device that is not mounted but it might have a different name | 17:00 |
| qqmber | I just sent you what I got | 17:01 |
| *** davidsha has quit IRC | 17:02 | |
| *** markvoelker has joined #openstack-nova | 17:02 | |
| *** factor has quit IRC | 17:03 | |
| *** openstackgerrit has quit IRC | 17:04 | |
| *** elod has quit IRC | 17:05 | |
| *** elod has joined #openstack-nova | 17:07 | |
| *** bnemec has joined #openstack-nova | 17:14 | |
| *** lpetrut has quit IRC | 17:16 | |
| *** ociuhandu has quit IRC | 17:26 | |
| dansmith | efried: can you save me some reading and summarize the opinion on bauzas' gpu numa spec | 17:39 |
| dansmith | ? | 17:39 |
| *** markguz_ has quit IRC | 17:39 | |
| dansmith | like, are you saying that we're really far enough along to fully model (and reshape to) gpus under numa nodes with allocation candidates during scheduling to not do any of what he prescribes? | 17:40 |
| dansmith | storing the child providers in the numa object and a weigher to look at it seems like (a) it might be useful for other things and (b) a weigher isn't a huge amount of debt | 17:40 |
| dansmith | like, I can imagine the "show me the instance topology" thing might use some info in that child object to construct a view of where your assigned gpu (or something else) fits into the topop | 17:41 |
| *** tbachman has quit IRC | 17:42 | |
| *** igordc has quit IRC | 17:45 | |
| *** luksky11 has quit IRC | 17:46 | |
| efried | dansmith: We're talking about coding a weigher that would provide a subset of the capability, and be 100% obsolete when we do placement-based topo modeling & affinity. | 17:56 |
| efried | reading material, like the placement side specs & patches? | 17:56 |
| dansmith | efried: even if placement returns an AC with hierarchical allocations for the proper gpus, we'd still have weigher activity to decide which of those we want right? | 17:58 |
| *** tbachman has joined #openstack-nova | 17:58 | |
| efried | what would we be weighing in that case? | 17:58 |
| efried | like, what's a use case where "pick the first one" isn't right. | 17:58 |
| dansmith | efried: there is always a decision to be made about which allocation we want right? | 17:58 |
| sean-k-mooney | we are weighign host that can fufile numa affitny above those that cant | 17:59 |
| *** BjoernT has quit IRC | 17:59 | |
| dansmith | placement may return two ACs, one where the gpu is on a node with the minimum amount of memory, and one where there's a bunch extra, and a weigher would choose the latter I would think | 18:00 |
| dansmith | or the weigher would opt for the AC with the best combination of cpu, memory and gpu, especially where there are more than one gpu and you want as even of a distribution as possible | 18:00 |
| sean-k-mooney | the other usecase for the weigher would be to avoid placeing instace on a host with a gpu if they did not request one | 18:01 |
| efried | I thought bauzas's spec proposes that we binarily weigh "satisfies affinity requirements" higher than "doesn't" | 18:01 |
| sean-k-mooney | by weighering them lower | 18:01 |
| dansmith | sean-k-mooney: that's what bauzas' proposed weigher is doing yeah, but I'm saying that may evolve to something still useful after placement is returning ACs with gpu elements | 18:01 |
| sean-k-mooney | we do that in the pci weigher i think | 18:01 |
| sean-k-mooney | right | 18:01 |
| efried | I'm certainly not saying there's no value in a weigher | 18:01 |
| sean-k-mooney | when modeled in placmenet if we have two AC on the same host it coudl weigh the ACs based on some other parmater | 18:02 |
| efried | but weighing "satisfies affinity requirements" is a thing we can do very soon with placement, if we're willing to put the work in it. | 18:02 |
| dansmith | I'm just saying I expect that the bones of the proposed weigher would still be useful after placement is returning more data | 18:02 |
