Earlier  
Posted Nick Remark
#openstack-nova - 2019-09-11
14:34:40 sean-k-mooney righ with the special case code it wont
14:34:42 alex_xu sean-k-mooney: we only implicit to create one in libvirt driver
14:34:46 sean-k-mooney so that is why it works
14:34:51 mloza Is there an option in nova for instance timeout to ERROR state when a compute node fails?
14:35:11 sean-k-mooney but was awsering your question about what does hw:numa_nodes=1 do
14:35:36 sean-k-mooney all other implcit numa toploges act exactly like hw:numa_nodes=1
14:35:44 alex_xu yes
14:35:50 sean-k-mooney hw:pmem is not the same as hw:numa_nodes=1
14:36:11 stephenfin okay, I hate the idea of the implicit NUMA affinity when VPMEM is specified working differently from implicit VPMEM for other things (hugepages, CPU pinning), but I agree it doesn't make sense to work that way if we can't provide VPMEM NUMA affinity
14:36:18 sean-k-mooney and we should not imply it is in any docs, comment or release notes
14:36:35 stephenfin I'm going to have a look to see how hard tacking on proper NUMA affinity is
14:37:02 sean-k-mooney stephenfin: its not hard it has been blocked because its not plamcent native.
14:37:04 stephenfin I don't buy the argument that we shouldn't add more things to the NUMATopology object. NUMA in placement isn't ready so we've to use what we got
14:37:14 stephenfin put politely, that's nonsense :)
14:38:03 sean-k-mooney its been nonsense for 3 years but that is not going to change. so we should enbrace doing it the plamcnet way. in U to give mriedem a nickel
14:38:10 stephenfin gimme a half hour to check out how tough this would be to do. If it's terrible, I'll hold my tongue
14:39:00 alex_xu stephenfin: yes, it will be very easy to support numa affinity once we have numa in placement
14:39:09 sean-k-mooney its 1 field in the host numa toplogy blob and a tweek to the numa toplogy filter to match on it in the numa_fit_instace_to_host funciton
14:39:12 alex_xu stephenfin: just like vgpu, the first step no numa affinity
14:39:33 sean-k-mooney alex_xu: its very easy to support it without placmeent but not polically viable
14:39:45 stephenfin vGPU didn't imply an implicit guest NUMA topology though, that's the key difference here
14:40:13 bauzas looks like we side-tracked the discussion
14:40:36 bauzas if that's about providing a NUMA support in placement, that's something we'll do in Unicorn
14:40:54 bauzas because we understand the concerns
14:41:09 alex_xu stephenfin: the imply an implicit guest NUMA and NUMA affinity are different thing
14:41:11 bauzas my PMs can pay for a nickel if that's the problem
14:41:28 sean-k-mooney alex_xu: not currently
14:41:34 sean-k-mooney form a nova point of view
14:41:38 stephenfin alex_xu: only for VPMEM though
14:41:43 sean-k-mooney but they woudl be with vpmem
14:42:04 sean-k-mooney so that break operators understand of what we men when we say there is an implcit numa toplogy
14:42:26 sean-k-mooney alex_xu: we can crate a numa toplogy as a impmeenation detail
14:42:53 sean-k-mooney but we should never telll people we are because they will assume it works the same way as the other cases we do that
14:44:40 sean-k-mooney alex_xu: just to clarify we use teh terms implcit and explcit numa topoyg in downstream docs and in bug reports and customer expect an implict numa toplogy to be the same as and expclti one. that is the only reason i bring up docs/release notes
14:45:23 mriedem here is an idea, defer vpmems since you're still arguing about designs 1 day before FF and just sort it out in Ussuri :)
14:45:31 mriedem rather than defer and promise to fix things later
14:45:54 alex_xu mriedem: come on...
14:46:53 sean-k-mooney as i said lest just follow the spec and use the code as it is
14:46:56 alex_xu mriedem: so that is works a later fix?
14:47:23 sean-k-mooney alex_xu: we would need to change qemu/libvirt to not do what the current code does
14:47:38 mriedem alex_xu: i'm not reviewing the series nor am i following the argument here, but i know when we say "we'll hack it in x and fix it for real in x+1" that x+1 rarely happens if ever (x+5)
14:47:40 sean-k-mooney or require a realy numa toplogy with affity
14:49:57 mriedem bauzas: you know you already have head count to work on numa in placement support for nova in ussuri?
