*** wolverineav has quit IRC | 00:02 | |
*** wolverineav has joined #openstack-nova | 00:03 | |
*** wolverineav has quit IRC | 00:07 | |
*** Swami has quit IRC | 00:08 | |
*** gyee has quit IRC | 00:09 | |
*** wolverineav has joined #openstack-nova | 00:20 | |
*** tbachman has joined #openstack-nova | 00:25 | |
*** wolverineav has quit IRC | 00:27 | |
*** ttsiouts has joined #openstack-nova | 00:38 | |
*** ttsiouts has quit IRC | 00:47 | |
*** ttsiouts has joined #openstack-nova | 00:48 | |
*** wolverineav has joined #openstack-nova | 00:49 | |
*** ttsiouts has quit IRC | 00:52 | |
*** wolverineav has quit IRC | 00:53 | |
*** lbragstad has joined #openstack-nova | 00:55 | |
*** ricolin has joined #openstack-nova | 01:02 | |
*** igordc has quit IRC | 01:06 | |
*** tpatil has joined #openstack-nova | 01:07 | |
*** tiendc has joined #openstack-nova | 01:20 | |
*** _alastor_ has joined #openstack-nova | 01:27 | |
jungleboyj | efried: Thanks. Is there time planned to discuss this at the ptg? | 01:31 |
---|---|---|
*** hongbin has joined #openstack-nova | 01:33 | |
openstackgerrit | Ivens Zambrano proposed openstack/nova-specs master: RMD Plugin: Energy Efficiency using CPU Core P-State control The power state of a core can be setup between a minimum and the maximum frequency on the cores as defined in the factory. Each core on a CPU can be defined individually to perform at specific f https://review.openstack.org/651024 | 01:34 |
*** nicolasbock has quit IRC | 01:39 | |
*** _alastor_ has quit IRC | 01:40 | |
*** ricolin_ has joined #openstack-nova | 01:50 | |
*** ricolin has quit IRC | 01:50 | |
*** tpatil has quit IRC | 02:17 | |
openstackgerrit | Boxiang Zhu proposed openstack/nova master: Cleanup migrate flags https://review.openstack.org/651056 | 02:17 |
*** tpatil has joined #openstack-nova | 02:25 | |
*** hongbin has quit IRC | 02:35 | |
openstackgerrit | zhufl proposed openstack/nova master: Remove conductor_api and cells_rpcapi from manager.py https://review.openstack.org/651059 | 02:35 |
*** whoami-rajat has joined #openstack-nova | 02:37 | |
*** hongbin has joined #openstack-nova | 02:38 | |
*** igordc has joined #openstack-nova | 02:40 | |
efried | jungleboyj: Nothing yet, swhat I was going to ask you about. We could do 11:15-lunch (with a break at 11:50 for team photo) if that works for you. Or after lunch, any time except 1400-1515. | 02:42 |
*** boxiang has quit IRC | 02:42 | |
*** boxiang has joined #openstack-nova | 02:42 | |
jungleboyj | I don't have anything scheduled yet so I am flexible. Let's see what the team has, if anything, and talk later in the week. | 02:44 |
jungleboyj | efried: on vacation this week so will have smcginnis follow up. | 02:45 |
efried | sure thing, thanks jungleboyj | 02:45 |
*** cfriesen has quit IRC | 02:46 | |
*** hongbin has quit IRC | 02:49 | |
*** hongbin has joined #openstack-nova | 02:51 | |
*** Sundar has quit IRC | 02:57 | |
*** tpatil has quit IRC | 02:57 | |
openstackgerrit | Merged openstack/nova master: Remove mox in unit/network/test_neutronv2.py (7) https://review.openstack.org/574974 | 03:22 |
openstackgerrit | melanie witt proposed openstack/nova master: Count instances from mappings and cores/ram from placement https://review.openstack.org/638073 | 03:38 |
openstackgerrit | melanie witt proposed openstack/nova master: Use instance mappings to count server group members https://review.openstack.org/638324 | 03:38 |
*** cfriesen has joined #openstack-nova | 03:40 | |
*** N3l1x has quit IRC | 03:42 | |
*** tpatil has joined #openstack-nova | 03:50 | |
*** lbragstad has quit IRC | 03:51 | |
openstackgerrit | Boxiang Zhu proposed openstack/nova master: Make evacuation respects anti-affinity rule https://review.openstack.org/649953 | 03:53 |
*** imacdonn_ has quit IRC | 04:02 | |
*** imacdonn_ has joined #openstack-nova | 04:03 | |
*** udesale has joined #openstack-nova | 04:04 | |
*** hongbin has quit IRC | 04:05 | |
*** bhagyashris has joined #openstack-nova | 04:13 | |
openstackgerrit | Merged openstack/nova master: Add placeholder migrations for Stein backports https://review.openstack.org/650964 | 04:29 |
*** markvoelker has quit IRC | 04:31 | |
openstackgerrit | Ghanshyam Mann proposed openstack/nova-specs master: Spec for API inconsistency cleanup https://review.openstack.org/603969 | 04:44 |
*** dpawlik has joined #openstack-nova | 04:45 | |
openstackgerrit | Ghanshyam Mann proposed openstack/nova-specs master: Spec for API inconsistency cleanup https://review.openstack.org/603969 | 04:46 |
*** zhongjun2 has quit IRC | 04:46 | |
*** wolverineav has joined #openstack-nova | 04:49 | |
*** dpawlik has quit IRC | 04:50 | |
*** wolverineav has quit IRC | 04:54 | |
*** tpatil has quit IRC | 04:58 | |
*** sidx64 has joined #openstack-nova | 05:08 | |
openstackgerrit | Tetsuro Nakamura proposed openstack/nova master: Add in_tree field to RequestGroup object https://review.openstack.org/649534 | 05:10 |
openstackgerrit | Tetsuro Nakamura proposed openstack/nova master: Add get_compute_nodes_by_host_or_node() https://review.openstack.org/650877 | 05:10 |
openstackgerrit | Tetsuro Nakamura proposed openstack/nova master: Pass target host to RequestGroup.in_tree https://review.openstack.org/650878 | 05:10 |
openstackgerrit | Tetsuro Nakamura proposed openstack/nova master: Query `in_tree` to placement https://review.openstack.org/649535 | 05:10 |
*** igordc has quit IRC | 05:11 | |
*** ratailor has joined #openstack-nova | 05:11 | |
eandersson | I still wish you could add your own filters and weights as plugins. | 05:28 |
eandersson | Instead of having to roll your own version of nova and drop them in the appropriate folders. | 05:29 |
*** tpatil has joined #openstack-nova | 05:32 | |
*** _alastor_ has joined #openstack-nova | 05:33 | |
openstackgerrit | Boxiang Zhu proposed openstack/nova master: Use the functional test test_parallel_evacuate_with_server_group https://review.openstack.org/649963 | 05:36 |
*** _alastor_ has quit IRC | 05:46 | |
*** sridharg has joined #openstack-nova | 05:50 | |
*** udesale has quit IRC | 05:52 | |
openstackgerrit | Merged openstack/nova master: Remove query_client from resource_tracker https://review.openstack.org/650616 | 05:53 |
*** ricolin_ has quit IRC | 05:58 | |
*** ileixe has quit IRC | 06:01 | |
*** ratailor has quit IRC | 06:01 | |
*** ileixe has joined #openstack-nova | 06:05 | |
*** ccamacho has joined #openstack-nova | 06:12 | |
*** ratailor has joined #openstack-nova | 06:13 | |
*** ricolin has joined #openstack-nova | 06:22 | |
*** boxiang has quit IRC | 06:27 | |
*** dpawlik has joined #openstack-nova | 06:27 | |
*** pcaruana has joined #openstack-nova | 06:30 | |
*** markvoelker has joined #openstack-nova | 06:32 | |
*** udesale has joined #openstack-nova | 06:33 | |
*** slaweq has joined #openstack-nova | 06:47 | |
*** cfriesen has quit IRC | 06:48 | |
*** ivve has joined #openstack-nova | 07:06 | |
*** markvoelker has quit IRC | 07:06 | |
*** phasespace has quit IRC | 07:08 | |
*** rpittau|afk is now known as rpittau | 07:12 | |
*** tesseract has joined #openstack-nova | 07:13 | |
*** awalende has joined #openstack-nova | 07:13 | |
*** sidx64 has quit IRC | 07:13 | |
*** jangutter has quit IRC | 07:15 | |
*** yankcrime has quit IRC | 07:15 | |
*** jangutter_ has joined #openstack-nova | 07:15 | |
*** sidx64 has joined #openstack-nova | 07:15 | |
*** tpatil has quit IRC | 07:16 | |
*** slaweq has quit IRC | 07:17 | |
kashyap | efried: When you're about, I'd suggest to _not_ approve (although it was "approved" for Stein) this spec right away: https://review.openstack.org/#/c/642030/ (Re-propose the spec to allow specifying a list of CPU models) | 07:17 |
*** slaweq has joined #openstack-nova | 07:17 | |
kashyap | The more I think of it, there are more questions than answers. And the author of the spec from Stein seems AWOL. | 07:18 |
*** tosky has joined #openstack-nova | 07:20 | |
*** tpatil has joined #openstack-nova | 07:21 | |
*** tssurya has joined #openstack-nova | 07:28 | |
openstackgerrit | Kashyap Chamarthy proposed openstack/nova-specs master: Add "CPU selection with hypervisor consideration" spec https://review.openstack.org/645814 | 07:29 |
bauzas | kashyap: just leave a comment in https://review.openstack.org/#/c/642030/ saying you'd like to not have the spec quick approved | 07:31 |
bauzas | quickly* (or fast) | 07:31 |
*** ileixe has quit IRC | 07:32 | |
*** tosky has quit IRC | 07:33 | |
*** tosky has joined #openstack-nova | 07:33 | |
*** boxiang has joined #openstack-nova | 07:33 | |
*** ileixe has joined #openstack-nova | 07:35 | |
*** ileixe has quit IRC | 07:35 | |
*** zhubx has joined #openstack-nova | 07:36 | |
*** zhubx has quit IRC | 07:36 | |
*** boxiang has quit IRC | 07:37 | |
*** boxiang has joined #openstack-nova | 07:38 | |
*** sidx64 has quit IRC | 07:43 | |
* kashyap clicks | 07:43 | |
kashyap | bauzas: I'm still thinking of the design. I didn't write it, the original person who wrote for Stein isn't answering questions on IRC or on the spec | 07:44 |
kashyap | bauzas: I'm writing a more thorough comment after some thinking last evening. | 07:45 |
*** ralonsoh has joined #openstack-nova | 07:48 | |
kashyap | bauzas: Maybe a silly question: Can tenant users create flavors? | 07:50 |
bauzas | kashyap: there is no silly question | 07:50 |
bauzas | kashyap: flavors are managed by admins, but that's just a policy | 07:50 |
kashyap | Right | 07:50 |
bauzas | that's the main different between image metadata and flavor extra specs | 07:51 |
bauzas | ones are admin-defined, the others are user-defined | 07:51 |
kashyap | bauzas: Ah, noted; so flavor extra_specs are user-defined. | 07:54 |
kashyap | Thanks! | 07:55 |
*** jonher_ has joined #openstack-nova | 07:57 | |
bauzas | kashyap: uh that's not what I said :) | 07:58 |
*** lee1 has joined #openstack-nova | 07:58 | |
bauzas | sorry, I flipped the answer | 07:58 |
*** jonher has quit IRC | 07:58 | |
*** ralonsoh has quit IRC | 07:58 | |
*** lyarwood has quit IRC | 07:58 | |
*** frickler has quit IRC | 07:58 | |
*** jonher_ is now known as jonher | 07:58 | |
kashyap | Err, it's the other way round. | 07:58 |
*** ralonsoh has joined #openstack-nova | 07:58 | |
bauzas | kashyap: https://docs.openstack.org/nova/latest/configuration/policy.html | 07:59 |
*** frickler has joined #openstack-nova | 07:59 | |
bauzas | os_compute_api:os-flavor-extra-specs:create Default rule:admin_api Operations POST /flavors/{flavor_id}/os-extra_specs/ | 07:59 |
bauzas | but as a user, you can see the flavor extra specs | 08:00 |
*** dtantsur|afk is now known as dtantsur | 08:00 | |
kashyap | Nod | 08:00 |
*** lee1 is now known as lyarwood | 08:00 | |
openstackgerrit | Kunpeng Zhang proposed openstack/nova master: Enhance live-migration progress log for memory and block data https://review.openstack.org/650621 | 08:00 |
*** rcernin has quit IRC | 08:01 | |
*** ccamacho has quit IRC | 08:01 | |
*** markvoelker has joined #openstack-nova | 08:03 | |
*** ileixe has joined #openstack-nova | 08:05 | |
*** ttsiouts has joined #openstack-nova | 08:05 | |
openstackgerrit | Dakshina Ilangovan proposed openstack/nova-specs master: Implements: blueprint nova-local-resource-management-that-uses-rmd https://review.openstack.org/651130 | 08:06 |