| dansmith | efried: and also, we've been on the verge of being able to do this for a while, so it's easy to never move forward on a sub-optimal approach in favor of something else that never comes, | 18:03 |
| dansmith | so if there's something re-usable about this after the better thing comes, I'm not as concerned about the "100% debt after merge" concern | 18:03 |
| efried | okay, so let's talk about how close we are to being able to move forward: | 18:03 |
| sean-k-mooney | i see tetsuro pushed up https://review.opendev.org/#/c/668376/ on monday but he has -w it so i assume it still have more work to do. | 18:05 |
| efried | For NUMA we need | 18:06 |
| efried | rg-to-rp mappings (implemented, microversion 1.34) https://docs.openstack.org/placement/latest/specs/train/approved/placement-resource-provider-request-group-mapping-in-allocation-candidates.html | 18:06 |
| efried | and pieces from https://docs.openstack.org/placement/latest/specs/train/approved/2005575-nested-magic-1.html as follows | 18:06 |
| efried | arbitrary group suffixes (implemented, microversion 1.33) | 18:06 |
| efried | same_subtree + resourceless request groups (https://review.opendev.org/#/c/668376/ -W as sean-k-mooney says, but pretty close) | 18:06 |
| efried | and then obviously we have to do the things in nova to 1) reshape and 2) request accordingly. | 18:06 |
| dansmith | yeah the reshape is what I'm concerned about.. that delaying us another cycle | 18:06 |
| efried | I feel it is very well understood what 1 and 2 will look like. | 18:07 |
| sean-k-mooney | 1 is https://review.opendev.org/#/q/topic:story/2006068+(status:open+OR+status:merged) right | 18:07 |
| dansmith | those are all placement, | 18:07 |
| dansmith | the reshape is in nova | 18:07 |
| sean-k-mooney | oh ya but the reshap cant happen until the placment stuff is done | 18:08 |
| dansmith | sure, just being clear | 18:08 |
| dansmith | the reshape is more complicated (IMHO, and IME), so unless that's "close" I'd be more concerned | 18:08 |
| sean-k-mooney | im not sure we fully underdand what the reshap will looklike | 18:08 |
| efried | yeah, those are a minor bugfix for rg-to-rp mappings. Really only cdent's is needed, the rest are what-ifs. | 18:09 |
| efried | The nested magic spec has several examples of what the topo will look like | 18:09 |
| dansmith | I think knowing what it will look like is not so much the problem | 18:09 |
| efried | deciding which resources need to go where? | 18:10 |
| sean-k-mooney | yes | 18:10 |
| dansmith | the actual mechanics of doing the reshape.. I know it's doable and we know what the end result will look like | 18:11 |
| sean-k-mooney | for exampel i think cpus are going to need to be moded in a nested resouce provider of the numa node not as an inventory of the numa node | 18:11 |
| sean-k-mooney | hugepages should proably be on the numa node but as a new resouce class per pagesize or as somthign else? | 18:12 |
| dansmith | efried: case in point ^ :) | 18:12 |
| sean-k-mooney | dose memroy_mb stay on the compute resouce proveiider? | 18:12 |
| dansmith | still plenty of discussion to derail the reshape | 18:12 |
| dansmith | efried: anyway, I think you oughta make a clear decision on this point, and -2 bauzas' thing if so, so that the messaging is clear what to expect and focus on | 18:15 |
| dansmith | I'm not saying that's a terrible thing, I just think it's not terrible to continue with both | 18:16 |
| dansmith | obviously I want the pure-placement approach for all the right reasons | 18:16 |
| efried | "not terrible to continue with both" is why I haven't downvoted | 18:17 |
| dansmith | okay | 18:17 |
| efried | but "want the pure-placement approach" is why I haven't upvoted. | 18:17 |
