| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2020-02-05 | |||
| 17:47:54 | stephenfin | melwitt: There's a super trivial follow-up to that too, btw, if you'd like to take a look https://review.opendev.org/#/c/705655/ | |
| 17:48:37 | efried | sean-k-mooney (sean-k-mooney1): https://blueprints.launchpad.net/nova/+spec/config-tsc-freq looks like something you would be able to grok. I talked to umbSublime and he's interested in working the implementation. At a glance, it seems like it should be fairly straightforward, but I'd like your opinion if you've got a moment to look. | |
| 17:48:46 | melwitt | stephenfin: will do | |
| 17:49:16 | sean-k-mooney | ya i was ment to write up a spec/blueprint | |
| 17:49:21 | sean-k-mooney | or help them | |
| 17:50:52 | sean-k-mooney | efried: the propoasal was to basically to add 2 extra specs one for the tsc frequency and the other for enableing invtsc | |
| 17:51:56 | sean-k-mooney | also add a trait to land on host that support invtsc | |
| 17:52:17 | efried | sean-k-mooney: cool. This would be umbSublime's first contribution, so he'll need some mentoring... | |
| 17:53:37 | sean-k-mooney | looking at my browser history we created 1 an i think i was ment to help draft the update to the bluepirnt. | |
| 17:53:41 | sean-k-mooney | https://etherpad.openstack.org/p/invtsc | |
| 17:53:56 | sean-k-mooney | is umbSublime about | |
| 17:54:18 | efried | That etherpad looks like the same content as what's in the bp description. | |
| 17:54:21 | sean-k-mooney | my isp decied to force recet my router again and broke my internet | |
| 17:54:23 | sean-k-mooney | yes | |
| 17:54:37 | sean-k-mooney | so i obvioulsy never wrote the updated draft | |
| 17:54:40 | efried | from what I could glean, "we" decided at some point that the blueprint could be specless. | |
| 17:54:54 | sean-k-mooney | yes i brought it up in the nova meeting | |
| 17:55:15 | sean-k-mooney | and we said if we spell out the minor changes need then it could be specless | |
| 17:55:22 | efried | great | |
| 17:56:52 | sean-k-mooney | umbSublime: this sliped off my radar sorry. | |
| 17:57:41 | umbSublime | reporting in o/ | |
| 17:58:04 | sean-k-mooney | umbSublime: if you are still up for working on the code cahgne ill try and draft some worth in the ether path to describe what is needed. if you are happy with tem we can update the blueprint and i can help you with reviewing your patch | |
| 17:58:57 | umbSublime | Sounds good to me! I don't think I'll have much time to work on this this week, but next week should be fine! | |
| 17:59:34 | umbSublime | sean-k-mooney, in which etherpad would that be ? | |
| 17:59:56 | efried | sean-k-mooney, umbSublime: we'll want something in approvable form by next week (spec freeze). | |
| 18:00:08 | sean-k-mooney | https://etherpad.openstack.org/p/invtsc | |
| 18:00:15 | umbSublime | efried, I see | |
| 18:00:30 | sean-k-mooney | umbSublime: so looking at my private irc logs from talking to you in novemeber | |
| 18:01:05 | sean-k-mooney | i think the proposal was to add a hw:tsc_freq_mhz flavor extra spec | |
| 18:01:40 | sean-k-mooney | and a hw:invtsc=on|off extra spec | |
| 18:02:12 | sean-k-mooney | that would be how you requested the feature i the flavor | |
| 18:02:41 | umbSublime | yes iirc that's we had discussed last time. I'll have to read back my notes to | |
| 18:02:58 | sean-k-mooney | form a shcduler point of view we would need a new trait for invtsc support | |
| 18:03:17 | sean-k-mooney | which would be auto added if you enabled the flavor extra spec | |
| 18:03:41 | sean-k-mooney | and then the last bit was just the libvirt driver chagne to add the xml elements and report support | |
| 18:04:43 | mriosfer | Hi , where i can find a suported list of gpu supported by queens? | |
| 18:05:01 | mriosfer | sean: looks like hw:video_ram dont report vram to windows for some reason :S | |