*** phasespace has joined #openstack-nova | 08:08 | |
*** priteau has joined #openstack-nova | 08:15 | |
*** wolverineav has joined #openstack-nova | 08:15 | |
*** luksky has joined #openstack-nova | 08:15 | |
kashyap | bauzas: One more: tenant users can't even set traits on flavors right? (My ans: They can't, since flavors are created by admins.) | 08:17 |
kashyap | s/flavors right?/flavors, right?/ | 08:17 |
bauzas | kashyap: you don't "set" a trait on a flavor, but you rather add a trait as an extra spec field | 08:17 |
bauzas | so, yeah to what you said | 08:17 |
kashyap | bauzas: Noted, thanks for correcting :-) | 08:17 |
bauzas | you 'set' a trait on a RP | 08:18 |
*** johanssone has quit IRC | 08:18 | |
kashyap | (Ah, Resource Provider; I don't much about them. For now, I'll table that topic. ) | 08:18 |
bauzas | kashyap: no worries, just explaining for you and also other folks | 08:18 |
kashyap | Thx. | 08:19 |
*** wolverineav has quit IRC | 08:20 | |
kashyap | yaawang: jackding: Hey, on this "select CPU model from a list" spec (https://review.openstack.org/#/c/642030/), I have some more unanswered design questions. | 08:20 |
kashyap | yaawang: jackding: I think we should sort them out first. Probably I will send an email to the 'openstack-discuss' list and Cc you both. | 08:20 |
*** johanssone has joined #openstack-nova | 08:24 | |
*** zbr has quit IRC | 08:28 | |
*** sidx64 has joined #openstack-nova | 08:28 | |
*** yankcrime has joined #openstack-nova | 08:33 | |
*** markvoelker has quit IRC | 08:36 | |
*** ttsiouts has quit IRC | 08:37 | |
*** ttsiouts has joined #openstack-nova | 08:37 | |
yaawang | kashyap: OK, do we need to open a zoom call to figure out some of the questions? | 08:39 |
kashyap | What is a "zoom call"? | 08:39 |
*** derekh has joined #openstack-nova | 08:40 | |
kashyap | I think I will write up the questions, so that those who are not present can also read them. | 08:40 |
kashyap | (And it serves as a record as well.) | 08:40 |
*** ttsiouts has quit IRC | 08:40 | |
*** ttsiouts has joined #openstack-nova | 08:40 | |
yaawang | s/zoom call/online meeting/ | 08:42 |
*** tkajinam has quit IRC | 08:43 | |
yaawang | kashyap: OK, I agree with you. When can you send the email? | 08:43 |
kashyap | yaawang: By end of today (I'm in CEST) | 08:44 |
yaawang | kashyap: Got it, thanks. | 08:46 |
openstackgerrit | Boxiang Zhu proposed openstack/nova-specs master: Expose live migration tunnelled in rest api https://review.openstack.org/651139 | 08:48 |
*** Luzi has joined #openstack-nova | 08:56 | |
*** bhagyashris has quit IRC | 08:57 | |
*** davidsha has joined #openstack-nova | 09:06 | |
*** luksky has quit IRC | 09:15 | |
*** bhagyashris_ has joined #openstack-nova | 09:20 | |
*** jaosorior has quit IRC | 09:24 | |
openstackgerrit | Boxiang Zhu proposed openstack/nova master: Make evacuation respects anti-affinity rule https://review.openstack.org/649953 | 09:29 |
*** jaosorior has joined #openstack-nova | 09:31 | |
openstackgerrit | Lee Yarwood proposed openstack/nova-specs master: Re-propose stable device rescue for Train https://review.openstack.org/651151 | 09:33 |
*** sidx64 has quit IRC | 09:34 | |
*** cdent has joined #openstack-nova | 09:39 | |
*** bhagyashris_ has quit IRC | 09:40 | |
*** sidx64 has joined #openstack-nova | 09:45 | |
*** tpatil has quit IRC | 09:47 | |
*** sidx64 has quit IRC | 09:51 | |
*** sidx64 has joined #openstack-nova | 09:54 | |
*** udesale has quit IRC | 09:57 | |
*** udesale has joined #openstack-nova | 09:58 | |
*** psachin has joined #openstack-nova | 09:59 | |
frickler | is there a list of todos regarding eliminating nova-network from the docs? this should be added to it https://docs.openstack.org/nova/latest/admin/networking-nova.html#metadata-service-deploy | 10:07 |
*** luksky has joined #openstack-nova | 10:08 | |
*** gary_perkins_ has quit IRC | 10:11 | |
*** gary_perkins has joined #openstack-nova | 10:11 | |
*** ttsiouts has quit IRC | 10:25 | |
*** ttsiouts has joined #openstack-nova | 10:26 | |
*** ttsiouts has quit IRC | 10:30 | |
*** markvoelker has joined #openstack-nova | 10:33 | |
*** bbowen has quit IRC | 10:36 | |
*** sidx64 has quit IRC | 10:41 | |
*** nicolasbock has joined #openstack-nova | 10:46 | |
*** tetsuro has quit IRC | 10:46 | |
*** ttsiouts has joined #openstack-nova | 10:46 | |
*** tbachman has quit IRC | 10:48 | |
*** owalsh has quit IRC | 10:53 | |
*** owalsh_ has joined #openstack-nova | 10:53 | |
openstackgerrit | Kashyap Chamarthy proposed openstack/nova-specs master: Add "Secure Boot support for KVM & QEMU guests" spec https://review.openstack.org/506720 | 10:54 |
*** udesale has quit IRC | 10:56 | |
kashyap | bauzas: Please don't "-1" based on "bad feelings": https://review.openstack.org/#/c/642030/ | 10:58 |
kashyap | If you have a good rationale, then do so. Otherwise, just leave a comment. | 10:58 |
sean-k-mooney | kashyap: i have not read it but i think -1 form a core because they have bad feels about something is not unwarrented | 10:59 |
kashyap | bauzas: Also, no -- just using the plural `cpu_models` is _not_ "simple". It can cause confusion | 10:59 |
kashyap | Please read the commit message, where I outlined _why_ we should use the name I proposed. | 10:59 |
kashyap | sean-k-mooney: Well, I see the point. But I'd prefer a -1 to be backed with rationale. (In this case bauzas did ask a few valid questions, though. It's all fine.) | 11:00 |
kashyap | Maybe I'm feeling a bit cranky with too many things going on ... don't take me _too_ seriously :-) | 11:01 |
sean-k-mooney | kashyap: when you have core right you can be more granular as you can -2 when you really have stong feeling somthing shoudl not be done | 11:01 |
kashyap | Nod | 11:02 |
sean-k-mooney | kashyap: by the way cpu_model_list = Conroe,Penryn,Nehalem,Westmere,SandyBridge,IvyBridge,Haswell,Broadwell,Skylake-Client,Skylake-Server | 11:03 |
sean-k-mooney | i assume that would prefer Conroe and proceed left to right | 11:03 |
*** sidx64 has joined #openstack-nova | 11:04 | |
openstackgerrit | Dakshina Ilangovan proposed openstack/nova-specs master: Nova local resource management that uses RMD https://review.openstack.org/651130 | 11:04 |
sean-k-mooney | in other words the operator is enabling the older cpu models first thereby ensuring the maxium compatiblity | 11:05 |
kashyap | sean-k-mooney: This whole idea of "CPUs must be oredered from least featureful to most" is flaky to me. | 11:05 |
kashyap | (I've asked the QEMU and libvirt devs for some more input, I'll respond on the change once I have clearer idea.) | 11:05 |
sean-k-mooney | i think it will work fine | 11:06 |
kashyap | sean-k-mooney: It is more tedious, isn't it, to get the order right, _while_ being aware of what CPUs one has locally? | 11:06 |
sean-k-mooney | we have prescedent for this already in neutron | 11:06 |
sean-k-mooney | it is more tedious but i doubt may people would list all of them | 11:06 |
sean-k-mooney | typically datacenters have only a handful of generations deployed concurrently | 11:07 |
*** markvoelker has quit IRC | 11:07 | |
sean-k-mooney | so i woudl expect the list to contain 2-3 in general | 11:07 |
sean-k-mooney | there is no point in listing a generation older then any of your servers afterall | 11:08 |
kashyap | Hmm, okay, that's somewhat palatable | 11:09 |
sean-k-mooney | so somthign like cpu_model_list = nehalam,ivybridge,skylake-server woudl be more typical | 11:09 |
*** zbr has joined #openstack-nova | 11:09 | |
*** sidx64 has quit IRC | 11:15 | |
*** sidx64 has joined #openstack-nova | 11:21 | |
kashyap | Right; but who knows | 11:25 |
*** sidx64 has quit IRC | 11:26 | |
*** sidx64 has joined #openstack-nova | 11:28 | |
*** mvkr has joined #openstack-nova | 11:30 | |
kashyap | sean-k-mooney: Thanks for the comments | 11:34 |
kashyap | sean-k-mooney: It seems like I have suddenly "inherited" this spec, as the original proposer for Stein is nowhere to be seen. | 11:34 |
sean-k-mooney | no worries. i need to write one sepc and then start makeing my way through the spec review queue | 11:37 |
kashyap | The more I think about this spec, the more questions arise. | 11:37 |
kashyap | (And yes, I'll move the rationale from commit message to the spec itself.) | 11:38 |
sean-k-mooney | i wrote a version of this sepc 3 years ago to satify telefonic. tye wanted to specify the cpu_model in the flavor. so i proposed adding a new cpu_mode=dynamic which allowed the flavor to set the cpu_model that was required | 11:39 |
sean-k-mooney | that spec predated placement so i think i also had a new schduler filter to do the compatiblity check | 11:39 |
sean-k-mooney | telcos have wanted somthign like this for a long time but its a blance between cloudines and hardware defined software | 11:41 |
sean-k-mooney | in this iteration the use of traits to exrpess the workload requirements is a nice abstration | 11:42 |
cdent | jaypipes++ | 11:45 |
jaypipes | cdent: ? | 11:46 |
cdent | rsd spec | 11:46 |
jaypipes | ah | 11:46 |
jaypipes | just starting off spec review day... | 11:46 |
kashyap | sean-k-mooney: What do you mean by "there is no easy way using qemu or libvirt for an operator to check the cpu model flags"? | 11:47 |
cdent | keep rolling like that and it should be a fine day | 11:47 |
kashyap | sean-k-mooney: You can trivially see what flags QEMU binary supports -- surely you that | 11:47 |
kashyap | a/"QEMU binary"/"a given QEMU binary on the host"/ | 11:48 |
sean-k-mooney | jaypipes: the reason that intel is proposeing that spec is because i pushed for that change in architure after it became apparent that doing it outside of nova had several disavantages | 11:48 |
sean-k-mooney | jaypipes: one of the main disadvanatges beside the fact taht efforts to do this outside of nova have failed at lest 3 time was when you go done that path using openstack at all becomes questionable | 11:49 |
*** bbowen has joined #openstack-nova | 11:50 | |
jaypipes | sean-k-mooney: so square peg, round hole it into Nova and that problem will go away? :) | 11:53 |
sean-k-mooney | jaypipes: part of the motivation of enableing RSD as a composible hypervior under nova is to ensure that end users can continue to use openstack api without needing to know their instances are provide by rsd but operators can leverage it | 11:53 |
jaypipes | sean-k-mooney: that is precisely why I believe it needs to be outside of Nova. | 11:53 |
sean-k-mooney | jaypipes: do you belive we shoudl delete the ironic driver? | 11:54 |
jaypipes | sean-k-mooney: no. | 11:54 |
sean-k-mooney | the integration will be almost identical | 11:54 |
jaypipes | sean-k-mooney: no it won't. | 11:54 |
jaypipes | specifically, having to create flavors on the fly | 11:55 |
sean-k-mooney | well i spend a year modeling it on how ironic worked with minor tweeks | 11:55 |
sean-k-mooney | jaypipes: that is not required | 11:55 |
sean-k-mooney | that was a suggesting to make testing eaier | 11:55 |
sean-k-mooney | the flavor it create just contain the reouces request for the custom_resouce class | 11:56 |
jaypipes | sean-k-mooney: I just don't think it's a good idea, sorry. | 11:56 |
sean-k-mooney | ok but for the record i did bring this idea up at the first denver ptg and the vancour summit to make sure people were ok with it | 11:57 |
sean-k-mooney | you werent at teh vancour summit | 11:57 |
sean-k-mooney | but i was pretty sure i ran this by you at some point | 11:57 |
