| *** tetsuro has joined #openstack-placement | 00:55 | |
| *** tetsuro has quit IRC | 02:16 | |
| *** tetsuro has joined #openstack-placement | 02:52 | |
| *** tetsuro has quit IRC | 03:28 | |
| *** tetsuro has joined #openstack-placement | 03:46 | |
| openstackgerrit | Merged openstack/placement master: Extra gabbi tests for same_subtree https://review.opendev.org/669930 | 04:42 |
|---|---|---|
| *** belmoreira has joined #openstack-placement | 06:13 | |
| *** artom has quit IRC | 06:32 | |
| *** artom has joined #openstack-placement | 06:33 | |
| *** belmoreira has quit IRC | 07:28 | |
| *** tssurya has joined #openstack-placement | 07:31 | |
| *** cdent has joined #openstack-placement | 07:36 | |
| *** e0ne has joined #openstack-placement | 07:58 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Use TraitCache for Trait.get_by_name https://review.opendev.org/673750 | 07:59 |
| *** belmoreira has joined #openstack-placement | 08:01 | |
| *** e0ne has quit IRC | 08:07 | |
| *** tetsuro has quit IRC | 08:11 | |
| *** ttsiouts has joined #openstack-placement | 08:11 | |
| cdent | every single time I use mock I want to burn the whole world down | 08:12 |
| *** tetsuro has joined #openstack-placement | 08:17 | |
| *** ttsiouts has quit IRC | 08:21 | |
| *** ttsiouts has joined #openstack-placement | 08:22 | |
| *** ttsiouts has quit IRC | 08:27 | |
| *** ttsiouts has joined #openstack-placement | 08:36 | |
| openstackgerrit | pengyuesheng proposed openstack/placement master: Blacklist sphinx 2.1.0 (autodoc bug) https://review.opendev.org/673760 | 08:42 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Optimize trait creation to check existence first https://review.opendev.org/673555 | 08:59 |
| *** belmoreira has quit IRC | 09:03 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add apache benchmark (ab) to end of perfload jobs https://review.opendev.org/673540 | 09:03 |
| *** ttsiouts has quit IRC | 09:05 | |
| *** tetsuro has quit IRC | 09:06 | |
| *** e0ne has joined #openstack-placement | 09:06 | |
| *** ttsiouts has joined #openstack-placement | 09:09 | |
| *** ttsiouts has quit IRC | 09:23 | |
| *** ttsiouts has joined #openstack-placement | 09:24 | |
| *** ttsiouts has quit IRC | 09:28 | |
| openstackgerrit | Merged openstack/osc-placement master: Update api-ref location https://review.opendev.org/672769 | 09:37 |
| *** ttsiouts has joined #openstack-placement | 09:45 | |
| *** ttsiouts has quit IRC | 10:00 | |
| *** ttsiouts has joined #openstack-placement | 10:01 | |
| *** ttsiouts has quit IRC | 10:06 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Run nested-perfload parallel correctly https://review.opendev.org/673505 | 10:19 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Implement a more complex nested-perfload topology https://review.opendev.org/673513 | 10:19 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add apache benchmark (ab) to end of perfload jobs https://review.opendev.org/673540 | 10:19 |
| *** belmoreira has joined #openstack-placement | 10:50 | |
| *** ttsiouts has joined #openstack-placement | 11:23 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Remove ProviderIds namedtuple https://review.opendev.org/673788 | 11:41 |
| *** ttsiouts has quit IRC | 13:25 | |
| *** ttsiouts has joined #openstack-placement | 13:26 | |
| *** mriedem has joined #openstack-placement | 13:27 | |
| *** ttsiouts has quit IRC | 13:31 | |
| *** ttsiouts has joined #openstack-placement | 13:41 | |
| *** belmoreira has quit IRC | 13:43 | |
| *** belmoreira has joined #openstack-placement | 13:45 | |
| *** belmoreira has quit IRC | 13:56 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add apache benchmark (ab) to end of perfload jobs https://review.opendev.org/673540 | 14:13 |
| *** belmoreira has joined #openstack-placement | 14:15 | |
| cdent | belmoreira: have you tweaked innodb_read_io_threads and innodb_write_io_threads on the mysql servers you have behind your placement servers? | 14:20 |
| cdent | efried: was there a reason you didn't +w the sphinx blacklist at https://review.opendev.org/#/c/673760/ ? | 14:21 |
| belmoreira | cdent I running with the default 4 | 14:34 |
| cdent | thanks belmoreira . that seems reasonable in a reasonably spread out set up, but I'm trying to do some performance testing and with just one database it's not seeming to be enough | 14:35 |
| belmoreira | cdent thanks for let me know. I'm interested in your results | 14:42 |