| efried | Who's going to work the code for bauzas's spec. | 18:17 |
| efried | ? | 18:17 |
| sean-k-mooney | bauzas: | 18:17 |
| dansmith | um, bauzas I hope :D | 18:17 |
| sean-k-mooney | assumign its approved if not he will bw working on other things | 18:18 |
| dansmith | mriedem: any thoughts on this? if you and efried are both -0.9 then it's probably not worth expecting it to happen | 18:18 |
| efried | If that person were dedicated instead to working on the topo reshape, that would get us closer | 18:18 |
| efried | faster | 18:18 |
| efried | even if it doesn't all come together in train. | 18:18 |
| efried | whereas I'm not sure what we gain by having the separate weigher. | 18:19 |
| sean-k-mooney | we get numa support for gpus in trian which will be the basis of our next longer support release donwstream | 18:19 |
| dansmith | the weigher gets us a lot closer to the correct behavior with a minimal amount of code and pretty predictable amount of work | 18:19 |
| dansmith | it doesn't get us to perfection, obviously | 18:19 |
| efried | closer to the correct behavior... for just vgpu, right? | 18:20 |
| dansmith | well, the bulk of this is reusable for other child devices, | 18:21 |
| sean-k-mooney | we already do the right thing for pci pasthough | 18:21 |
| sean-k-mooney | and for hugepages and cpu pinning | 18:21 |
| dansmith | although I expect the weigher would probably be specific to gpus | 18:21 |
| efried | other child devices - that we don't model in placement yet | 18:21 |
| efried | so n/a? | 18:21 |
| efried | and please let's not start modeling devices in non-NUMA-aware placement and adding weigher code to do the NUMA stuff for that. | 18:22 |
| dansmith | efried: other things we hang off of the compute RP that aren't specific in placement, but yeah, nothing pending, just saying this isn't adding a bunch of stuff to do gpu-specific things | 18:22 |
| sean-k-mooney | we dont track them in plamcneet and can provide better placmente and affintiy because of that | 18:22 |
| efried | until we track them *properly* in placement, right. | 18:22 |
| *** BjoernT has joined #openstack-nova | 18:22 | |
| sean-k-mooney | which has been promsied for 2 years at this point | 18:23 |
| efried | good, so let's get it the f done instead of keep kicking it down the road | 18:23 |
| dansmith | efried: just FYI I hate myself for even arguing about this, because pure-placement is obviously the right way, I'm just hedging because of the pain of things not actually being in train | 18:23 |
| dansmith | because seriously... years. | 18:23 |
| efried | I've said before (or maybe just thought it) vgpu may have been nice as a foot in the door proving device modeling in placement, but premature. And now we're trying to retrofit to make it work instead of shelving it until we can make the infrastructure support it properly. | 18:24 |
| efried | but | 18:25 |
| efried | I still don't feel strongly enough to block | 18:25 |
| sean-k-mooney | efried: for context the CAT stuff im working on here https://review.opendev.org/#/c/662264/ which you would also like me to track only with numa awer placment was requested for pike | 18:25 |
| dansmith | if we didn't have basic gpu stuff that we have today at this point, I think we'd probably be having a much different conversation.. and not a fun one. | 18:25 |
| *** amodi has quit IRC | 18:25 | |
| efried | Tell ya, though, if sam_subtree lands pretty soon, as appears likely to do, that may be enough to make me -1 the vgpu affinity spec. | 18:26 |
| dansmith | I thought you were the "don't wait forever for perfect" guy? :) | 18:26 |
| efried | Totally. And I have no illusions that numa-in-placement will be perfect right away | 18:26 |
| dansmith | well, if you think that's going to happen, you should just preemtively -1 it, or | 18:27 |