*** sidx64 has quit IRC | 11:59 | |
kashyap | gibi: On your (good) question on the change, please read this spec, too (if you have time). It aims to answer your question: https://review.openstack.org/#/c/645814/4/specs/train/approved/cpu-selection-with-hypervisor-consideration.rst | 11:59 |
*** tbachman has joined #openstack-nova | 11:59 | |
*** ttsiouts has quit IRC | 12:00 | |
*** jmlowe has quit IRC | 12:01 | |
jaypipes | sean-k-mooney: OK, but for the record, I've had the exact same position for 4 years since Intel proposed it when I was working for Mirantis. | 12:01 |
jaypipes | :) | 12:01 |
*** ttsiouts has joined #openstack-nova | 12:01 | |
jaypipes | and you were working for Intel :) | 12:01 |
*** jmlowe has joined #openstack-nova | 12:02 | |
*** nicolasbock has quit IRC | 12:02 | |
*** nicolasbock has joined #openstack-nova | 12:02 | |
sean-k-mooney | :) ok yes but i guess we have always dissagreed then we never tought vlance was a viable approch in my team | 12:02 |
gibi | kashyap: ack. There is too many specs :) | 12:03 |
sean-k-mooney | it lost the cloud abstration meaning existing workloads/workflows needed to be modifyed to know what rsd was | 12:03 |
jaypipes | sean-k-mooney: if you think RSD is "cloudy" as opposed to just a tool for operators to reduce power consumption in their datacenters, then I suppose you might think that. | 12:07 |
jaypipes | sean-k-mooney: however, I don't. I don't see RSD as anything other than a compute node dynamic composer. I don't see it as the next generation of "packaging" for a workload. | 12:08 |
jaypipes | sean-k-mooney: it's hardware, plain and simple. It's not the next Kubernetes abstraction for a new type of workload. | 12:08 |
sean-k-mooney | yes i agree which is why i think openstacks apis need to be the abstration | 12:09 |
jaypipes | sean-k-mooney: we will just have to agree to disagree on this topic :) | 12:09 |
sean-k-mooney | i think of it as a bearmental hypervior | 12:09 |
sean-k-mooney | like ironic | 12:09 |
*** davidsha has quit IRC | 12:10 | |
*** elod has quit IRC | 12:11 | |
*** elod has joined #openstack-nova | 12:11 | |
kashyap | gibi: Hehe, sorry. Yeah, context overload; completely understand it. | 12:13 |
kashyap | gibi: Take a look at it, on a non-spec day; I'll assume everyone is maxed out today | 12:14 |
gibi | kashyap: I kept that spec open, but I move by the etherpad order https://etherpad.openstack.org/p/nova-spec-review-day | 12:14 |
kashyap | (And looking at too many specs parallely will only splinter the attention to hell and back, which is costly.) | 12:14 |
*** wolverineav has joined #openstack-nova | 12:16 | |
*** sidx64 has joined #openstack-nova | 12:18 | |
jaypipes | sean-k-mooney: "bearmental hypervior" <-- nice. | 12:18 |
kashyap | Folks: On line-42 I added a spec I meant to add on 04-Apr (as mentioned on the list to Eric's email on pre-PTG review); didn't post it in the last minute :-) | 12:19 |
*** wolverineav has quit IRC | 12:20 | |
sean-k-mooney | :) | 12:21 |
*** tiendc has quit IRC | 12:22 | |
openstackgerrit | Merged openstack/nova master: Improve test coverage of nova.privsep.path. https://review.openstack.org/648601 | 12:22 |
openstackgerrit | Merged openstack/nova master: Improve test coverage of nova.privsep.fs. https://review.openstack.org/648602 | 12:22 |
gibi | kashyap: you really want me to read that spec :) | 12:23 |
openstackgerrit | Merged openstack/nova master: Improve test coverage of nova.privsep.fs, continued. https://review.openstack.org/648603 | 12:24 |
*** ratailor has quit IRC | 12:24 | |
kashyap | gibi: Hehe, the above mentioned line-42 one is about Secure Boot one (which jaypipes reviewed back in the day; before we identified work to be done in lower layers) | 12:25 |
gibi | kashyap: ohh, I looked at the line 41 instead. | 12:25 |
kashyap | gibi: Yep, that is fairly uncontroversial. (Says the guy who proposed it.) | 12:25 |
gibi | kashyap: anyhow I'm reading the spec at l41 now :) | 12:25 |
kashyap | Thanks! | 12:25 |
kashyap | I need to be AFK in 35 mins; will respond on the change for any feedback | 12:26 |
kashyap | jaypipes: If you still have the stomach for it this week, might want to have a gander at the Secure Boot thing, in your "copious free time" :D | 12:26 |
* kashyap has posted periodic updates on the spec over the year, as work got completed in lower layers (OVMF, QEMU and libvirt) | 12:26 | |
jaypipes | kashyap: yep, today is spec review day. | 12:27 |
* kashyap is aware; but I'm aware of the limited resource that is human attention. So I don't expect people to get to it today | 12:30 | |
*** jamesdenton has quit IRC | 12:33 | |
openstackgerrit | Kashyap Chamarthy proposed openstack/nova-specs master: Re-propose the spec to allow specifying a list of CPU models https://review.openstack.org/642030 | 12:37 |
*** tetsuro has joined #openstack-nova | 12:37 | |
*** tetsuro has quit IRC | 12:40 | |
bauzas | kashyap: sean-k-mooney: sorry was afk for lunch | 12:40 |
bauzas | so I missed your pings | 12:40 |
*** maciejjozefczyk has quit IRC | 12:40 | |
bauzas | don't take a -1 for a passive-aggressive method | 12:40 |
bauzas | it's just "I disagree with the current patch, because of : X, Y" | 12:40 |
kashyap | bauzas: No worries; it's all sorted | 12:41 |
bauzas | and my -1 is not about the naming, but rather about the fact that operators should provide a list ordered | 12:41 |
kashyap | bauzas: Yeah, saw your question; see the response from Sean and me... | 12:41 |
sean-k-mooney | bauzas: yes i know. that is how i use it too. | 12:41 |
bauzas | I'll look at the comments, for the moment, I'm looking at other spec | 12:42 |
*** priteau has quit IRC | 12:42 | |
*** openstackgerrit has quit IRC | 12:44 | |
*** dtantsur is now known as dtantsur|brb | 12:46 | |
*** Luzi_ has joined #openstack-nova | 12:46 | |
*** openstackgerrit has joined #openstack-nova | 12:48 | |
openstackgerrit | Theodoros Tsioutsias proposed openstack/nova master: Add instance hard delete https://review.openstack.org/570202 | 12:48 |
openstackgerrit | Theodoros Tsioutsias proposed openstack/nova master: Add requested_networks to RequestSpec https://review.openstack.org/570201 | 12:48 |
openstackgerrit | Theodoros Tsioutsias proposed openstack/nova master: Enable rebuild for instances in cell0 https://review.openstack.org/570203 | 12:48 |
*** lbragstad has joined #openstack-nova | 12:49 | |
openstackgerrit | Theodoros Tsioutsias proposed openstack/nova master: Introduce the PENDING instance state https://review.openstack.org/566473 | 12:49 |
openstackgerrit | Theodoros Tsioutsias proposed openstack/nova master: Allow rebuild for instances in PENDING state https://review.openstack.org/637585 | 12:49 |
*** Luzi has quit IRC | 12:49 | |
efried | kashyap: ack | 12:50 |
sean-k-mooney | its surpring how often "sudo rm -rf /var/lib/docker/" fixs your docker issues | 12:51 |
efried | I at least diff from the last release's version before doing that; this one has changed substantially so I'll wait | 12:51 |
efried | What's the adage? "If you try to solve a problem using docker, now you have two problems." ? | 12:52 |
efried | dansmith: Would you please update the channel topic to something like | 12:53 |
efried | Welcome to Train pre-PTG spec review day #1: https://etherpad.openstack.org/p/nova-spec-review-day | 12:53 |
*** jroll has quit IRC | 12:55 | |
*** owalsh_ is now known as owalsh_afk | 12:58 | |
*** jroll has joined #openstack-nova | 12:59 | |
*** mdbooth has quit IRC | 12:59 | |
*** mdbooth has joined #openstack-nova | 12:59 | |
*** mchlumsky has joined #openstack-nova | 13:00 | |
*** eharney has joined #openstack-nova | 13:00 | |
*** dikonoor has joined #openstack-nova | 13:04 | |
*** sidx64 has quit IRC | 13:05 | |
*** dikonoor has quit IRC | 13:08 | |
*** tbachman has quit IRC | 13:10 | |
openstackgerrit | Huqiang Wang proposed openstack/nova-specs master: Separate the vCPUs into different pools based on priority https://review.openstack.org/649882 | 13:13 |
*** tbachman has joined #openstack-nova | 13:16 | |
*** Luzi_ has quit IRC | 13:17 | |
*** amodi has joined #openstack-nova | 13:19 | |
openstackgerrit | Dakshina Ilangovan proposed openstack/nova-specs master: Nova LLC allocation - RMD plugin for RDT CAT https://review.openstack.org/651233 | 13:19 |
openstackgerrit | Lee Yarwood proposed openstack/nova master: Use migration_status during volume migrating and retyping https://review.openstack.org/637224 | 13:21 |
*** awaugama has joined #openstack-nova | 13:24 | |
*** dakshina-ilangov has joined #openstack-nova | 13:25 | |
*** belmoreira has joined #openstack-nova | 13:26 | |
*** ChanServ sets mode: +o dansmith | 13:28 | |
*** priteau has joined #openstack-nova | 13:28 | |
*** IvensZambrano has joined #openstack-nova | 13:28 | |
*** dansmith changes topic to "Welcome to Train pre-PTG spec review day #1: https://etherpad.openstack.org/p/nova-spec-review-day -- Current runways: https://etherpad.openstack.org/p/nova-runways-train -- This channel is for Nova development. For support of Nova deployments, please use #openstack." | 13:28 | |
*** tbachman has quit IRC | 13:32 | |
*** udesale has joined #openstack-nova | 13:36 | |
kashyap | sean-k-mooney: On "Docker", you're so 2014. 'Everyone' has moved on from Docker to "CRI-O" and "podman", etc. | 13:37 |
kashyap | (Please update your "container jargon" :D) | 13:37 |
* kashyap stops trolling, lest he gets banned | 13:37 | |
*** tbachman has joined #openstack-nova | 13:38 | |
*** maciejjozefczyk has joined #openstack-nova | 13:38 | |
sean-k-mooney | kashyap: i am using docker and plan to never use podman. i might use CRI-O | 13:38 |
kashyap | sean-k-mooney: I was massively trolling; sorry. (I have no clue about their compatibilities.) | 13:39 |
sean-k-mooney | podman has 1:1 commandline compatible with docker | 13:41 |
sean-k-mooney | or the docker cli | 13:41 |
*** lpetrut has joined #openstack-nova | 13:42 | |
*** malekmar has joined #openstack-nova | 13:42 | |
sean-k-mooney | cri-o is interesting but its non tivial to setup currently | 13:42 |
*** malekmar has quit IRC | 13:42 | |
*** mriedem has joined #openstack-nova | 13:43 | |
mriedem | mordred: as our resident compute api/sdk opinion czar, i'd like your opinion on this spec https://review.openstack.org/#/c/638629/9/specs/train/approved/add-locked-reason.rst@198 | 13:43 |
mriedem | specifically the response format for those new fields | 13:43 |
mriedem | i.e. nested or flat | 13:44 |
*** liuyulong|away is now known as liuyulong | 13:44 | |
kashyap | gibi: Thanks! (I spent a lot of time wordsmithing to get words right in that spec) | 13:45 |
*** belmorei_ has joined #openstack-nova | 13:46 | |
*** phasespace has quit IRC | 13:48 | |
*** belmoreira has quit IRC | 13:48 | |
*** awalende has quit IRC | 13:58 | |
*** awalende has joined #openstack-nova | 13:59 | |
*** maciejjozefczyk has quit IRC | 13:59 | |
*** mlavalle has joined #openstack-nova | 14:00 | |
*** awalende has quit IRC | 14:03 | |
*** davidsha_ has joined #openstack-nova | 14:04 | |
*** priteau has quit IRC | 14:04 | |
*** lpetrut has quit IRC | 14:04 | |
*** ttsiouts has quit IRC | 14:06 | |
*** ttsiouts has joined #openstack-nova | 14:07 | |
*** ttsiouts has quit IRC | 14:09 | |
*** ttsiouts has joined #openstack-nova | 14:09 | |
mriedem | tssurya: some questions in your ironic power state sync spec https://review.openstack.org/#/c/636132/ | 14:11 |
