| *** artom has joined #openstack-placement | 00:08 | |
| *** takashin has joined #openstack-placement | 02:57 | |
| *** belmoreira has joined #openstack-placement | 06:19 | |
| *** tssurya has joined #openstack-placement | 07:20 | |
| *** belmoreira has quit IRC | 07:30 | |
| *** helenafm has joined #openstack-placement | 07:41 | |
| *** cdent has joined #openstack-placement | 07:56 | |
| *** ttsiouts has joined #openstack-placement | 08:03 | |
| *** belmoreira has joined #openstack-placement | 08:05 | |
| *** belmoreira has quit IRC | 08:48 | |
| *** belmoreira has joined #openstack-placement | 08:54 | |
| *** tetsuro has joined #openstack-placement | 09:00 | |
| *** takashin has left #openstack-placement | 09:30 | |
| openstackgerrit | Merged openstack/placement master: Add Python 3 Train unit tests https://review.opendev.org/669481 | 09:42 |
|---|---|---|
| *** belmoreira has quit IRC | 09:44 | |
| cdent | stephenfin: I made some comments in here yesterday about the pdf goal: http://eavesdrop.openstack.org/irclogs/%23openstack-placement/%23openstack-placement.2019-07-15.log.html#t2019-07-15T18:45:10 | 10:15 |
| *** tetsuro has quit IRC | 10:16 | |
| cdent | I got a build, but it was messy, required I hit return twice manually to get past some stuff, and produce a vast volume of warnings/errors of which I could make no sense | 10:16 |
| stephenfin | Yup :-\ I've seen the same thing for other projects (mostly oslo - nova still isn't building). I think for this to succeed we're going to need some help from people that know the LaTeX toolchain in some depth | 10:17 |
| *** ttsiouts has quit IRC | 10:18 | |
| cdent | I was hoping to sort of get ahead of the wave a bit, but I think I'm going to put it back on the back burner. | 10:18 |
| * cdent mixes his metaphors | 10:18 | |
| *** ttsiouts has joined #openstack-placement | 10:19 | |
| *** ttsiouts has quit IRC | 10:23 | |
| *** ttsiouts has joined #openstack-placement | 11:22 | |
| *** helenafm has quit IRC | 12:52 | |
| *** helenafm has joined #openstack-placement | 12:53 | |
| cdent | efried, gibi : fun new bug related to resource class cache, for your reading pleasure: https://storyboard.openstack.org/#!/story/2006232 Probably not urgent, but something that may require some thought on the best choice. | 13:16 |
| efried | ack | 13:16 |
| *** helenafm has quit IRC | 13:34 | |
| *** ttsiouts has quit IRC | 13:34 | |
| *** ttsiouts has joined #openstack-placement | 13:35 | |
| *** helenafm has joined #openstack-placement | 13:36 | |
| *** ttsiouts has quit IRC | 13:38 | |
| *** ttsiouts has joined #openstack-placement | 13:38 | |
| efried | cdent: commented. Definitely not urgent IMO. | 14:06 |
| cdent | efried: responded, need to check a thing which might make it more relevant | 14:16 |
| efried | back atcha | 14:29 |
| cdent | double back atcha efried | 14:48 |
| gibi | efried, cdent: replied in the bug | 15:00 |
| efried | cdent: btw, I'm not ignoring your callout in the last pupdate. I started to compose a response, but was afraid of looking like an idiot. | 15:07 |
| efried | cdent: http://paste.openstack.org/raw/754444/ | 15:07 |
| efried | I was going to continue by nominating some people who have boots on the ground, like CERN and/or our infra teams. But it quickly became clear to me that, while those are definitely large deployments that use openstack, there's really no evidence that they're in any way *typical* | 15:08 |
| cdent | neither CERN nor infra care (yet) about nested | 15:09 |
| cdent | so it either needs to be a cyborg driven model, or a network bandwidth driven model, or something in those realms | 15:10 |
| gibi | efried, cdent: I have no real usage data from bandwidth yet as my employeer still on Pike so they cannot use the bandwidth feature yet | 15:19 |
| efried | oh, was "caring right now" a criterion? | 15:20 |
| cdent | I'm assuming we want some form of realism | 15:20 |
| cdent | if not, then either of you two are plenty of capable, as you've both done it tons of times already in your heads | 15:21 |
| cdent | the issue isn't so much that topology be truly real, but rather that it exercises the relevant features | 15:21 |
| efried | yeah, but (at least on my part) it's just guesswork | 15:21 |
| efried | oh, well then, whyncha use that gabbi fixture I made for nested magic? | 15:21 |
| gibi | efried: good point, that is better than nothing | 15:22 |
| efried | it was specifically designed to exercise the features | 15:22 |
| cdent | because a) i'm not sure that's sufficient, b) I'm hoping that someone who is not me will do it because we need that | 15:22 |
| cdent | you know, delegation. According to the project guide I'm supposed to do that more. | 15:22 |
| cdent | and you know I'm a stickler for the rules and all that | 15:22 |
| * cdent goes to confession | 15:23 | |
| gibi | if I had time I would ask from some data from my downstream counterparts about deployed hardware and try to guess how a bandwidth config will look like on those hardwares | 15:23 |
| efried | cdent: and tbc, we're concerned about scale here, so we don't care about e.g. edge deployments with three nodes | 15:23 |
| cdent | correct | 15:24 |
| efried | okay, how bout we ask stx? | 15:24 |
| efried | don't they do big deployments that give a crap about numa? | 15:24 |
| efried | they're the ones driving all the numa topo reporting and live migration stuff atm iirc | 15:25 |
| cdent | the end goal here is to be able to slaughter a placement server with a topology that will causes a_c queries to hit all the corners | 15:25 |
| cdent | efried: that makes sense, but they've also made it clear that they are months or even years behind their upstream | 15:25 |
| efried | that seems less important than the fact that they care | 15:27 |
| efried | unless we can find someone who cares *right now*. | 15:27 |
| cdent | well I'm pretty sure I care less than them or either of you | 15:28 |
| gibi | I don't have time to care right now, sorry | 15:28 |
| efried | I'll poke dtroyer and see if he can offer any insight. | 15:28 |
| cdent | thanks, that's a good idea | 15:28 |
| cdent | efried: i discovered a jsonpath feature that you might find useful in the future: `parent`: https://github.com/cdent/placecat/pull/3/files#diff-ea604bb547ddae3f9a383248673a5c9bR194 | 15:32 |
| *** e0ne has joined #openstack-placement | 15:32 | |
| * cdent ends day | 15:45 | |
| * cdent waves | 15:45 | |
| *** cdent has quit IRC | 15:45 | |
| *** ttsiouts has quit IRC | 15:51 | |
| * gibi waves too | 15:52 | |
| *** ttsiouts has joined #openstack-placement | 15:52 | |
| *** helenafm has quit IRC | 15:52 | |
| *** tssurya has quit IRC | 15:55 | |
| *** ttsiouts has quit IRC | 15:57 | |
| *** atahardjebbar has joined #openstack-placement | 16:43 | |
| *** amodi has quit IRC | 16:48 | |
| *** atahardjebbar has quit IRC | 17:13 | |
| *** e0ne has quit IRC | 17:20 | |
| *** efried has quit IRC | 18:12 | |
| *** efried has joined #openstack-placement | 18:21 | |
| *** amodi has joined #openstack-placement | 21:28 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!