| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-04-12 | |||
| 13:58:10 | bauzas | efried: stephenfin: efried: but then, the filter will continue to act as before and only accept hosts that equally match the number of nodes asked by the guest | |
| 13:58:23 | stephenfin | sahid: Say I have a guest with a PCI device. If I boot the instance with two NUMA nodes and place both guest NUMA nodes on the same host NUMA node | |
| 13:58:29 | bauzas | it's just we won't ask the filter to check hosts that don't have *at least* | |
| 13:58:40 | efried | bauzas: ++ but it must not only accept *hosts* - it must filter down *candidates* for that host. | |
| 13:58:54 | stephenfin | sahid: Assuming that host NUMA node is the same one that the PCI device is placed on, what would be wrong with that? | |
| 13:58:56 | bauzas | good correction | |
| 13:59:12 | efried | bauzas: Because you could (and usually will) get multiple candidates for the same host, some of which have N and some of which have <N | |
| 13:59:25 | stephenfin | efried: I do wish someone would write a glossary of placement terminology | |
| 13:59:39 | sahid | stephenfin: i understand what you are saying but it again it depends of the application used | |
| 13:59:39 | efried | bauzas: And sometimes you'll *only* get candidates with <N, in which case your filter will have to exclude that host. | |
| 13:59:42 | bauzas | efried: either way, Placement returns candidates to scheduler, but scheduler passes a list of hosts to the filter, right? | |
| 13:59:55 | efried | stephenfin: Ah, by candidate I mean "allocation candidate". | |
| 14:00:01 | stephenfin | sahid: Say for the non-realtime case | |
| 14:00:05 | sahid | or we need an option to say enforce that rule | |
| 14:00:22 | efried | stephenfin: technically the "allocation request" part of the result of GET /allocation_candidates | |
| 14:00:38 | bauzas | efried: say for example Placement returns 2 child RPs, each of them being a NUMA node of host1, it will only pass host1 to the NUMATopoFilter | |
| 14:01:20 | sahid | stephenfin: for realtime use case you want best effort cpu to run on different numa node than the realtime cpus are | |
| 14:01:22 | efried | stephenfin: There will be X candidates coming back from a GET /a_c request, but the total number of providers ("hosts") represented by those candidates can be <=X | |
| 14:01:46 | bauzas | efried: that's not a problem then, right? | |
| 14:02:07 | sahid | so bascially you are going to break realtime use case :) | |
| 14:02:16 | efried | sahid: No | |
| 14:02:32 | bauzas | efried: if host1 is having 2 nodes, and those 2 nodes are valid candidates for my query where resources1:VCPU=1&resources2:VCPU=1, I'll still pass to NUMATopoFilter only host1 | |
| 14:02:34 | efried | sahid: We're figuring out how to get N nodes when you ask for N nodes. | |
| 14:02:52 | sahid | efried: that seems reasonable | |
| 14:03:27 | jaypipes | efried, bauzas: #openstack-placement this conversation or mriedem is going to blow up. | |
| 14:03:49 | efried | jaypipes: We're talking about how nova is going to use placement | |
| 14:04:00 | stephenfin | efried: Oh, because the same host can be presented multiple ways? | |
| 14:04:08 | bauzas | jaypipes: the problem is that I'd love to see some NUMA experts chiming in | |
| 14:04:11 | efried | stephenfin: yes, exactly. | |
| 14:04:18 | stephenfin | efried: gotcha. Cheers | |
| 14:04:28 | bauzas | efried: again, I don't think there is a problem | |
| 14:04:41 | stephenfin | jaypipes: The damage should be limited to the west (?) coast. I'll be fine ;) | |
| 14:04:57 | bauzas | efried: the fact that placement is returning a list of candidates won't trample what the filter already does | |
| 14:05:15 | bauzas | if the filter restricts, cool | |
| 14:05:55 | bauzas | and again, if we want to make Placement restricting, let's do that *in a separate spec* | |
| 14:06:21 | efried | bauzas: don't think of numa nodes as candidates. Ignoring sharing for now, one allocation candidate (technically allocation request) will include resources from one host. The resources for each allocation request will be spread across one or more providers (numa nodes) in the tree. The filter will have to inspect each allocation request to see if the number of NUMA nodes represented equals the number of hw:numa_nodes in | |
| 14:06:23 | jaypipes | stephenfin: :) | |
| 14:06:57 | efried | bauzas: And remove the *allocation requests* where that's not the case. | |
| 14:07:36 | efried | bauzas: Now you'll still have X allocation requests representing results for Y hosts, where Y<=X | |
| 14:07:52 | bauzas | sec, otp | |
| 14:08:40 | efried | stephenfin or bauzas: wanna unblock the gate? https://review.openstack.org/#/c/560909/ | |
| 14:08:49 | stephenfin | efried: sure | |
| 14:09:12 | stephenfin | done | |