mriedem | dansmith: ^ you should probably review that as well | 14:11 |
tssurya | mriedem: ack, looking | 14:11 |
mriedem | jaypipes maybe too since he loves ironic nowadays | 14:11 |
jaypipes | mriedem: errmmm, not quite :) | 14:12 |
jaypipes | but I'll review regardless. | 14:12 |
dansmith | mriedem: on meetings for the next couple hours, but willt ry | 14:15 |
kashyap | Frelling rST | 14:22 |
*** belmoreira has joined #openstack-nova | 14:22 | |
mriedem | gibi: i've replied to your questions in the cross-cell resize re-proposal https://review.openstack.org/#/c/642807/ | 14:23 |
mriedem | suppose dansmith and melwitt should look at that one as well since they are familiar with the original in stein | 14:23 |
*** belmorei_ has quit IRC | 14:25 | |
* alex_xu has wonderful spec review day with have cold and a lot of snot and no enough sleep and full day meeting | 14:25 | |
kashyap | alex_xu: Gosh, it is 22:26. Nothing is more important than sleep... | 14:26 |
alex_xu | \o/ sleep now | 14:27 |
*** belmoreira has quit IRC | 14:27 | |
* kashyap plugs the fantastic book by the amazing Matthew Walker (https://www.penguin.co.uk/books/295/295665/why-we-sleep/9780141983769.html) | 14:27 | |
mriedem | alex_xu: i still have some chinese cold medicine from the shenzhen airport i can send you :) | 14:27 |
mriedem | i don't know what it says on the packaging but i know i'm only supposed to take 2 of each every 12 hours | 14:27 |
*** dtantsur|brb is now known as dtantsur | 14:30 | |
openstackgerrit | Stephen Finucane proposed openstack/nova-specs master: Add 'flavor-extra-spec-image-property-validation-extended' spec https://review.openstack.org/638734 | 14:37 |
*** belmoreira has joined #openstack-nova | 14:37 | |
*** cfriesen has joined #openstack-nova | 14:39 | |
stephenfin | kashyap: Fair to say https://review.openstack.org/#/c/625216/ means https://review.openstack.org/#/c/651139/ is unnecessary? | 14:40 |
* kashyap looks | 14:40 | |
*** N3l1x has joined #openstack-nova | 14:41 | |
*** owalsh_afk is now known as owalsh | 14:41 | |
kashyap | stephenfin: Just took a cursory look, so they're proposing to expose _tunnelled via REST API. That's baloney | 14:42 |
kashyap | We don't want _tunnelled at all, we will deprecate it, as you know | 14:42 |
*** igordc has joined #openstack-nova | 14:42 | |
stephenfin | kashyap: Yup, that's what I was thinking but I just wanted to be sure it was as baloney as I thought it was | 14:43 |
kashyap | "We" as in, once we bump the min versions to have hypervisor-native TLS, then Nova can nuke _tunnelled support | 14:43 |
openstackgerrit | Eric Fried proposed openstack/nova-specs master: Tools & docs for backlog & abandoned spec process https://review.openstack.org/648800 | 14:44 |
efried | stephenfin: Did I do that right? ^ | 14:44 |
* kashyap also links to: http://git.openstack.org/cgit/openstack/nova/tree/doc/source/admin/secure-live-migration-with-qemu-native-tls.rst | 14:44 | |
kashyap | stephenfin: Aside: please deliver me from my misery by sharing the "genuine table" syntax. I'm only seeing these "naked tables": https://kashyapc.fedorapeople.org/Plain-rendering-of-table-Sphinx.png | 14:44 |
kashyap | (Also, I like the .. code-block: text tip from efried; it's visually clean and immediately shows what a user wants) | 14:45 |
stephenfin | kashyap: Ahh, that's the theme's fault. I'd been testing with the default Sphinx theme | 14:45 |
kashyap | (s/what a user wants/what the API entails/) | 14:45 |
stephenfin | Bad openstackdocstheme | 14:45 |
*** belmoreira has quit IRC | 14:46 | |
stephenfin | kashyap: OK, ignore that comment so. That's certainly more difficult to read that one would like | 14:46 |
kashyap | Yeah, phew. Thanks for confirming that I'm missing to see something that only you seem to see :-) | 14:46 |
kashyap | s/that I'm missing/that I'm _not_ missing/ | 14:46 |
stephenfin | efried: 'ish. You don't need to define the 'nodeps' testenv | 14:46 |
efried | stephenfin: There was another place where the lack of (even the ability to request) table borders was very annoying. | 14:46 |
gibi | mriedem: thanks. the cross resize patch looks good to me. I did +1 it only as I feel others should approve it | 14:47 |
efried | stephenfin: I figured I had the choice between defining that env or repeating 'deps=' in both, right? | 14:47 |
efried | figured the former was more DRY-ish and less prone to future slippage | 14:47 |
efried | can flip it if you prefer it the other way, whatevs. | 14:48 |
stephenfin | efried: Not really. deps for a given section will inherit from '[testenv]' by default | 14:48 |
stephenfin | So you've reverted back to 'deps = -r{toxinidir}/doc/requirements.txt' for those two sections | 14:48 |
efried | eh? | 14:48 |
*** lpetrut has joined #openstack-nova | 14:48 | |
efried | Even though I said envdir= ? | 14:48 |
efried | okay, I'll twiddle it again. | 14:49 |
stephenfin | yeah, envdir has no bearing on deps | 14:49 |
efried | ight | 14:49 |
mriedem | kashyap: stephenfin: gibi: i've left several comments and questions in this baselineHypervisorCPU spec https://review.openstack.org/#/c/645814/ and i must be missing something that everyone else understands about what this solves | 14:49 |
kashyap | Will look | 14:49 |
openstackgerrit | Eric Fried proposed openstack/nova-specs master: Tools & docs for backlog & abandoned spec process https://review.openstack.org/648800 | 14:49 |
stephenfin | other than if you use the same envdir with different deps, in which case you get the aforementioned rebuilding on environment each time | 14:50 |
efried | whee | 14:50 |
efried | done | 14:50 |
efried | ---^ | 14:50 |
efried | Thanks for the help stephenfin | 14:50 |
*** jistr is now known as jistr|call | 14:51 | |
gibi | mriedem: I think https://review.openstack.org/#/c/645814/ and https://review.openstack.org/#/c/642030/ are connected | 14:53 |
gibi | mriedem: so when nova will automatically select a specific cpu_model from a list of models it will select a better suited one | 14:53 |
mriedem | the former doesn't mention anything about the latter | 14:54 |
*** maciejjozefczyk has joined #openstack-nova | 14:54 | |
*** maciejjozefczyk has quit IRC | 14:55 | |
stephenfin | efried: +2'd | 14:55 |
efried | thank you sir. | 14:55 |
gibi | mriedem: that is true | 14:55 |
*** lbragstad has quit IRC | 14:56 | |
gibi | mriedem: kashyap knows more about the connection | 14:56 |
*** lbragstad has joined #openstack-nova | 14:56 | |
kashyap | I did mention about this https://review.openstack.org/#/c/645814/ in https://review.openstack.org/#/c/645814/ | 14:57 |
*** NostawRm has joined #openstack-nova | 14:57 | |
mriedem | i assume you meant to link to https://review.openstack.org/#/c/642030/ | 14:57 |
mriedem | which i haven't read | 14:57 |
kashyap | mriedem: The "executive summary" is that the two libvirt APIs Nova uses today don't take account what the host KVM, QEMU and libvirt are capable of, when computing CPU models. And the proposed two new APIs will solve a class of issues. | 14:57 |
kashyap | Among other things, Nova can now ask: "Are CPU flags X and Y supported by KVM, QEMU and libvirt on the host"? | 14:58 |
kashyap | How does that help operators? | 14:58 |
kashyap | It gives them more _useful_ (and in some cases _secure_ -- e.g. CPU flags introduced to solve Meltdown/Spectre) guest CPU configuration | 14:59 |
kashyap | (Not everything is linked to migration.) | 14:59 |
kashyap | I'll respond on the review, to save some typing. | 14:59 |
dansmith | kashyap: are you talking about for scheduling or for generating more optimal xml for a guest? | 14:59 |
mriedem | it's not clear to me if the operator is doing anything here, or if operators today are doing x to workaround limitations in nova's libvirt driver using these older apis | 15:00 |
mriedem | the spec just isn't clear to me on what the end user visible benefit is | 15:01 |
mriedem | i'm sure there is one, but it's just not clear to me | 15:01 |
*** ttsiouts has quit IRC | 15:01 | |
mriedem | 1. use new apis, 2. ?, 3. profit. | 15:01 |
kashyap | dansmith: Not scheduling. In some ways you can call it "optimal XML for a guest", but not everything is about XML generation. It is more about how Nova's libvirt driver can take into account multiple important factors (listed in the spec) when coming up with guest CPUs | 15:02 |
mordred | mriedem: I think either thing is fine from an openstacksdk POV - but I could see the shift from bool to dict being maybe a little more annoying for the more staticly typed folks like go - whereas the flat approach like it is in the notification payload would be potentially friendlier to deal with there | 15:02 |
mriedem | i should probably re-watch your summit talk that is related to this | 15:02 |
*** ttsiouts has joined #openstack-nova | 15:02 | |
dansmith | kashyap: okay, I | 15:02 |
mordred | mriedem: so I think I'd vote, in this case, for flat, mostly because of the existence of the locked key already | 15:02 |
dansmith | have only skimmed, but.. probably need to be more clear about that | 15:02 |
kashyap | mriedem: It is not directly a clean one-to-one mapping of: "if you click on this button, elephants will dance on your screen" | 15:02 |
kashyap | mriedem: The operator doesn't have to do anything much here | 15:03 |
mriedem | is there a very obvious case of "this is an issue we have today which is a limitation due to using the older apis which would be solved by using the new" | 15:03 |
mriedem | like an actual example | 15:03 |
kashyap | Yes, let me link to the code | 15:04 |
mriedem | mordred: ok, if you can take a sec to comment on the spec that would go a long way i think | 15:04 |
mordred | sure! | 15:04 |
kashyap | mriedem: libvirt upstream added the two APIs (partly) based on the need that came out of Nova, noted here: https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L3947,#L3958 | 15:05 |
kashyap | (That comment should be removed, now that the libvirt RFE I filed is fixed.) | 15:05 |
openstackgerrit | sean mooney proposed openstack/nova-specs master: Libvirt: add vPMU spec for train https://review.openstack.org/651269 | 15:05 |
*** ttsiouts has quit IRC | 15:05 | |
*** ttsiouts has joined #openstack-nova | 15:06 | |
mordred | mriedem: done | 15:06 |
kashyap | mriedem: A practical example is: the existing libvirt APIs can't answer this: Is this combination of 'IvyBridge' CPU model with 'pcid' and 'pdpe1gb' supported by the KVM+QEMU+libvirt on the host? | 15:07 |
sean-k-mooney | kashyap: isnt that a defct in libvirt | 15:07 |
kashyap | sean-k-mooney: Yes, it is. That's why I asked libvirt upstream to provide the two new RFEs: https://bugzilla.redhat.com/show_bug.cgi?id=1559832 | 15:08 |
openstack | bugzilla.redhat.com bug 1559832 in libvirt "[RFE] Fine-grained API to validate if a given CPU model and flags are supported by QEMU / KVM" [Unspecified,Closed: nextrelease] - Assigned to jdenemar | 15:08 |
sean-k-mooney | it could tell if pdpe1gb was supporte by looking the the host cpu flags | 15:08 |
kashyap | (To solve the problem for Nova.) | 15:08 |
mriedem | kashyap: so today operators can just populate cpu_model_extra_flags with ['pcid', 'pdpe1gb'] and we'll blindly throw those into the guest cpu config, which may or may not cause the guest to fail to launch? | 15:08 |