| efried | which btw is why we should not be arguing about hugepages and splitting CPUs finer than per-numa-node. | 18:27 |
| dansmith | or -1 it contingent on that merging when you think it will | 18:27 |
| dansmith | because really, messaging the likelihood of this being a thing is pretty important to us | 18:27 |
| *** _erlon_ has quit IRC | 18:28 | |
| efried | and why we punted on can_split | 18:28 |
| efried | "messaging the likelihood of this being a thing" is actually the best argument I've heard so far. | 18:28 |
| dansmith | it's the most important I think | 18:29 |
| sean-k-mooney | well if we cant have memory fine then a numa node we can implemenat CAT | 18:29 |
| sean-k-mooney | the cpus inventories and cache inventores shoudl be in the same RP | 18:30 |
| sean-k-mooney | and cache regins are finer then numa node | 18:30 |
| sean-k-mooney | we can have mulplie l3 cache regins in the same numa node | 18:30 |
| efried | understood. | 18:30 |
| efried | So | 18:30 |
| sean-k-mooney | it would basicaly be a 3 level tree | 18:31 |
| efried | when we implement CAT *after* having implemented NUMA topo as its own thing, we can reshape the CPUs and L3s down into grandchild providers. | 18:31 |
| sean-k-mooney | yes we could | 18:31 |
| efried | but we should not be trying to solve CAT preemptively. | 18:32 |
| sean-k-mooney | or we could do that first and reshap them under the numa node when they are added | 18:32 |
| *** shilpasd has quit IRC | 18:32 | |
| efried | I owe you a response in the pqos spec, but that doesn't buy us anything at this point. | 18:32 |
| sean-k-mooney | we talked about it a little today in our squad call | 18:32 |
| sean-k-mooney | im going to simplfy it and proably add a depency on some of the placmenet stuff | 18:33 |
| sean-k-mooney | i will likely make memory bandwith depend on the numa in placment spec | 18:33 |
| sean-k-mooney | i would like to keep cache not depenent on tha tbut may on same tree | 18:33 |
| sean-k-mooney | well it depends i have to see is there a way to us more of the plamcent stuff | 18:34 |
| sean-k-mooney | it significantly incresase the risk we wont have CAT in train however | 18:35 |
| *** artom|gmtplus3 has quit IRC | 18:35 | |
| *** ralonsoh has quit IRC | 18:35 | |
| efried | I'm addressing similar concerns in the providers.yaml spec | 18:35 |
| sean-k-mooney | ill admit i have nto really been following that one. i kindof got burt out on that topic last cycle | 18:36 |
| sean-k-mooney | is it approching agreement or still in flux | 18:37 |
| efried | Dakshina has me convinced that there's a reason to do cache in child providers. But only to ensure that one request group gets all its cacheways from one package -- *not* because we can do anything about affinity. | 18:37 |
| efried | however | 18:37 |
| efried | in tree | 18:37 |
| efried | we can do affinity after GET /a_c via the NUMATopologyFilter | 18:37 |
| efried | was gonna say that couples the filter to libvirt, but that's already dismally true anyway. | 18:39 |
| efried | sean-k-mooney: I haven't posted those comments yet, but will ping you once I have. | 18:40 |
| sean-k-mooney | ok. im goign to leave the rebasing it so that we can continue the disucssion in one place | 18:41 |
| sean-k-mooney | are you off for the 4th of july weekend? | 18:41 |
| sean-k-mooney | i assume yes | 18:41 |
| efried | and dansmith, I'll review tetsuro's same_subtree patch. I have the broad strokes from an earlier iteration, but it's now been squashed and reworked a little bit, so I need to refresh and see how far off it still is. And if it's as close as I think, I'll take a stand on the vgpu affinity spec. | 18:41 |
| dansmith | thanks | 18:42 |
| efried | sean-k-mooney: yes, I'll be off Th & F | 18:42 |
| efried | Friday is my 22nd anniversary :) | 18:42 |
| *** luksky11 has joined #openstack-nova | 18:42 | |