| 18:05:07 | sean-k-mooney | umbSublime: do you want to take go at writing that up? or shall i? | |
| 18:05:42 | sean-k-mooney | umbSublime: i also linked https://review.opendev.org/#/c/671338/ to you as an example of how to add the feature just to refresh your memory | |
| 18:06:22 | sean-k-mooney | mriosfer: so in the image its hw_video_ram and you need to set hw_video:max_ram in the flavor i think | |
| 18:06:38 | mriosfer | yes i got it | |
| 18:06:45 | mriosfer | im going to double check | |
| 18:06:50 | umbSublime | Considering I'm not really familiar with the codebase yet, I think it would help me if you'd point me in the right direction (right I had forgotten that link). I'll read on this and take a closer look this afternoon. I'll report back if have any questions/concerns | |
| 18:08:49 | sean-k-mooney | umbSublime: ok i wont get to it today but ill write a rough draft of the chagne in the etherpad tomorrow. then we can update the blue print and review it with the nova team | |
| 18:09:35 | sean-k-mooney | umbSublime: would you be able to attend the nova team meeting tommorow in case people have questions about your usecase | |
| 18:09:39 | umbSublime | Thank you so much, I hope this isn't too much of a burden :/ | |
| 18:09:47 | umbSublime | Yes I should be available tomorrow | |
| 18:10:55 | sean-k-mooney | well normally if we had more time its something that we would help to do. its almost the feature proposal deadline so it will need to be approve this week or next so i can help a little more then normal | |
| 18:12:17 | umbSublime | There a few few "newcoming contributer" docs out there for OS, but is there one more specific to nova I should up on ? | |
| 18:12:57 | sean-k-mooney | we do have a nova one im not sure if it differe much form the generic one let me see if i can find it | |
| 18:14:01 | sean-k-mooney | umbSublime: they are here https://docs.openstack.org/nova/latest/contributor/index.html | |
| 18:14:13 | umbSublime | \o/ | |
| 18:15:33 | sean-k-mooney | umbSublime: you might find this one useful https://docs.openstack.org/nova/latest/contributor/how-to-get-involved.html | |
| 18:16:14 | umbSublime | I'll try and get up to speed on this tonight | |
| 18:38:41 | openstackgerrit | Merged openstack/nova master: nova-net: Remove now unnecessary nova-net workaround https://review.opendev.org/702440 | |
| 18:49:39 | openstackgerrit | Eric Fried proposed openstack/nova master: WIP: refactor: Do network & accel discovery near volumes https://review.opendev.org/706083 | |
| 18:49:53 | efried | dansmith: Does that ^ refactor look like what you had in mind? | |
| 18:51:21 | dansmith | efried: from a skim, yeah | |
| 18:52:58 | efried | dansmith: If so, my main objections to the preceding patch are addressed. The other bits I noted are still worth doing, but could conceivably be FUP'd if there's motivation not to respin the series. | |
| 18:53:16 | dansmith | ack | |
| 20:07:00 | openstackgerrit | Mykola Yakovliev proposed openstack/nova master: Fix boot_roles in InstanceSystemMetadata https://review.opendev.org/698040 | |
| 20:33:24 | sean-k-mooney | umbSublime: i have updated https://etherpad.openstack.org/p/invtsc with what i think woudl be sufficent for the blueprint. | |
| 20:34:20 | sean-k-mooney | umbSublime: i have not spellchecked it and as efried will attest that is likely required. | |
| 20:35:05 | sean-k-mooney | ill add this to the end of the nova meeting agendra for tomorrow but if you are happy with the content can you update the blueprint. | |
| 20:49:43 | umbSublime | sean-k-mooney, thanks! I did a first pass on spellcheck. I think I'll wait until tomorrow to post it in case I think of something to add in the meantime | |
| 20:55:18 | sean-k-mooney | umbSublime: ya if you want to reframe the problem statement etc feel free | |
