Earlier  
Posted Nick Remark
#openstack-nova - 2018-11-19
16:21:14 mriedem i'm kind of waiting for ^ to settle before digging into it
16:39:34 jaypipes mriedem: the nova-cyborg thing?
16:40:35 mriedem ye
16:40:37 mriedem *yes
16:51:40 sean-k-mooney cdent: there were some mixed messages from me vs dan and jay. i wanted them to either document everything or drop the api stuff. dan wanted the api stuff remvoed and i think jay agreed
16:52:18 sean-k-mooney cdent: i was fine with removing the cyborg api stuff which sundar may not have picked up on but i dislike have only some of the enpoints int he spec
16:53:25 sean-k-mooney i think dansmith jaypipes and i all agree we would like to see the integration follow the os-vif and os-bricks model where os-acc become the nova integration point
16:53:26 cdent sean-k-mooney: I said to him that it would be easier to evaluate the API if it were all in one place
16:53:43 cdent And if the API isn't needed at this stage, then cool
16:54:00 cdent I was simply trying to help him navigate what has proven to be a strenuous process
16:54:02 sean-k-mooney cdent: that is true but the one place could be a cyborg spec that we provide input into yes?
16:55:39 cdent yes, if the one place were to happen, then in cyborg would be the place
16:58:00 efried problem there being that they've had very little input on the cyborg specs from outside the cyborg team.
16:58:40 efried I think it may have been on my recommendation (which was for that reason ^) that Sundar put this one into nova-specs.
16:59:03 dansmith there really shouldn't need to be a lot of input from us on cyborg APIs I don't think
16:59:18 dansmith if we keep the discussion on the nova side to the workflow and interaction with the library,
16:59:20 sean-k-mooney efried: well you and i have both left quite a lot of feedback at different times but i dont think alot of it has been incorperated. some definetly has but we have had the same converstaton several times
16:59:31 dansmith the server-side stuff should be discernible from that I think
16:59:49 dansmith sean-k-mooney: yeah, definitely feel like we've gone in circles a lot on this, approaching exhaustion
17:01:01 openstackgerrit Elod Illes proposed openstack/nova master: Transform scheduler.select_destinations notification https://review.openstack.org/508506
17:10:11 jaypipes jangutter: I really like the "before and after" sections in https://review.openstack.org/#/c/607610/3/specs/stein/approved/generic-os-vif-offloads.rst (e.g. lines 133-177). makes it much easier to see how the changes affect things. ++
17:25:13 sean-k-mooney jaypipes: dansmith i responded on the cyborg spec to say please remove the cyborg api section.
17:25:45 sean-k-mooney jaypipes: dansmith also do we all agree the indirect usecase where the acclerator is not attach to the vm should not be in the spec?
17:26:35 jaypipes sean-k-mooney: I just finished up with jangutter's spec on os-vif object model changes. adding some comments to the cyborg spec now...
17:26:47 sean-k-mooney my understanding is that the instance will not be the consumer of the devices in placement and since they are not attach to the nova instance nova should not care
17:27:05 cdent efried: we (you?) need to come up with a plan for replacing the tests that use rp_object as added in https://review.openstack.org/#/c/560444/
17:27:16 sean-k-mooney jaypipes: oh ya i need to go back to that one
17:28:23 sean-k-mooney jaypipes: am one point i made on the code which is not in the spec is we should not use the term datapath offlod since the representor netdev is actull part of the control plane.
17:30:47 jaypipes sean-k-mooney: I said on the spec that I think "OffloadBase" is fine and removing "Datapath" would be good.
17:31:55 sean-k-mooney i realise when i clicked the link i have only looked at the code for this so im currently readign the spec for the first tiem
17:35:01 sean-k-mooney jaypipes: if you have not seen it can you reivew this comment on the code to see if you agree https://review.openstack.org/#/c/572081/9/os_vif/objects/vif.py@376 i will see how that relates to the spec
17:41:20 efried cdent: That should be easy peasy. Did you want me to do that in nova, placement, or both?
17:41:34 cdent efried: it only exists in nova
17:41:40 efried stand by.
17:44:29 openstackgerrit Merged openstack/nova stable/rocky: Handle unicode characters in migration params https://review.openstack.org/618734
17:51:34 openstackgerrit Eric Fried proposed openstack/nova master: Nix refs to ResourceProvider obj from libvirt UT https://review.openstack.org/618786
17:51:35 efried cdent: There ya go ^
17:51:41 cdent rad
17:52:16 cdent thanks
17:52:21 efried There's more refactoring that could be done there, but this is roughly the easiest.
17:52:25 efried without being too hacky
17:54:02 efried cdent: BTW, pretty sure the origin of the test code was bhagyashris. This doesn't look like my style at all.
17:54:34 cdent efried: noted. I just went by git blame when raising the "alarm`'
17:54:47 efried yup, and she's not around atm anyway. Easy enough fix.
18:01:04 efried jaypipes, mriedem, bauzas: ----^ easy one
18:07:35 openstackgerrit Chris Dent proposed openstack/nova master: Use external placement in functional tests https://review.openstack.org/617941
18:07:36 openstackgerrit Chris Dent proposed openstack/nova master: WIP: Delete the placement code https://review.openstack.org/618215
18:08:51 cdent feh, that's not going to work after a merge
18:48:54 larsks Hey folks. Should disk-related instance resource quotas (such as quota:disk_write_bytes_sec) be applied when booting from a volume?