14:50:50 bauzas mriedem: if I don't have other things like internal crisises, then yes
14:50:59 bauzas or escalations if you prefer
14:51:00 sean-k-mooney at this point not having it block so much of our roadmap since we cant do it the non plamcent way i dont see we have any choice but to dedicate headcount to it
14:51:29 mriedem bauzas: so that means no
14:51:41 bauzas sean-k-mooney: yeah and I went out of steam trying to find some other way for vGPU affinity
14:52:16 mriedem let's just not bullshit ourselves here is all i'm saying
14:52:21 bauzas mriedem: nah, it just means I'm telling the truth : we can't continue implementing NUMA related features without having NUMA support in placement
14:52:30 mriedem sure you can!
14:52:36 bauzas and that's something I'll make it clea to who it could be
14:52:39 mriedem stephenfin doesn't see why ont
14:52:40 mriedem *not
14:52:47 stephenfin jaypipes
14:52:54 stephenfin that was the main reason why, in the past :)
14:52:56 sean-k-mooney no not jaypies
14:53:04 mriedem god rest his soul
14:53:08 stephenfin RIP
14:53:08 stephenfin placement will solve <X>
14:53:11 sean-k-mooney we can implemntate we jsut cant get it merged
14:53:50 sean-k-mooney anyone here if his house/partner/dogs were safe after the storm
14:53:53 mriedem so what's the actual decision that's blocking the pmems stuff at this point? whether numa is implied or not?
14:54:27 stephenfin by specifying VPMEM, you get an implicit NUMA topology but that implicit NUMA topology works differently to every other implicit NUMA topology
14:54:39 mriedem differently how
14:54:41 dansmith that's BS
14:54:44 dansmith don't do that
14:54:54 alex_xu dansmith: you said yes in PTG :)
14:54:57 sean-k-mooney it does not give numa affinity fo the cpus and memory
14:55:02 stephenfin normally, each guest NUMA node is mapped to a unique host NUMA node
14:55:07 stephenfin with this, they'll float
14:55:10 sean-k-mooney so its not the same as hw:numa_nodes=1
14:55:15 dansmith alex_xu: sean-k-mooney said I said that too, but I don't agree that I did
14:55:34 stephenfin so the guest has a NUMA topology but it's literally there to work around a quirk in libvirt
14:55:38 dansmith alex_xu: I know that because I'm quite sure I didn't say anything positive about vpmems
14:55:52 sean-k-mooney dansmith: it proably was not you but apprently it was form redhat an i know stephn and i relly hate that idea
14:55:58 alex_xu dansmith: haha
14:56:15 stephenfin I really don't want to blow something out of the water, but I feel by doing this we're really just masking a problem with libvirt (or QEMU)
14:56:35 stephenfin and we're also creating yet another special case that we'll have to clean up later
14:57:04 mriedem so with pmem do you have to have an implicit numa topology?
14:57:12 stephenfin yup, see here
14:57:26 stephenfin https://github.com/libvirt/libvirt/blob/master/src/qemu/qemu_domain.c#L11604-L11615
14:57:28 mriedem i mean, will libvirt fail to boot the guest w/o numa defined in the xml?
14:57:31 stephenfin yup
14:57:32 mriedem ah ok
14:57:34 alex_xu sean-k-mooney: stephenfin ok, let me try to remove that, maybe it is super fast and simple than our discussion :)
14:57:50 mriedem alex_xu: remove what? qemu code?
14:57:58 stephenfin alex_xu: Remove the implicit NUMA topology? Doesn't look like that's an option. See the above link
14:58:03 mriedem that is likely years before it hits LTS
14:58:33 stephenfin libvirt is telling us we need a NUMA topology, so we must give it a NUMA topology
14:58:46 alex_xu stephenfin: add numa topology explicityly
14:58:54 mriedem ok so if qemu requires implicit numa, then our options are hack it and create a unicorn (bad for nova), change qemu and defer the feature (bad for intel), or do numa affinity as usual in nova,
14:58:55 bauzas oh gosh
14:58:55 mriedem right?
14:59:14 stephenfin mriedem: Right. But
14:59:17 sean-k-mooney mriedem: yes
14:59:21 openstackgerrit Eric Fried proposed openstack/nova master: WIP Include error 'details' in dynamic vendordata log https://review.opendev.org/681329
14:59:39 stephenfin the reason they avoided doing it that way was because they wanted to avoid doing more stuff in NUMATopologyFilter
14:59:42 dansmith artom: commented on the top numa functional patch
14:59:46 bauzas ... and all Hatters technically need to bail out for attending some meeting :)

Earlier   Later