| 21:19:10 | spatel | sean-k-mooney: hey | |
| 21:20:21 | spatel | I am seeing strange performance issue, I am running erlang application on openstack instance which is giving me very poor performance but if i run same application on bare metal performance is really good | |
| 21:21:01 | spatel | I have created big VM and allocated all vCPU to instance so only single VM running on compute node.. | |
| 21:21:17 | spatel | trying to understand where is the bottleneck | |
| 21:31:36 | openstackgerrit | Stephen Finucane proposed openstack/nova master: trivial: Fetch 'Service' objects once when building AZs https://review.opendev.org/705653 | |
| 21:31:37 | openstackgerrit | Stephen Finucane proposed openstack/nova master: trivial: Merge unnecessary 'NovaProxyRequestHandlerBase' separation https://review.opendev.org/705655 | |
| 21:31:37 | openstackgerrit | Stephen Finucane proposed openstack/nova master: trivial: Bump minimum version of websockify https://review.opendev.org/705654 | |
| 21:31:38 | openstackgerrit | Stephen Finucane proposed openstack/nova master: tox: Integrate mypy https://review.opendev.org/676208 | |
| 21:31:38 | openstackgerrit | Stephen Finucane proposed openstack/nova master: trivial: Remove 'run_once' helper https://review.opendev.org/705656 | |
| 21:31:39 | openstackgerrit | Stephen Finucane proposed openstack/nova master: mypy: Add nova.cmd, nova.conf, nova.console https://review.opendev.org/705657 | |
| 21:31:39 | openstackgerrit | Stephen Finucane proposed openstack/nova master: mypy: Add type annotations to 'nova.pci' https://review.opendev.org/676209 | |
| 21:31:40 | openstackgerrit | Stephen Finucane proposed openstack/nova master: WIP: mypy: Add type annotations to top-level modules https://review.opendev.org/705658 | |
| 21:32:42 | sean-k-mooney | spatel: erlang is pretty memory sensitive so using hugepages in the gust and likely on the host for the vm should help | |
| 21:33:20 | sean-k-mooney | spatel: you shoudl also look into http://erlang.org/doc/man/HiPE_app.html | |
| 21:35:15 | openstackgerrit | Merged openstack/nova master: Handle neutron without the fip-port-details extension https://review.opendev.org/705760 | |
| 21:38:42 | spatel | sean-k-mooney: i am using hugepage | |
| 21:38:52 | spatel | I have tried all possible best practices | |
| 21:39:56 | spatel | earlier i gave all CPU to vm (32 core) so thinking i should just create VM with 16 vCPU so it stay one side of NUMA zone to see how its performing.. | |
| 21:42:30 | sean-k-mooney | ya numa affinitising the vm might help or create a dual numa guest | |
| 21:42:56 | sean-k-mooney | that would report the numa affintiy to erlang but if you are using hugepags it has a numa toplogy of 1 already | |
| 21:43:03 | spatel | I am only curious if i run same benchmark on bare metal then i get good result but if i run on VM then it gives poor result | |
| 21:43:16 | sean-k-mooney | what it could be is the cpu model | |
| 21:43:34 | sean-k-mooney | if you are not useing host-passthouhg you might not have all the cpu feature in the vm that you have in the host | |
| 21:43:36 | spatel | my CPU mode is passthrough | |
| 21:43:58 | sean-k-mooney | well if you have it numa affined and cpu passthohg | |
| 21:44:03 | sean-k-mooney | are you using cpu pinning | |
| 21:44:06 | spatel | Yes | |
| 21:44:11 | spatel | CPU pinning | |
| 21:44:27 | sean-k-mooney | then the only bottelnecs you have left are disk io or netwrok | |
| 21:44:52 | sean-k-mooney | what disk cache mode are you using | |
| 21:44:53 | spatel | I am using SR-IOV and not seeing any single packetloss or performance issue on network layer | |
| 21:45:08 | spatel | We have SSD | |
| 21:45:15 | spatel | not sure about disk cache mode? | |
| 21:45:47 | sean-k-mooney | https://docs.openstack.org/nova/latest/configuration/config.html#libvirt.disk_cachemodes | |