sean-k-mooney | kashyap: yes but this is somthing that qemu should also be able to do today but it does not provde a cli to ask to question | 15:08 |
kashyap | sean-k-mooney: I think we're hand-waving; if you see the upstream patch `diff` for libvirt that introduced the APIs, you'll see what is involved | 15:08 |
kashyap | And also please see the spec I posted. Which compares the two APIs | 15:08 |
kashyap | mriedem: Yep, exactly | 15:09 |
kashyap | sean-k-mooney: Here libvirt is taking advantage of the enormous work done in QEMU. (Look up: `query-cpu-model-expansion`) | 15:09 |
sean-k-mooney | mriedem: yes that why we originally had a whitelist of jsut pcid | 15:09 |
kashyap | sean-k-mooney: See this blog post from the QEMU's CPU modelling infra maintainer: https://habkost.net/posts/2017/03/qemu-cpu-model-probing-story.html | 15:09 |
mriedem | ok, so that would be good information for the spec, i.e. today operators have to be careful about how they configure nova-compute for the libvirt driver with regard to cpu_model and cpu_model_extra_flags because the wrong combination can lead to failures to launch a guest - with this change the libvirt driver will automatically figure that out so the operator can set it and forget it | 15:10 |
kashyap | mriedem: Yep, whatever it takes to make it clear. It's also the "curse of having too much context". I hesitate to add needless detail | 15:10 |
mriedem | mordred: thanks | 15:10 |
sean-k-mooney | i dont think we should forget it | 15:10 |
sean-k-mooney | if you enable somting in the config that is invalid i would prefer if the nova compute agent failed to start | 15:11 |
kashyap | (Yes, the whitelist was for stable branches) | 15:11 |
sean-k-mooney | kashyap: not just for stable branches | 15:11 |
mriedem | kashyap: i realize you're close to the libvirt / qemu details on this stuff, but consider your audience is not so as a nova spec reviewer i understand nova things and need to know how this translates to nova problems and how they are fixed by this spec | 15:11 |
kashyap | mriedem: I go to painful lengths to be aware of that. | 15:12 |
sean-k-mooney | kashyap: i reasied the problem of doing the validation on the reviews for you extra flags patches | 15:12 |
kashyap | mriedem: If you closely read what I write, I make a lot of effort to write with "no assumptions", and with a full-assed beginning, middle and an end :-) | 15:12 |
kashyap | sean-k-mooney: Well, I remember the change. You're forgetting something: | 15:12 |
kashyap | sean-k-mooney: Until Rocky we white-listed; after that we lifted it. | 15:13 |
sean-k-mooney | kashyap: yes and i pushed back on lifting it until we have validation of the extra flags | 15:13 |
mriedem | kashyap: the writing isn't the problem, i understand the words, they just don't have context for me | 15:13 |
mriedem | anyway if others are cool with approving this fine | 15:13 |
mriedem | i didn't -1 it | 15:13 |
kashyap | mriedem: Noted, I should have also linked the relevant Nova code, from which it came from. | 15:13 |
sean-k-mooney | i was over ruled but i personally still thing we should not have lifted the whitelist untill we have a way to validate it | 15:13 |
kashyap | mriedem: Yeah, noticed it. Appreciate the restraint :-) | 15:13 |
kashyap | mriedem: That spec (and the libvirt work) is derived directly from the problem Nova was facing, BTW | 15:14 |
kashyap | (Anyway, you see that now. I shold've made it clearer) | 15:14 |
kashyap | sean-k-mooney: I think there we disagree. Lifting the restrictions then made sense, and now it does. | 15:15 |
*** tbachman has quit IRC | 15:15 | |
sean-k-mooney | we only lifted the restition because of all the specultive execution vulnerablitys and the fact we tought there would need to be a lot of extra cpus flags in the coming months | 15:16 |
kashyap | sean-k-mooney: Recall the solid reason *why* we allowed free-form: at that time we had no damn idea what else CPU flags Intel et al will introduce | 15:16 |
dansmith | sean-k-mooney: are you talking about the cpu flags thing that we added for spectre | 15:16 |
dansmith | ? | 15:16 |
kashyap | Yep | 15:16 |
dansmith | sean-k-mooney: the whitelist was *only* for stable, never intended to have that in master going forward | 15:17 |
kashyap | Exactly. I think sean-k-mooney got things mixed up. | 15:17 |
sean-k-mooney | no | 15:17 |
sean-k-mooney | that was the desicion we came to in the end | 15:17 |
dansmith | um, what? | 15:18 |
sean-k-mooney | but i did not think we should add the feature at all without validation logic | 15:18 |
sean-k-mooney | i was ok with not having the validation logic if we had a whitelist | 15:18 |
sean-k-mooney | and i was ok with removing the whitelist due to the security concerns | 15:18 |
dansmith | tbh, I don't recall you even being involved :) | 15:19 |
sean-k-mooney | but i still think we should have the validation | 15:19 |
sean-k-mooney | i was at intel a the time and was not allowed to talk publically about it much | 15:19 |
sean-k-mooney | so im not sure how much of my comment were recored upstream and how much i said to indivigual people | 15:20 |
mriedem | efried: nit and question in https://review.openstack.org/#/c/648800/6 | 15:20 |
* gibi only was able to read half of the specs in the etherpad during the day and has to leave now | 15:21 | |
dansmith | well, anyway, the plan was to make it free-form going forward, and the compromise for stable was to add the whitelist instead of a different-for-stable config flag | 15:21 |
sean-k-mooney | ya and kashyap's spec https://review.openstack.org/#/c/645814/4/specs/train/approved/cpu-selection-with-hypervisor-consideration.rst add the validation i was originally hoping for so that makes me happy :) | 15:22 |
dansmith | the only reason for the whitelist was to avoid having a different config flag for stable than master | 15:22 |
kashyap | Yes, what dansmith said. | 15:22 |
*** sapd1_x has joined #openstack-nova | 15:22 | |
*** tbachman has joined #openstack-nova | 15:22 | |
openstackgerrit | Eric Fried proposed openstack/nova-specs master: Tools & docs for backlog & abandoned spec process https://review.openstack.org/648800 | 15:23 |
efried | mriedem: answered ^ | 15:23 |
efried | mriedem: stephenfin helped out with the tox thing, so it must be right. | 15:23 |
efried | gibi: Thank you for your participation! | 15:24 |
*** sapd1_x has quit IRC | 15:27 | |
*** lpetrut has quit IRC | 15:29 | |
sean-k-mooney | kashyap: ok but my request for validating the feature is there in https://review.openstack.org/#/c/534384/5/nova/conf/libvirt.py@538 anyway im almost through your spec which seams to provide it anyway. | 15:31 |
mriedem | efried: ok final question https://review.openstack.org/#/c/648800/7 | 15:31 |
*** ivve has quit IRC | 15:32 | |
kashyap | sean-k-mooney: Nod; I recall seeing the comment | 15:33 |
*** awalende has joined #openstack-nova | 15:33 | |
*** gyee has joined #openstack-nova | 15:35 | |
*** udesale has quit IRC | 15:37 | |
*** awalende has quit IRC | 15:37 | |
*** dave-mccowan has joined #openstack-nova | 15:40 | |
efried | mriedem: responded | 15:41 |
efried | mriedem: Believe me, I would have preferred to put the glob in this change instead of individually in the four on top :( | 15:41 |
*** tbachman has quit IRC | 15:41 | |
efried | I think gerrit will at least be smart enough not to generate a merge conflict on the three remaining when the first one merges... | 15:42 |
*** luksky has quit IRC | 15:42 | |
*** tbachman has joined #openstack-nova | 15:44 | |
efried | dakshina-ilangov: A couple of things about https://review.openstack.org/#/c/651130/ just procedurally: | 15:46 |
efried | 1) Please wrap the body of the commit message at 72c | 15:46 |
efried | 2) You can build locally by running `tox -e docs` to resolve formatting issues | 15:46 |
*** ttsiouts has quit IRC | 15:56 | |
openstackgerrit | Stephen Finucane proposed openstack/nova-specs master: Add 'flavor-extra-spec-image-property-validation-extended' spec https://review.openstack.org/638734 | 15:56 |
*** ttsiouts has joined #openstack-nova | 15:57 | |
*** jistr|call is now known as jistr | 16:00 | |
*** ttsiouts has quit IRC | 16:01 | |
*** dakshina-ilangov has quit IRC | 16:04 | |
openstackgerrit | François Palin proposed openstack/nova master: nova diagnostics command is not working with all interfaces https://review.openstack.org/648123 | 16:08 |
*** priteau has joined #openstack-nova | 16:13 | |
stephenfin | kaboom | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove cells v1 jobs https://review.openstack.org/651289 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'nova-cells' service https://review.openstack.org/651290 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove '/os-cells' REST APIs https://review.openstack.org/651291 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Stop handling cells v1 in '/os-hypervisors' API https://review.openstack.org/651292 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Stop handling cells v1 in '/os-servers' API https://review.openstack.org/651293 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'nova-manage cell' commands https://review.openstack.org/651294 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Stop handling cells v1 for console authentication https://review.openstack.org/651295 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove old-style cell v1 instance listing https://review.openstack.org/651296 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'bdm_(update_or_create|destroy)_at_top' https://review.openstack.org/651297 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'instance_fault_create_at_top' https://review.openstack.org/651298 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'instance_info_cache_update_at_top' https://review.openstack.org/651299 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'get_keypair_at_top' https://review.openstack.org/651300 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'instance_update_at_top', 'instance_destroy_at_top' https://review.openstack.org/651301 | 16:15 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'instance_update_from_api' https://review.openstack.org/651302 | 16:16 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Stop handling 'update_cells' on 'BandwidthUsage.create' https://review.openstack.org/651303 | 16:16 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Stop handling cells v1 for instance naming https://review.openstack.org/651304 | 16:16 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Stop handling 'cell_name' field on Instance.save() https://review.openstack.org/651305 | 16:16 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove cells code https://review.openstack.org/651306 | 16:16 |
sean-k-mooney | cells v1 removal | 16:16 |
stephenfin | I _think_ all but the last two or three should be passing unit test, functional tests and pep8 | 16:16 |
* stephenfin crosses fingers | 16:16 | |
stephenfin | sean-k-mooney: yuuup | 16:16 |
sean-k-mooney | well the gate is going to be busy for a while | 16:16 |
stephenfin | First time in days that my laptop hasn't sounded like a rocket | 16:17 |
mriedem | stephenfin: are you going to throw that blueprint in the meeting agenda for thursday since it's a specless blueprint? | 16:17 |
stephenfin | mriedem: Good call. Sure | 16:17 |