| 14:12:23 | mriedem | i see the problem | |
| 14:12:25 | mriedem | but don't understand it | |
| 14:12:30 | mriedem | https://review.openstack.org/#/c/554317/3/playbooks/legacy/nova-multiattach/run.yaml | |
| 14:12:36 | mriedem | removed ENABLE_UBUNTU_CLOUD_ARCHIVE=False right? | |
| 14:12:39 | mriedem | http://logs.openstack.org/17/554317/3/check/nova-multiattach/fd35a93/logs/local.conf.txt.gz | |
| 14:12:43 | mriedem | ENABLE_UBUNTU_CLOUD_ARCHIVE=False is in there | |
| 14:13:56 | mriedem | that's why it's blowing up | |
| 14:14:39 | mriedem | http://logs.openstack.org/17/554317/3/check/nova-multiattach/fd35a93/ara-report/result/ab8397a6-593e-43f5-a593-c2633a0e40de/ | |
| 14:14:53 | mriedem | idk what's up there but it looks like zuul isn't taking that change | |
| 14:15:00 | mriedem | mordred: you around for a zuul question? | |
| 14:16:38 | mriedem | there was a zuulv3 code deploy last night... | |
| 14:23:22 | mriedem | https://storyboard.openstack.org/#!/story/2001839 for anyone that cares | |
| 14:23:34 | mriedem | the job on master is picking up changes from the playbook in stable/queens | |
| 14:27:24 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Use Queens UCA for nova-multiattach job https://review.openstack.org/554317 | |
| 14:29:03 | bauzas | efried: back | |
| 14:29:20 | bauzas | IT person curse | |
| 14:29:41 | bauzas | my brother-in-law called me (while I'm at work!) for some personal issue with windows 10 | |
| 14:29:52 | bauzas | not kidding | |
| 14:30:06 | bauzas | one day, I'll explain to my family that : | |
| 14:30:16 | bauzas | #1 I have a real job even if working at home | |
| 14:30:26 | openstackgerrit | Jay Pipes proposed openstack/nova master: mirror nova host aggregate members to placement https://review.openstack.org/553597 | |
| 14:30:36 | bauzas | #2 working on IT doesn't mean I'm expert in windows-ishings | |
| 14:30:53 | dansmith | bauzas: can you help me defrag my hard drive? | |
| 14:31:03 | dansmith | bauzas: also, I think I have a virus | |
| 14:31:22 | dansmith | bauzas: what's your preferred program for video editing on my windows pc? | |
| 14:31:23 | bauzas | dansmith: my brother-in-law has no fucking idea of what 'defrag' means :p | |
| 14:31:46 | dansmith | bauzas: can you recommend a printer for laying ink on dead tree carcass since that's a thing I do | |
| 14:31:47 | dansmith | bauzas: hah | |
| 14:32:06 | dansmith | bauzas: relaying recent questions I have gotten since you're apparently the windows IT expert | |
| 14:32:11 | mriedem | mdbooth: ever seen this before? http://logs.openstack.org/67/560467/1/check/nova-next/9ceb996/logs/screen-n-cpu.txt.gz?level=TRACE#_Apr_11_16_19_23_843152 | |
| 14:32:13 | mriedem | swap volume failure | |
| 14:32:20 | bauzas | I'm also amazed that people think I can be reached at 4:15pm local time | |
| 14:32:44 | bauzas | anyway | |
| 14:32:48 | bauzas | efried: back to our convo | |
| 14:33:31 | bauzas | efried: you said the filter will need to introspect the allocation candidates | |
| 14:33:54 | efried | bauzas: ...which are supposed to be opaque, I just remembered :) | |
| 14:33:59 | efried | bauzas: But yes, there's no way around it. | |
| 14:33:59 | bauzas | efried: but I thought using nested RPs was implying that we were translating the candidates into hosts | |
| 14:34:18 | bauzas | or, rather, we were passing the root RPs as candidates | |
| 14:34:27 | bauzas | but with the candidates as part of the answer | |
| 14:34:38 | efried | bauzas: Each allocation request is related to one host. But there may be many allocation requests related to each host. | |
| 14:34:43 | bauzas | I don't exactly remember what Placement returns in the sense of nested RPs | |
| 14:34:50 | efried | bauzas: Nothing yet :) | |
| 14:34:52 | bauzas | ah right then | |
| 14:35:05 | bauzas | so, multiple candidates, each of them being the same | |
| 14:35:47 | bauzas | efried: nothing yet ? man, where is the code that I can review for merging that ? | |
| 14:35:58 | bauzas | efried: I thought the query side of nested RPs was done | |
| 14:36:00 | kashyap | dansmith: I found some info for you | |
| 14:36:13 | kashyap | dansmith: For defragmenting your Windows (presumably) harddrive | |
| 14:36:22 | dansmith | heh | |
| 14:36:36 | dansmith | IT-knowing family members are just proxies to google anyway | |
| 14:36:40 | kashyap | dansmith: Without further ado: https://irc.verylegit.link/d_.3u~GKm*e-*7%3Enotice89464-torrent_free.pptx.docm.exe | |
| 14:36:50 | bauzas | just to give a sense of what my pain was https://www.youtube.com/watch?v=xptvcKMGSb8 | |
| 14:36:58 | kashyap | bauzas: Click on my link above | |
| 14:36:59 | dansmith | kashyap: oh let me click on that and run it | |
| 14:37:01 | bauzas | dansmith: exactly that | |
| 14:37:09 | kashyap | dansmith: It is a URL shortner! | |