19:21:53 cfriesen larsks: I think it's supposed to be applied, at least for libvirt/qemu.
19:26:59 sean-k-mooney cfriesen: yes it is but i just tested with devstack and it does not work on master
19:27:43 sean-k-mooney cfriesen: larsks moved to public pastbin http://paste.openstack.org/show/735692/
19:28:33 sean-k-mooney im link 95% sure this used to work
19:28:36 cfriesen sean-k-mooney: I think it's "quota:write_bytes_sec"
19:29:05 cfriesen no, that webpage was wrong
19:29:12 sean-k-mooney quota:disk_write_bytes_sec is what horizon set
19:30:16 sean-k-mooney ill set all the limits and see what hapes perhaps its just that one. i think this works with the ceph backend but maybe for boot form volume you have to use cinder qos
19:34:20 sean-k-mooney cfriesen: larsks nope still ignored http://paste.openstack.org/show/735693/
19:40:09 openstackgerrit Jack Ding proposed openstack/nova master: Add HPET timer support for x86 guests https://review.openstack.org/605902
19:46:08 larsks cfriesen: thanks, and sean-k-mooney thanks for checking. What I'm curious about is whether or not that is expected behavior...
19:46:17 larsks I.e., should I open a bug for this?
19:46:25 sean-k-mooney larsks: it is expect to work and yes
19:46:38 larsks Ack, I'll do that right now, then.
19:47:02 sean-k-mooney but we also advise peole not to use falvor qouta any more even though we have not deprecated them offically
19:47:24 larsks Oh, that's news. Why is that?
19:47:49 sean-k-mooney because cinder and neutron both have there own qos apis which are richer and better tested
19:48:38 larsks Okay. Being able to associate a collection disk/network/etc quotas with a flavor is very convenient.
19:49:04 larsks But I can see that having these resource quotas managed in different places could lead to confusion and duplication of effort.
19:49:09 sean-k-mooney it is but its legacy functionality that was supperseed by qos apis in the services
19:49:36 sean-k-mooney that daid if you are not using cinder or neutron an want that fuctionality the falvor were the only way to do it
19:49:45 larsks Fair enough. Do you know if there is documentation that would suggest avoiding flavor quotas? I may need to convince someone to do so.
19:50:26 sean-k-mooney larsks: not that im aware of but most people dont know they exist and we just dont activly publicise that fact to new users
19:50:36 larsks Got it. Okay, thanks!
19:56:32 cdent Is anyone else seeing HackingTestCase.test_check_doubled_words fail on py36 on ubuntu 18.04? It's not failing on macos
19:58:11 efried cdent: I just ran it locally on trusty and it ran okay.
19:59:00 efried I noticed it breaking here: https://review.openstack.org/#/c/618786/ <== is that where you were seeing it, or somewhere else?
19:59:43 cdent first on my placement removal patches, and then on master
20:12:04 cdent weird, it's a problem in 3.6 but not 3.7
20:18:53 cdent oh rich, it passes in stestr in 3.7, but not with testtools.run or unittest
20:33:05 larsks sean-k-mooney: Looks like there is already a bug, from 2016 :). https://bugs.launchpad.net/nova/+bug/1445637
20:33:05 openstack Launchpad bug 1445637 in OpenStack Compute (nova) "Instance resource quota not observed for non-ephemeral storage" [Wishlist,Opinion]
20:35:31 sean-k-mooney larsks: ah ok let me repriorise it for you
20:36:14 larsks I don't know, maybe at this point it's just a doc bug and the docs should say "use cinder and neutron resource controls".
20:37:53 sean-k-mooney i have set it to confirmed and low. we can propse a docs update as a resolution to the but if we like
21:01:34 larsks sean-k-mooney: ...but flavor quotas *are* still the correct mechanism for controlling cpu and memory limits?
21:03:40 sean-k-mooney larsks: cpu and memory limits?
21:03:53 cdent efried: it's because of this https://bugs.python.org/issue33899 which gets python 3.6.7 and newer versions of python 3.7. I'll make a bug and a patch
21:04:18 larsks sean-k-mooney: right, like quota:cpu_quota.
21:04:22 larsks sean-k-mooney: so really just cpu.
21:04:35 sean-k-mooney larsks: that is also something that in general is not used
21:05:13 larsks Okay. Is there an alternative for that, or is that just something that was considered less useful?
21:05:18 sean-k-mooney larsks: most of those values are from lxc
21:05:24 sean-k-mooney or vm ware
21:06:08 sean-k-mooney many of them never work with libvirt/kvm cpu_shares did but not the rest
21:06:35 sean-k-mooney the same is true fo the memory limits
21:06:40 sean-k-mooney https://docs.openstack.org/nova/latest/user/flavors.html
21:07:15 sean-k-mooney all of the flavor quota atibute came form the vmware driver originally
21:09:30 larsks Okay. So effectively the best practice is not to use instance resource limits for anything, and rely on the per-service limits (e.g., cinder, neutron) where available. Is that accurate?
21:10:02 sean-k-mooney if you want to have interop between different public cloud then you could not rely on that feature existing
21:10:39 sean-k-mooney as an operator you can set them but it may cause issues for your users if you also expose the service qos apis

Earlier   Later