| dustinc | efired: congrats! | 18:43 |
| sean-k-mooney | 22nd, surly not wedding? unless you were married quite young | 18:43 |
| dustinc | efried.. | 18:43 |
| sean-k-mooney | as in before its legal. | 18:44 |
| efried | hehe. I was married at 20 | 18:44 |
| *** tesseract has quit IRC | 18:44 | |
| efried | so it was legal to marry, just not legal to buy alcohol :P | 18:44 |
| efried | thanks dustinc | 18:44 |
| sean-k-mooney | well congrats and i always get confused about how the us legislates for mimium ages for things | 18:45 |
| efried | you're not the only one | 18:45 |
| dustinc | can active duty still buy beer/wine at 18 on base? | 18:46 |
| dustinc | gets even more confusing | 18:46 |
| efried | 21 for alcohol is country-wide at this point. Louisiana was (I think) the last one to capitulate in the late '90s (from 18) when the fed cut off their road funding. | 18:46 |
| sean-k-mooney | you can drive from 16 and get a gun licence at 16 or 18 right and marry at like 16-18 depending on the state | 18:47 |
| efried | sean-k-mooney: I grew up near London, where the official age was 18 but not enforced at all. I was pubbing from 14. | 18:47 |
| sean-k-mooney | and vote at 18 too right | 18:47 |
| dustinc | the problem with voting here is you can vote at 18 but nobody does until 60 sigh... | 18:48 |
| efried | vote at 18, yes, that was a big deal ~20y ago. Driving depends on the state, but it's 16 in most places. You can get an exception for "hardship" or farming (which are the same thing to me :P ) | 18:48 |
| efried | that's not a problem | 18:48 |
| efried | the problem with voting is that you are allowed to do so despite being completely ignorant | 18:48 |
| dustinc | that doesn't change with age | 18:48 |
| efried | swhat I'm sayin. | 18:48 |
| dustinc | gotcha :) | 18:49 |
| dustinc | voting is really complicated in some states too | 18:49 |
| sean-k-mooney | ya you can get a work vechles lices that allows you do drive a tracker or combin haveser at 16 in ireland but need to be 18 to get a car license which is ... odd | 18:49 |
| dustinc | my state votes by mail, others you have to stand in line for hours | 18:49 |
| dustinc | do you have graduated licensing for motorcycles? | 18:50 |
| sean-k-mooney | yes | 18:50 |
| dustinc | we get 18 year olds buying a 1000cc for their first bike here | 18:50 |
| dustinc | so annoying | 18:50 |
| sean-k-mooney | http://www.rsa.ie/Documents/Learner%20Drivers/Third%20Directive/bikes_chart_3rd_directive_V2.pdf | 18:51 |
| efried | In the UK didn't they pass a thing where you have to have a decal of some sort for the first year after you get your license, so other motorists are warned that you still suck? | 18:51 |
| sean-k-mooney | http://www.rsa.ie/en/RSA/Learner-Drivers/Motorcyclists/Licence-categories-explained/ | 18:51 |
| sean-k-mooney | efried: thats in ireland | 18:52 |
| sean-k-mooney | you have N placnete for novice | 18:52 |
| sean-k-mooney | *plates | 18:52 |
| sean-k-mooney | i like the metric system but i really dont understadn inutitivly how much 11KW is vs horsepower | 18:53 |
| dustinc | about 14hp according to google..dunno how you can even get around with so little | 18:54 |
| dustinc | a common starter bike here is 250cc twin and they make around 35hp I think | 18:55 |
| dustinc | and are sooo slow | 18:55 |
| sean-k-mooney | what does a 125cc engin normally put out | 18:55 |
| dustinc | not sure, guessing 11kw/14hp is probably pretty typical | 18:55 |
| dustinc | we just don't have many street-legal 125cc here | 18:55 |
| sean-k-mooney | ya that is the smalest license you get when your 16 | 18:56 |
| sean-k-mooney | so its mainly scotters and dirt bikes | 18:56 |
| sean-k-mooney | that said i was not expecting thise to qualify https://www.bennetts.co.uk/bikesocial/news-and-views/advice/new-riders-and-training/top-10-a1-licence-friendly-bikes-2018 | 18:57 |