gmann | mriedem: I would like to keep only first 3 fixes in this spec, hoping that will not be big for review/doc ? - https://review.openstack.org/#/c/603969/9 | 16:17 |
mriedem | because i may or may not have several new edge use cases and features for cells v1 and will block it's removal | 16:18 |
gmann | mriedem: success code change is something need more audit and can be done later ? | 16:18 |
mriedem | gmann: those first 3 are probably ok and straight-forward enough | 16:19 |
gmann | ok. let me update the spec. | 16:19 |
*** phasespace has joined #openstack-nova | 16:25 | |
openstackgerrit | Sylvain Bauza proposed openstack/nova-specs master: Proposes NUMA affinity for vGPUs https://review.openstack.org/650963 | 16:25 |
bauzas | stephenfin: thanks for reviewing https://review.openstack.org/#/c/650963/, just wrote PS2 | 16:25 |
*** tjgresha has joined #openstack-nova | 16:26 | |
tssurya | mordred, mriedem, dansmith: after the back and forths on changing the type of the "locked" response key since all of you are for the seperate key style and no change on type I will just update the spec then to that approach. | 16:33 |
dansmith | tssurya: thanks | 16:34 |
tssurya | separate* | 16:34 |
mordred | ++ | 16:34 |
dansmith | I was shocked to have mordred come in and agree with me | 16:34 |
dansmith | he might not have realized what he was doing though | 16:34 |
dansmith | stephenfin: melwitt bauzas mriedem: I didn't see any discussion about how stephenfin only created five placeholders for the main db this time, but I do think that's probably more than enough and would be good for the future | 16:39 |
bauzas | yeah, I just feel 5 is enough given the history | 16:39 |
stephenfin | dansmith: I _think_ I called it out in the commit message or a comment, so I figured everyone was ok with that | 16:39 |
dansmith | stephenfin: I didn't see it anywhere | 16:40 |
stephenfin | (did it because you mentioned that we'd never used them anyway) | 16:40 |
mordred | dansmith: wait. I change my mind. whatever the opposite things is please | 16:40 |
dansmith | mordred: :) | 16:40 |
stephenfin | dansmith: Yeah, it wasn't as obvious as it should have been but it's there alright https://review.openstack.org/#/c/650964/1/nova/tests/unit/db/test_migrations.py@183 | 16:41 |
dansmith | ack, okay | 16:41 |
* stephenfin 's head hurts from spec reviews, -> home | 16:41 | |
mriedem | he put it in a comment | 16:44 |
mriedem | oh yeah i'm late | 16:44 |
dansmith | I didn't read that deep since it was already merged, just saw the fewer file count | 16:44 |
*** dpawlik has quit IRC | 16:51 | |
*** derekh has quit IRC | 17:00 | |
*** psachin has quit IRC | 17:00 | |
*** dtantsur is now known as dtantsur|afk | 17:03 | |
*** priteau has quit IRC | 17:04 | |
gmann | dansmith: mriedem mordred can we remove 'locked' field in favor of 'locked_by' in https://review.openstack.org/#/c/638629/9/specs/train/approved/add-locked-reason.rst@198 | 17:06 |
dansmith | I'd rather we didn't | 17:06 |
mordred | gmann: it would be nicer to not remove the field | 17:06 |
gmann | it will be change for old user so no strong opinion just a thought. | 17:06 |
dansmith | it's just senseless change for no real gain | 17:06 |
gmann | yeah | 17:06 |
mriedem | pretty sure that was discussed in some of the earlier patch sets | 17:07 |
gmann | ok. | 17:07 |
bauzas | stephenfin: I'm a terrible man, I left you a -1 on https://review.openstack.org/#/c/555081 after you left | 17:09 |
*** tjgresha has quit IRC | 17:10 | |
bauzas | but I feel it's worth discussing on the upgrade path | 17:10 |
openstackgerrit | Ghanshyam Mann proposed openstack/nova-specs master: Spec for API inconsistency cleanup https://review.openstack.org/603969 | 17:10 |
bauzas | I'm in favor of just saying "heh dude, migrate your workloads first before touching anything or your compute will blow up" | 17:10 |
gmann | mriedem: mordred updated the API cleanup spec with only 3 cleanups to target. https://review.openstack.org/603969 | 17:10 |
dansmith | bauzas: um, really? | 17:11 |
dansmith | bauzas: you want cern to have to migrate all their workloads before/during an upgrade? | 17:11 |
bauzas | dansmith: no, that's not what I meant | 17:11 |
bauzas | dansmith: I meant, before changing the opts | 17:12 |
bauzas | if you upgrade and just keep the same conf, you're not impacted | 17:12 |
bauzas | actually, I left another comment saying this | 17:12 |
bauzas | ie. keep existing behaviour in Train exactly like in Stein if untouched | 17:13 |
dansmith | okay let me re-read | 17:13 |
bauzas | dansmith: context being https://review.openstack.org/#/c/555081/22 | 17:13 |
dansmith | right I know :) | 17:13 |
bauzas | https://review.openstack.org/#/c/555081/22/specs/train/approved/cpu-resources.rst@171 is the existing backwards compat concern I have => don't change anything | 17:14 |
mriedem | tssurya: gmann: i left some comments on the locked reason spec regarding filtering and sorting, but i'm not sure why locked is not a valid sort/filter parameter but locked_by is - even though we dont currently expose the latter | 17:14 |
mordred | gmann: all of those fields with OS-EXT prefixes are readonly aren't they? | 17:14 |
bauzas | https://review.openstack.org/#/c/555081/22/specs/train/approved/cpu-resources.rst@181 is the "what happens when you start playing with those new opts" | 17:14 |
tssurya | mriedem: I was just about to push :) | 17:15 |
dansmith | bauzas: okay, so what happens if you do make those changes on a compute node with instances? | 17:15 |
bauzas | oh man, I really thought the locked reason spec was not controversial :) | 17:15 |
tssurya | oh had just added filtering based on the new keys will be allowed | 17:15 |
mordred | gmann: in any case - we expose all of those without the prefix in sdk already - so I'm definitely in favor of that cleanup :) | 17:15 |
mriedem | tssurya: you can't filter on something that's stored in system_metadata | 17:16 |
dansmith | bauzas: as long as the same config means you don't have to move stuff that's okay, I thought you were kindof implying that they would have to move everything at some point though | 17:16 |
bauzas | dansmith: that's my point, in the spec, we say we reshape | 17:16 |
bauzas | when we start using those opts | 17:16 |
tssurya | bauzas: I thought so too! (when I initially proposed it) | 17:16 |
mordred | (although I think I need to add a way to express inside of sdk "this property comes from whichever of OS-EXT-AZ:availability_zone or availability_zone exists" | 17:16 |
bauzas | dansmith: nope, nope, keep same things in Train if config unchanged | 17:16 |
dansmith | bauzas: I would expect someone like CERN to really care about enabling this sort of thing and requiring everyone to move many gigs of data around just so we don't have to reshape is kinda icky | 17:16 |
dansmith | I mean, really icky | 17:17 |
bauzas | dansmith: I totally agree | 17:17 |
tssurya | mriedem: oh yea forgot ! I guess we discussed this too earlier in the spec when we were deciding where to put it | 17:17 |
mriedem | mordred: not necessarily, OS-DCF:diskConfig is on server create | 17:17 |
gmann | mriedem: they are all in response so did not get readonly thing completely | 17:17 |
bauzas | dansmith: that's why I'm in favor of a full service stop if operators start playing with new config opts in Train and have existing instances | 17:17 |
mordred | mriedem: ah - so it is. ok - that one will take a little extra effort - but shouldn't be too bad | 17:17 |
mriedem | gmann: OS-DCF:diskConfig is in server create and resize requests | 17:18 |
mordred | consuming whichever field name happens to be in the response is easy - knowing which field name one is supposed to send to the api is more effort | 17:18 |
mordred | but still doable and fine | 17:18 |
mriedem | i don't know that gmann was thinking about request params, and i just thought of OS-DCF:diskConfig since you saked | 17:19 |
mriedem | *asked | 17:19 |
mriedem | there could be others, it's a crusty api | 17:19 |
gmann | ok, user data is one case? | 17:19 |
dansmith | bauzas: but I'm saying I think that we should be able to provide support for them stopping a compute, tweaking the params to enable, and then restart... with instances. | 17:19 |
mriedem | user_data is not prefixed in the create request | 17:19 |
mordred | mriedem: I was mostly looking at the list in https://review.openstack.org/#/c/603969/10/specs/train/approved/api-consistency-cleanup.rst | 17:19 |
bauzas | dansmith: with a reshape ? | 17:19 |
gmann | OS-DCF:diskConfig will not be changed right | 17:19 |
bauzas | dansmith: sorry if I'm not getting you | 17:19 |
dansmith | bauzas: we can abort on startup if they have specified something that can't be reshaped, requiring them to move some stuff off before doing it, but requiring clearing every single compute node (over time) to turn that on really sucks | 17:20 |
bauzas | dansmith: okay so you're in favor of some automation | 17:20 |
bauzas | lemme reconsider that | 17:20 |
mordred | dansmith: just adopt cloud native design principles and stop writing files to disk. solves all your problems | 17:20 |
dansmith | mordred: heh | 17:21 |
gmann | mriedem: tssurya locked_by in valid sort/filter list was mistake. we can improve that in tssurya spec i think. | 17:21 |
mriedem | mistake because it's a field we don't even expose today? | 17:22 |
tssurya | gmann, mriedem: I will add locked" as a sort/filter key | 17:22 |
mriedem | i know at one point there was a comment in the code near these fields that said "don't use locked anymore" becaues locked_by implied locked | 17:22 |
tssurya | but as for the existing "locked_by" key in the whitelist, isn't that just a refactor of code ? since we don't expose it anyways ? | 17:22 |
gmann | mriedem: yes, we did not expose that and had in valid sort/fitler | 17:23 |
mriedem | tssurya: refactor of code? | 17:23 |
gmann | i tried to check the reason in original change but did not find any rational to add in whitelist | 17:23 |
mriedem | tssurya: as i mentioned in the spec, supporting filtering on locked_by could get weird, | 17:23 |
tssurya | mriedem: no I don't want to support locked_by based filtering as well | 17:24 |
tssurya | its useless IMO | 17:24 |
mriedem | becaues if we expose locked_by=admin or other, but only store admin or owner, then filtering on locked_by=other means we have to do something like "select * from instances where locked_by!='admin'" | 17:24 |
dansmith | filtering in the api on a non-db field isn't *super* terrible, but sorting is | 17:24 |
gmann | this one - https://review.openstack.org/#/c/408571/ | 17:24 |
bauzas | dansmith: just to make sure https://review.openstack.org/#/c/555081/22/specs/train/approved/cpu-resources.rst@725 requires a reshape, right? | 17:24 |
mriedem | dansmith: i don't think sorting is a problem really here | 17:25 |
bauzas | ie. libvirt sees the options be set, raises a ReshapeNeeded, gets the inventories and allocations, move them | 17:25 |
mriedem | filtering would have to be special cased though, and locked_by is in the db | 17:25 |
dansmith | bauzas: properly representing allocations of dedicated cpus as PCPU right? Not even migration will fix those instances, so I would think reshape is the *only* way to fix that, no? | 17:26 |
mriedem | my pizza is getting dangerously cold | 17:27 |