| dustinc | it should be that way here, maybe start at 250cc since 125cc is hard to find, but people starting out on 110hp 600cc sport bikes (or worse...180-200hp liter bikes) is common and causes lots of accidents..and then expensive insurance and stupid laws | 18:57 |
| sean-k-mooney | i guess you really dont need the same level of raw horse power vs a car. anyway i better finish up my email and sign off for the evening o/ | 18:59 |
| dustinc | you guys get some cool little bikes there... | 19:00 |
| sean-k-mooney | unfortunetly i dont have a bike license so cant ride any of them | 19:00 |
| sean-k-mooney | i have though about getting it but mainly since i never drive my car and have been thinink of gettin rid of it but still should have somthing | 19:01 |
| qqmber | sean-k-mooney: I just ceated a new instance and in this one worked (attaching a new volume)... Crap... maybe the instance is screwed :( | 19:02 |
| sean-k-mooney | qqmber: it might be related to your iscsi backend | 19:03 |
| sean-k-mooney | do you have issue conencting to it form other vms on the same host | 19:03 |
| qqmber | that's what I'll try now | 19:03 |
| sean-k-mooney | you could also try moving the vm to anohter host and see if that fixes it | 19:04 |
| *** maciejjozefczyk has quit IRC | 19:04 | |
| sean-k-mooney | but it could be that its not propely exported | 19:04 |
| mriedem | dansmith: efried: i was out getting my quarterly sheering, do i need to weigh in on this numa affinity spec thing? b/c i don't really want to. | 19:04 |
| mriedem | *shearing | 19:04 |
| mriedem | sorry | 19:04 |
| efried | mriedem: Apparently we all have to take a stand one way or another, or dansmith won't be our friend anymore. | 19:05 |
| dansmith | mriedem: if you want, but I think I nudged efried in enough of a direction that it won't matter | 19:05 |
| efried | wait | 19:05 |
| *** igordc has joined #openstack-nova | 19:05 | |
| mriedem | i haven't reviewed the latest updates on sylvain's spec | 19:06 |
| dansmith | efried: not all of us, just you (have to take a stand) | 19:06 |
| mriedem | nor do i really have much skin in this game | 19:06 |
| mriedem | does france vacation season end before or after FF.... | 19:08 |
| dansmith | does it end...at all? | 19:09 |
| mriedem | touche | 19:10 |
| mriedem | fyi for anyone following along, there is 1 queens change to merge https://review.opendev.org/#/q/status:open+project:openstack/nova+branch:stable/queens+label:Code-Review=2 and then i'll be doing a release | 19:11 |
| *** whoami-rajat has quit IRC | 19:12 | |
| sean-k-mooney | i was expecting that to auto merge whne the parent did | 19:12 |
| qqmber | sean-k-mooney: one (maybe) important point... the instance with the (apparent) problem was migrated from one host to another (I shutdown the instance first) | 19:12 |
| mriedem | sean-k-mooney: it would have if not for gate failures | 19:12 |
| sean-k-mooney | mriedem: ya i just clicked in an notice the job that failed faild before the test ran | 19:13 |
| sean-k-mooney | qqmber: did you confimr the migration | 19:13 |
| sean-k-mooney | if not the instace might still be pointing at the source node. | 19:14 |
| * sean-k-mooney goes to check | 19:14 | |
| qqmber | yes, I did | 19:16 |
| qqmber | is there any way to confirm that? (I'm 99.9999999% sure I did) | 19:17 |
| sean-k-mooney | ok i was just wondering if the instace.host could still be set to the old host | 19:17 |
| sean-k-mooney | which would likely mess up attaching new volumes | 19:17 |
| qqmber | this instance had 2 volumes before the migration... after the migration one volume "died" (maybe because of the current problem), and now I'm not able to attach new ones... | 19:18 |