*** mriedem is now known as mriedem_away | 17:27 | |
dansmith | bauzas: unless you depend on scheduler/conductor to generate a new allocation for a properly hierarchical instance I guess, but.. that's just really super expensive when all you want to do is move some numbers around | 17:27 |
bauzas | dansmith: well, in a world before pcpu_shared_set, all instances are seen as using shared set of CPUs, so they need to be VCPU anyway, but the problem is that if we reserve, say CPU1, then we need to arbitratly assign instances using this CPU1 as PCPU | 17:27 |
dansmith | mriedem_away: I know we can filter on locked_by in the db, but I'm saying if we allow filtering on locked and then drop locked in the future, we can continue honoring that behavior from other fields | 17:28 |
bauzas | dansmith: but then there is an allocation ratio problem, since we don't want to oversubscribe on PCPU | 17:28 |
dansmith | bauzas: but L725 is talking about instances that already have dedicated cpus, but represented as VCPU right? | 17:28 |
cdent | efried: I'm moving on to the cross project nova+placement etherpad for my pre-ptg emails. a) cool with that? b) okay with the process I've been following c) anything especially I should or should not skip? | 17:28 |
bauzas | dansmith: yup, but the above example is still valid, nope? | 17:29 |
bauzas | it's just undocumented in the spec | 17:29 |
* dansmith is confused | 17:29 | |
bauzas | that's what I tried to address in https://review.openstack.org/#/c/555081/22/specs/train/approved/cpu-resources.rst@181 | 17:29 |
bauzas | actually, my wife yells at me because I need to go eating melted cheese | 17:30 |
bauzas | I need to disappear | 17:30 |
bauzas | https://en.wikipedia.org/wiki/Vacherin#Mont_d'Or | 17:31 |
gmann | mriedem_away: mordred let me check those fields from request point of view. | 17:32 |
sean-k-mooney | bauzas: stephenfin can ye review https://review.openstack.org/#/c/649882/ its related to both the cpus in placemnt spec and numa in placement | 17:33 |
bauzas | dansmith: left another comment trying to describe my concerns in https://review.openstack.org/#/c/555081/22/specs/train/approved/cpu-resources.rst@181 | 17:37 |
bauzas | hopefully it will help | 17:37 |
bauzas | now I need to bail out | 17:37 |
dansmith | bauzas: I just commented summarizing my view | 17:37 |
dansmith | maybe that will help | 17:37 |
dansmith | go eat cheese | 17:37 |
*** tjgresha has joined #openstack-nova | 17:38 | |
bauzas | dansmith: okay, i'll join later | 17:38 |
*** gmann is now known as gmann_afk | 17:40 | |
efried | cdent: a) Sure, if you're done with nova reviews :P b) Yes, me likey. (Note that I am reading all of them, but only responding when I have something to say; I'm not sure there's positive value in an empty ack.); c) other than things you already mentioned (perhaps those should be re/moved from the main part of the etherpad?) | 17:41 |
jmlowe | cdent: Got a quick question about placement and vgpus | 17:43 |
efried | cdent: also on a) and b), I am likely to hold off on any further responses until tomorrow. Trying to focus solely on nova spec reviews today (though the gods are conspiring to make that challenging). | 17:43 |
*** ralonsoh has quit IRC | 17:44 | |
cdent | a) I clicked on every link on the spec etherpad this morning and left reviews where I felt comfortable. sometimes I did not because my reaction was "oh look, yet more complexity for the five people who care" and didn't think that was too helpful, so moved on to something else, b) cool, c) I'm trying to make sure the etherpad exposes what has thread | 17:44 |
cdent | jmlowe: hola | 17:44 |
cdent | jmlowe: I can try to help, but the specifics of how the vgpus are being managed is not much in my wheelhouse | 17:45 |
*** mriedem_away is now known as mriedem | 17:46 | |
mriedem | tssurya: so maybe a question that you don't want to hear right now, but do we really need to expose that locked_by field? it seems to be causing a lot more unnecessary churn on this spec than is needed | 17:47 |
mriedem | if the point of the spec is just expose a reason field, can we just....do that? | 17:47 |
*** wolverineav has joined #openstack-nova | 17:48 | |
jmlowe | cdent: So we got 6 nodes with 4 nvidia v100's each for a total of 24, the thing that's confusing me is that in the docs the example is "openstack flavor set vgpu_1 --property "resources:VGPU=1"", I can slice these 8 ways, how do I manipulate the resources in placement so that I have 4*8 vgpus? Or do I have a GCE when it comes to resources, filter scheduler, and placement? | 17:48 |
* cdent is digesting | 17:50 | |
jmlowe | seems like a major omission, how do I tell placement how many of some arbitrary thing I have? | 17:50 |
openstackgerrit | Surya Seetharaman proposed openstack/nova-specs master: Support adding the reason behind a server lock https://review.openstack.org/638629 | 17:51 |
cdent | jmlowe: what release are you on these days? | 17:51 |
jmlowe | Rocky | 17:52 |
sean-k-mooney | jmlowe: bauzas will be able to answer but you should not need too | 17:52 |
tssurya | mriedem: just saw your comment here :/ | 17:52 |
mriedem | i left them on the spec as well, you can answer whenever | 17:52 |
*** wolverineav has quit IRC | 17:52 | |
tssurya | hmm, ok if we are going to expose "locked" after all then locked_by is not needed | 17:52 |
*** wolverineav has joined #openstack-nova | 17:53 | |
sean-k-mooney | jmlowe: there are some limitation in what we currently supprot but as of stien we shoudl now have a nested resoucp provider per phyical gpu | 17:53 |
*** tjgresha has quit IRC | 17:53 | |
sean-k-mooney | each phyical gpu will have an inventory of 8 vGPU reources | 17:53 |
jmlowe | any chance this is explained in the Stein admin guide? | 17:53 |
cdent | jmlowe: the issue you've got is that what's possible in Stein is a lot different from what's possible in Rocky | 17:54 |
sean-k-mooney | jmlowe: in rocky i know we did not support multiple vgpus type on the same host but we might have supported multipel pysical gpus | 17:54 |
jmlowe | *looking ahead to stein admin guide | 17:54 |
sean-k-mooney | jmlowe: i think bauzas wrote something up yes but not sure if its in the admin guide | 17:55 |
*** sridharg has quit IRC | 17:55 | |
*** davidsha_ has quit IRC | 17:55 | |
mriedem | tssurya: we already expose locked :) | 17:55 |
jmlowe | ok, this is starting to make more sense now that I understand the limitations of <= Rocky | 17:56 |
mriedem | https://docs.openstack.org/nova/rocky/admin/virtual-gpu.html | 17:56 |
mriedem | that's the vgpu admin guide for rocky | 17:56 |
mriedem | https://docs.openstack.org/nova/stein/admin/virtual-gpu.html#checking-allocations-and-inventories-for-virtual-gpus is new content for stein | 17:57 |
mriedem | the placement CLI examples in there may or may not be useful to you in rocky | 17:57 |
sean-k-mooney | ya so in rocky we supporte 1 gpu type per node but i belive we had supported muplip phyical gpus that were reported in 1 inventory on the compute node RP | 17:57 |
mriedem | the main difference is in stein there are nested providers for VGPU inventory and allocations | 17:57 |
sean-k-mooney | and in stien we reshap them into child resouce providers | 17:58 |
cdent | jmlowe: this may be informative: https://docs.openstack.org/nova/latest/admin/virtual-gpu.html#checking-allocations-and-inventories-for-virtual-gpus | 17:58 |
cdent | oh, sorry, mriedem just did that | 17:58 |
cdent | was off looking. | 17:58 |
* mriedem has it memorized | 17:59 | |
* cdent is being trained to just wait for mriedem to find everything | 17:59 | |
cdent | Is dinner time | 18:00 |
* cdent waves | 18:00 | |
*** cdent has quit IRC | 18:00 | |
*** sidx64 has joined #openstack-nova | 18:00 | |
tssurya | efried: wait wait I am updating that spec to remove that part which you just commented | 18:02 |
efried | tssurya: Okay. Don't worry, I wasn't merging it :) | 18:02 |
efried | you're going to just not display locked_by at all? | 18:02 |
efried | Sigh, I guess that's one way to skin this cat. | 18:03 |
efried | And you would know better than I whether a user would want to report on "locked by me" and/or "locked by not-me". | 18:03 |
tssurya | efried: yea | 18:03 |
tssurya | I am going to kick locked_by | 18:04 |
*** ricolin has quit IRC | 18:04 | |
tssurya | that's causing all the confusion | 18:04 |
efried | And I guess that could be added later, though it would seem like a lot of work to spin a whole new blueprint/spec/microversion just for that. | 18:04 |
tssurya | efried: yea that's why I had intially added it here | 18:04 |
tssurya | but I hope nobody would want locked_by ? | 18:04 |
efried | that's the thing | 18:05 |
tssurya | when the user tries to unlock they would know if it was "locked by me" or not | 18:05 |
efried | like I'm doing maintenance, so I lock a bunch of things, and then my maint is done, so I want to unlock the things. But I forgot to keep track of the ones I locked. So I want a report that says "hey, what did I lock?" | 18:05 |
dansmith | I was told that the way to tell if it's locked by me is to unlock it :P | 18:06 |
efried | yeah. I could script to unlock "everything" and tolerate the failures ^ | 18:06 |
dansmith | which seems silly to me | 18:06 |
efried | that's pretty ugly | 18:06 |
*** sidx64_ has joined #openstack-nova | 18:06 | |
dansmith | because checking policy of an unlock operation during get is not okay for some reason | 18:06 |
*** sidx64 has quit IRC | 18:06 | |
efried | becuase the policy is conditional on the owner of the lock? | 18:07 |
efried | a thing that can't be determined at policy enforcement time? | 18:07 |
dansmith | I dunno | 18:07 |
dansmith | what I want is a boolean that tells me "can I unlock this thing?" | 18:07 |
efried | locked_by == 'owner' | 18:07 |
efried | tssurya: Okay, so I'm with dansmith on this. Who's the dissenter? mriedem? | 18:08 |
dansmith | but if I'm an admin, I can unlock it too | 18:08 |
dansmith | or if the policy is such that I can unlock.. | 18:08 |
dansmith | efried: just to be clear, I'm fine with not exposing locked_by | 18:08 |
mriedem | efried: locked_by won't tell you "hey, what did I lock?" anyway | 18:08 |
efried | Okay, | 18:08 |
efried | can_I_unlock = am_I_admin or locked_by == 'owner' | 18:08 |
mriedem | and owner = the instance is in a project that i'm in | 18:09 |
dansmith | not sure that's enough | 18:09 |
mriedem | i think... | 18:09 |
dansmith | can't I write complex policy to allow a role that can unlock things for other projects? | 18:09 |
dansmith | regardless, I'd really suggest we not get back into this, remove locked_by and let tssurya proceed | 18:10 |
tssurya | dansmith: thanks :) | 18:10 |
tssurya | efried you okay with that ? | 18:10 |
efried | sure | 18:10 |
mriedem | i also really don't want to think about locked_by anymore, so if the goal is add a reason field, let's do that - making locked_by not suck is a whole other can of worms | 18:11 |
mriedem | and likely a separate spec | 18:11 |
efried | I can get on board with that ^ | 18:11 |
efried | I had actually drafted a comment that read a lot like that | 18:11 |
tssurya | ok thanks everyone I am deleting locked_by from my memory | 18:11 |
efried | tssurya: you can't delete it from memory. | 18:11 |
efried | you should put something in Alternatives so we remember we had this conversation. | 18:12 |
mriedem | not persistent memory anyway https://review.openstack.org/#/c/601596/ | 18:12 |
mriedem | hi-o!!!! | 18:12 |
* efried high-five-sikes mriedem | 18:12 | |
efried | tssurya: ...even if it's just a link to the eavesdrop of this conversation | 18:12 |
openstackgerrit | Surya Seetharaman proposed openstack/nova-specs master: Support adding the reason behind a server lock https://review.openstack.org/638629 | 18:13 |
mriedem | i feel like at this point in my day my time would be better served raking up dead grass outside since it's supposed to dump a foot a snow on us tomorrow | 18:14 |
mriedem | *of snow | 18:14 |
*** gmann_afk is now known as gmann | 18:21 | |
*** rpittau is now known as rpittau|afk | 18:21 | |
*** luksky has joined #openstack-nova | 18:22 | |
gmann | tssurya: mriedem efried dansmith can we remove the locked_by from valid sort/filter key also as that is not going to be exposed anyway. | 18:24 |
*** lpetrut has joined #openstack-nova | 18:25 | |
efried | gmann: If it does nothing right now? Then sure, why don't we do that in a totally unrelated patch? | 18:26 |
gmann | efried: that need microversion bump so doing in unrelated patch need another spec and version bump | 18:27 |
efried | gmann: So it *does* do something right now? | 18:27 |
efried | It actually works, you just don't see the locked_by field in your result? | 18:28 |
efried | or it's ignored, but removing it will cause a 400? | 18:28 |
dansmith | anybody want to blow their brains out when I'm done with mine? | 18:28 |
*** lpetrut has quit IRC | 18:29 | |
gmann | efried: yeah it does not show locked_by in results so user will just assume it worked well but no confirmation. in code we do consider that. | 18:30 |
efried | then why remove it? | 18:30 |
gmann | there is no way for user to know what is valid value of locked_by as nova does not expose that at all | 18:30 |
efried | rather than put it off until later and fix it properly | 18:30 |
ganso | Hi efried. Sorry to interrupt. When you have some time, could you please review https://review.openstack.org/#/c/650437/ and https://review.openstack.org/#/c/649421/ ? Thanks in advance | 18:31 |
efried | ganso: Is tomorrow soon enough? I'd like to focus on specs as much as possible today. | 18:31 |
ganso | efried: sure, thank you! | 18:32 |
gmann | ok, if we do something with locked_by later then only fix sort/filter too / that is ok for me. | 18:32 |
efried | ganso: Oh, those are stein. I'm not stable, so I can't really help you :) | 18:32 |
efried | gmann: Okay, I think that's the plan. | 18:32 |
ganso | efried: oh, sorry... I'll look for the stable maintainers | 18:32 |
gmann | +1 | 18:32 |
efried | gmann: Current bp is going to add sorting/filtering by the locked bool. | 18:33 |
efried | iiuc | 18:33 |
efried | but ignore locked_by entirely - not fix it or remove it. | 18:33 |
mriedem | ganso: https://review.openstack.org/#/admin/groups/540,members | 18:34 |
gmann | efried: you mean current spec will start ignoring 'locked_by' from sort/filter ? | 18:34 |
mriedem | ganso: but also, those can't merge yet anyway b/c we haven't GA'ed stien | 18:34 |
mriedem | *stein | 18:34 |
ganso | mriedem: oh right! release is friday, correct? | 18:35 |
gmann | i feel leave it as it is currently. | 18:35 |
mriedem | hence my -W on the bottom patch | 18:35 |
mriedem | ganso: tomorrow i think | 18:35 |
efried | gmann: No, sorry, I mean the current spec will do nothing to change locked_by in any way. | 18:35 |
mriedem | dansmith: grab a rake | 18:35 |
gmann | efried: +1 | 18:35 |
* efried waits to see how dansmith plans to blow his brains out with a rake. | 18:35 | |
*** tjgresha has joined #openstack-nova | 18:36 | |
*** betherly has joined #openstack-nova | 18:38 | |
*** eharney has quit IRC | 18:42 | |
*** betherly has quit IRC | 18:43 | |
*** IvensZambrano has quit IRC | 18:43 | |
*** wolverineav has quit IRC | 18:46 | |
openstackgerrit | Lee Yarwood proposed openstack/nova master: Block swap volume on volumes with >1 rw attachment https://review.openstack.org/572790 | 18:46 |
openstackgerrit | Lee Yarwood proposed openstack/nova master: Keep attach_mode as top-level field in _translate_attachment_ref https://review.openstack.org/574413 | 18:46 |
*** wolverineav has joined #openstack-nova | 18:47 | |
lyarwood | mriedem: ^ hope you don't mind, this came up downstream last week, finally getting back around to it now and noticed some whitespace. | 18:47 |
*** wolverineav has quit IRC | 18:52 | |
*** betherly has joined #openstack-nova | 18:59 | |
*** betherly has quit IRC | 19:04 | |
*** ociuhandu has joined #openstack-nova | 19:04 | |
*** bbowen_ has joined #openstack-nova | 19:06 | |
*** bbowen has quit IRC | 19:09 | |
*** tbachman has quit IRC | 19:15 | |
*** whoami-rajat has quit IRC | 19:17 | |
*** betherly has joined #openstack-nova | 19:20 | |
*** wolverineav has joined #openstack-nova | 19:23 | |
*** wolverineav has quit IRC | 19:23 | |
*** wolverineav has joined #openstack-nova | 19:24 | |
*** betherly has quit IRC | 19:25 | |
*** sidx64_ has quit IRC | 19:27 | |
*** efried is now known as efried_afk | 19:30 | |
tssurya | mriedem, dansmith: thanks for reviewing https://review.openstack.org/#/c/636132/ , sorry didn't get time to respond, will go through it tomorrow. | 19:34 |
dansmith | tssurya: hopefully we can make this one every bit as painful as the locked_by one! | 19:35 |
dansmith | (not) | 19:35 |
*** boxiang has quit IRC | 19:37 | |
*** boxiang has joined #openstack-nova | 19:37 | |
*** eharney has joined #openstack-nova | 19:38 | |
*** tssurya has quit IRC | 19:39 | |
*** dakshina-ilangov has joined #openstack-nova | 19:40 | |
*** betherly has joined #openstack-nova | 19:41 | |
mriedem | lyarwood: fine with me, but you'll have to convince mdbooth since i think it was held up on him, plus this tempest change https://review.openstack.org/#/c/573025/ | 19:41 |
*** bbowen__ has joined #openstack-nova | 19:43 | |
lyarwood | mriedem: ack thanks will do | 19:44 |
*** awaugama has quit IRC | 19:44 | |
mriedem | lyarwood: btw i was told you own this now https://review.openstack.org/#/c/551349/ | 19:44 |
*** betherly has quit IRC | 19:45 | |
lyarwood | mriedem: yeah, I've just not found my way back to it since the last PS | 19:46 |
*** bbowen_ has quit IRC | 19:46 | |
mriedem | let your heart guide you young lee | 19:46 |
lyarwood | mriedem: I'll queue it up for the morning, thanks for the review! | 19:47 |
*** betherly has joined #openstack-nova | 20:01 | |
*** sean-k-mooney has quit IRC | 20:05 | |
*** betherly has quit IRC | 20:06 | |
*** bbowen__ has quit IRC | 20:08 | |
*** sean-k-mooney has joined #openstack-nova | 20:09 | |
*** jackding has quit IRC | 20:11 | |
*** jackding has joined #openstack-nova | 20:12 | |
*** amodi has quit IRC | 20:15 | |
*** betherly has joined #openstack-nova | 20:22 | |
mriedem | so we agree we're not going to sort/filter on locked_reason right | 20:24 |
mriedem | ? | 20:24 |
*** betherly has quit IRC | 20:26 | |
openstackgerrit | Matt Riedemann proposed openstack/nova-specs master: Support adding the reason behind a server lock https://review.openstack.org/638629 | 20:27 |
*** wolverineav has quit IRC | 20:27 | |
gmann | mriedem: right. filter/sort by 'locked' only | 20:28 |
mriedem | efried_afk: well, got one spec approved today | 20:29 |
mriedem | who has the champagne? is dansmith still alive? | 20:29 |
dansmith | heh | 20:31 |
*** pcaruana has quit IRC | 20:31 | |
*** wolverineav has joined #openstack-nova | 20:32 | |
*** pcaruana has joined #openstack-nova | 20:33 | |
*** pcaruana has quit IRC | 20:36 | |
*** wolverineav has quit IRC | 20:37 | |
openstackgerrit | Ghanshyam Mann proposed openstack/nova-specs master: Spec for API inconsistency cleanup https://review.openstack.org/603969 | 20:38 |
gmann | mriedem: updated the readonly info for OS_EXt prefix fields ^^. OS-DCF:diskConfig i propose to change in request also. | 20:38 |
gmann | checked all the extensions fields and listed the one with prefix | 20:38 |
*** pcaruana has joined #openstack-nova | 20:39 | |
*** pcaruana has quit IRC | 20:47 | |
openstackgerrit | Merged openstack/nova-specs master: Support adding the reason behind a server lock https://review.openstack.org/638629 | 20:48 |
*** wolverineav has joined #openstack-nova | 20:50 | |
*** ttsiouts has joined #openstack-nova | 20:55 | |
*** wolverineav has quit IRC | 20:57 | |
*** tbachman has joined #openstack-nova | 20:58 | |
*** wolverineav has joined #openstack-nova | 20:59 | |
*** wolverineav has quit IRC | 21:00 | |
*** wolverineav has joined #openstack-nova | 21:00 | |
*** betherly has joined #openstack-nova | 21:03 | |
*** betherly has quit IRC | 21:08 | |
*** bbowen__ has joined #openstack-nova | 21:09 | |
*** wolverineav has quit IRC | 21:15 | |
*** wolverineav has joined #openstack-nova | 21:16 | |
*** eharney has quit IRC | 21:19 | |
*** wolverineav has quit IRC | 21:21 | |
*** wolverineav has joined #openstack-nova | 21:23 | |
*** efried_afk is now known as efried | 21:25 | |
*** igordc has quit IRC | 21:25 | |
*** wolverineav has quit IRC | 21:27 | |
efried | mriedem: \o/ but a lot got done today. | 21:27 |
efried | and especially leading up to today, with people fleshing their specs. | 21:27 |
*** wolverineav has joined #openstack-nova | 21:28 | |
*** wolverineav has quit IRC | 21:28 | |
*** wolverin_ has joined #openstack-nova | 21:29 | |
*** wolverin_ has quit IRC | 21:31 | |
*** wolverineav has joined #openstack-nova | 21:33 | |
*** wolverineav has quit IRC | 21:38 | |
*** mchlumsky has quit IRC | 21:44 | |
*** betherly has joined #openstack-nova | 21:44 | |
*** betherly has quit IRC | 21:49 | |
*** tesseract has quit IRC | 21:50 | |
*** luksky has quit IRC | 21:50 | |
*** wolverineav has joined #openstack-nova | 21:58 | |
*** awalende has joined #openstack-nova | 22:00 | |
*** slaweq has quit IRC | 22:01 | |
*** awalende has quit IRC | 22:04 | |
*** hongbin has joined #openstack-nova | 22:13 | |
*** slaweq has joined #openstack-nova | 22:14 | |
*** ttsiouts has quit IRC | 22:16 | |
*** ttsiouts has joined #openstack-nova | 22:17 | |
*** slaweq has quit IRC | 22:18 | |
*** ttsiouts has quit IRC | 22:21 | |
*** rcernin has joined #openstack-nova | 22:28 | |
*** jamesdenton has joined #openstack-nova | 22:32 | |
mordred | efried: heh libvirt gorp | 22:37 |
openstackgerrit | Dakshina Ilangovan proposed openstack/nova-specs master: Nova LLC allocation - RMD plugin for RDT CAT https://review.openstack.org/651233 | 22:38 |
*** tkajinam has joined #openstack-nova | 22:53 | |
*** betherly has joined #openstack-nova | 22:57 | |
*** betherly has quit IRC | 23:02 | |
*** betherly has joined #openstack-nova | 23:18 | |
*** betherly has quit IRC | 23:22 | |
alex_xu | mriedem: hah, hope that is cold medicine :) | 23:31 |
*** wolverineav has quit IRC | 23:31 | |
*** mlavalle has quit IRC | 23:32 | |
*** wolverineav has joined #openstack-nova | 23:33 | |
*** tosky has quit IRC | 23:36 | |
*** betherly has joined #openstack-nova | 23:38 | |
*** wolverineav has quit IRC | 23:41 | |
*** betherly has quit IRC | 23:44 | |
*** slaweq has joined #openstack-nova | 23:48 | |
*** hongbin has quit IRC | 23:48 | |
*** dklyle has quit IRC | 23:52 | |
*** slaweq has quit IRC | 23:52 | |
*** betherly has joined #openstack-nova | 23:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!