| sean-k-mooney | is the dead volume form the same cinder backend as the one that does not show up | 19:19 |
| *** markvoelker has quit IRC | 19:19 | |
| *** markvoelker has joined #openstack-nova | 19:19 | |
| sean-k-mooney | dansmith: by the way for atroms revert resize change you want to avoid the api change by doing the db lookup as mriedem and i were discussing and then have the followup patch on master change to passing the migration object down and remvoe the db lookup right | 19:21 |
| dansmith | sean-k-mooney: I | 19:21 |
| sean-k-mooney | cool | 19:21 |
| dansmith | sean-k-mooney: I'm fine with doing the one thing and backporting it, but mriedem is not, so then yes, those two steps | 19:21 |
| sean-k-mooney | i kind of prefered that too but was not sure how you would feel about the extra db query | 19:22 |
| *** cfriesen has quit IRC | 19:22 | |
| dansmith | sean-k-mooney: IMHO, backporting the interface change is not a problem, so we could avoid the lookup and the two-step dance | 19:22 |
| dansmith | but mriedem wants the interface to be stable, which is also fine | 19:22 |
| dansmith | what I'm not okay with is the duck typing on our internal interface | 19:23 |
| mriedem | we'll have a two step either way (to remove the TypeError handling) | 19:23 |
| sean-k-mooney | yep | 19:23 |
| dansmith | mriedem: I'm not okay with the typeerror either.. like, at all. | 19:23 |
| *** mrch_ has joined #openstack-nova | 19:23 | |
| dansmith | mriedem: we don't need it though, because it's all an internal interface.. we could just pass the thing in and change all the drivers, and backport that | 19:23 |
| sean-k-mooney | well the type error handeling is for oot drivers | 19:23 |
| dansmith | sean-k-mooney: right, which I don't care about at all | 19:24 |
| sean-k-mooney | which i know we technicaly dont support | 19:24 |
| mriedem | i only really care here b/c of backports | 19:24 |
| mriedem | if it weren't being backported i'd just say change the interface | 19:24 |
| mriedem | i'm not losing sleep over this though | 19:24 |
| sean-k-mooney | ok im just trying to make sure we know what to change while both of ye are on vacation for the next two days | 19:24 |
| dansmith | right, but I don't get why the backport matters, since this is all inside nova-compute, except for the OOT drivers, which .. I don't care about | 19:25 |
| dansmith | but whatever | 19:25 |
| mriedem | oh i'm here friday | 19:25 |
| dansmith | me too | 19:25 |
| sean-k-mooney | oh ok | 19:25 |
| mriedem | i am, however, gone the entire week of the 15th | 19:25 |
| mriedem | with no laptop | 19:25 |
| dansmith | so wait until the 15th and then we merge? :D | 19:25 |
| mriedem | and the revert on the 16th? sure. | 19:25 |
| dansmith | geg | 19:25 |
| dansmith | er, heh | 19:26 |
| sean-k-mooney | i mean its only merged once 3 times the charm right | 19:26 |
| mriedem | i really look forward to having to eyeball the diff yes | 19:27 |
| *** eharney has quit IRC | 19:27 | |
| sean-k-mooney | so to sumerise do the two step dance with the db lookup to minimes the backport size | 19:28 |
| dansmith | yes | 19:29 |
| sean-k-mooney | cool | 19:29 |
| *** whoami-rajat has joined #openstack-nova | 19:29 | |
| sean-k-mooney | ok im catully going to drop for the evenying enjoy tommorow o/ | 19:30 |
| *** igordc has quit IRC | 19:31 | |
| *** openstackgerrit has joined #openstack-nova | 19:34 | |
| openstackgerrit | Merged openstack/python-novaclient master: Modify the url of upper_constraints_file https://review.opendev.org/665934 | 19:35 |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: Follow up for pre-filter-disabled-computes series https://review.opendev.org/668986 | 19:56 |
| efried | sean-k-mooney: I posted that review. The part that pertains to what we were discussing earlier (child providers for LLC) is here https://review.opendev.org/#/c/612497/4/specs/train/approved/provider-config-file.rst@109 | 19:56 |
| efried | sean-k-mooney, dansmith: That spec is probably close enough for y'all to start looking at now. https://review.opendev.org/#/c/612497/ | 19:57 |
| *** damien_r has joined #openstack-nova | 19:57 | |
| efried | There's still work to be done, as noted, but it's in a place where there are no longer *major* issues that I am seeing. So a good time for *you* to identify major issues :) | 19:57 |
| mriedem | this all looks dead https://review.opendev.org/#/q/topic:bp/add-emulated-virtual-tpm+(status:open+OR+status:merged) | 20:03 |
| *** eharney has joined #openstack-nova | 20:04 | |
| *** hemna has quit IRC | 20:27 | |
| *** hemna has joined #openstack-nova | 20:27 | |
| *** jaypipes has quit IRC | 20:34 | |
| *** trident has quit IRC | 20:43 | |
| *** trident has joined #openstack-nova | 20:45 | |
| *** xek has quit IRC | 20:47 | |
| *** damien_r has quit IRC | 20:55 | |
| *** _alastor_ has quit IRC | 20:59 | |
| efried | mriedem: as in, been in merge conflict for a while, or as in, you looked at the code and it doesn't seem like it matches the spec etc? | 21:08 |
| mriedem | former | 21:11 |
| mriedem | i feel my abandon finger itching | 21:11 |
| *** slaweq has quit IRC | 21:11 | |
| *** igordc has joined #openstack-nova | 21:14 | |
| *** pcaruana has quit IRC | 21:16 | |
| efried | mriedem: We reapproved that spec for this release (in April). Go abandon something *really* old. | 21:21 |
| mriedem | blarg | 21:24 |
| openstackgerrit | Eric Fried proposed openstack/nova master: Fix GET /servers/detail host_status performance regression https://review.opendev.org/666042 | 21:26 |
| *** slaweq has joined #openstack-nova | 21:27 | |
| *** slaweq has quit IRC | 21:32 | |
| *** slaweq has joined #openstack-nova | 21:38 | |
| openstackgerrit | Merged openstack/nova master: [FUP] Follow-up patch for SR-IOV live migration https://review.opendev.org/659101 | 21:41 |
| openstackgerrit | Merged openstack/nova stable/stein: Stabilize unshelve notification sample tests https://review.opendev.org/668806 | 21:42 |
| openstackgerrit | Merged openstack/nova master: Follow-up for I6a777b4b7a5729488f939df8c40e49bd40aec3dd https://review.opendev.org/664967 | 21:42 |
| *** slaweq has quit IRC | 21:43 | |
| *** takashin has joined #openstack-nova | 21:51 | |
| *** rcernin has quit IRC | 22:04 | |
| *** qqmber has quit IRC | 22:11 | |
| *** bbowen has quit IRC | 22:15 | |
| *** _alastor_ has joined #openstack-nova | 22:17 | |
| *** whoami-rajat has quit IRC | 22:42 | |
| *** tbachman_ has joined #openstack-nova | 22:43 | |
| *** tbachman has quit IRC | 22:43 | |
| *** tbachman_ is now known as tbachman | 22:43 | |
| efried | dansmith: I voted on the vgpu affinity spec | 22:45 |
| efried | gotta run, see y'all Monday o/ | 22:45 |
| *** efried is now known as efried_pto | 22:45 | |
| *** luksky11 has quit IRC | 22:48 | |
| *** tkajinam has joined #openstack-nova | 22:54 | |
| openstackgerrit | Merged openstack/nova master: Remove 'MultiattachSupportNotYetAvailable' exception https://review.opendev.org/651315 | 23:00 |
| openstackgerrit | Merged openstack/nova master: Rename CinderFixtureNewAttachFlow to CinderFixture https://review.opendev.org/668561 | 23:00 |
| *** rcernin has joined #openstack-nova | 23:09 | |
| *** owalsh_ has joined #openstack-nova | 23:10 | |
| *** owalsh has quit IRC | 23:10 | |
| *** slaweq has joined #openstack-nova | 23:14 | |
| *** slaweq has quit IRC | 23:19 | |
| *** gyee has quit IRC | 23:39 | |
| *** nicolasbock has quit IRC | 23:41 | |
| openstackgerrit | Merged openstack/nova master: Add neutron-tempest-iptables_hybrid job to experimental queue https://review.opendev.org/667154 | 23:45 |
| *** gyee has joined #openstack-nova | 23:54 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!