| cdent | there's a related stack of work happening around here: https://review.opendev.org/673505 and I wrote about some of it here: https://anticdent.org/profiling-placement-in-docker.html | 14:43 |
| cdent | the `ab` on the end of that stack of changes is timing out becaue the database isn't keeping up. uwsgi+placement seems to be doing okay, but the db doesn't want to do enough. | 14:44 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Optimize trait creation to check existence first https://review.opendev.org/673555 | 14:46 |
| *** belmoreira has quit IRC | 14:51 | |
| efried | cdent: Yeah, it seems like a slam dunk, but I thought having stephenfin send it would be appropriate, since he was involved in the sphinx issue and he's on the core list. | 14:58 |
| cdent | efried: meh, based on reading the issue and stephenfin's already awareness of it, I'd say slam it. Unless there are substantial reasons to delay things, why delay? | 15:00 |
| *** belmoreira has joined #openstack-placement | 15:03 | |
| efried | cdent: sure, you can +W it if you like. | 15:04 |
| * cdent does | 15:04 | |
| efried | cdent: btw, your opposition to mock is noted, but it exists for exactly this reason. The test as written in PS4 would blow up against the code as written in PS1, but that's the only way (other than review) that that bug would have been found short of a real field repro. | 15:18 |
| cdent | efried: I know, and I'm sure you know I know. My opposition is not about its utility. | 15:18 |
| cdent | its about its necessity | 15:19 |
| cdent | if we write code that needs mocks then we've written code that could be better | 15:19 |
| cdent | but at a certain point you just gotta say screw it and deal | 15:20 |
| efried | interesting. | 15:20 |
| cdent | I know that in a nova-oriented world, the idea of writing code that doesn't need mocks is insanity, but it is possible (and better, in my opinion, which I ack is much different from others) | 15:20 |
| efried | so in this case you would have to use, like, a database lock to avoid the possibility of races. | 15:20 |
| cdent | and, in any case I don't really mind: "bug would have been found short of a real field repro". Bugs are how we make sure a community exists | 15:21 |
| cdent | part of the reason we don't have a community around placement is because it is too damn good :) | 15:21 |
| efried | first world problems | 15:21 |
| efried | or something | 15:22 |
| cdent | quite | 15:22 |
| efried | But I'm sure you've seen the graphs every software shop loves to put in devs' faces | 15:22 |
| efried | about the exponential increase in cost of a bug the later it is discovered. | 15:23 |
| efried | to that end, avoiding field escapes is our job, and in that sense, I *do* mind. | 15:23 |
| efried | (clearly you don't actually object; this is just conversation) | 15:24 |
| cdent | I think the economics of those sorts of graphs are complicated in open source, especially open source which is as socially and economically complex as openstack | 15:25 |
| cdent | to such an extent that I'm not sure we can use standard metrics for in house development health | 15:25 |
| cdent | i mean yeah, if I objected that much, I wouldn't have written the test, but I object enough to moan about it | 15:26 |
| efried | Agree open source is different for sure | 15:26 |
| efried | The slope of the graph may be shallower, but the direction is still the same. | 15:27 |
| cdent | If I had to pick one of the primary factors in why openstack has a bad architecture one of them (but not the sole) would be over-reliance on mocked unit tests | 15:27 |
| efried | because downstreams still have to package etc. | 15:27 |
| cdent | oh don't get me started on packaging :) | 15:27 |
| efried | oh, I have no problem with the philosophy of "only use mock and unit tests for things you can't (practically) cover with functional tests" | 15:28 |
| efried | and e.g. nova has been moving in that direction | 15:28 |
| cdent | it's kind of too late for nova | 15:28 |
| cdent | i mean we can do a better job of finding and fixing bugs with the new style, but changing the architecture, though possible, is very slow | 15:29 |
| cdent | we're making progress | 15:29 |
| cdent | it's good | 15:29 |
| cdent | i'm not trying to be a totally negative ninny | 15:29 |
| cdent | but let's just say I have thoughts | 15:29 |
| cdent | welp, I was hoping the gate was going to give me something to read before I ran out of time, but no. | 15:34 |
| cdent | until tomorrow | 15:34 |
| * cdent waves | 15:34 | |
| efried | o/ | 15:34 |
| *** cdent has quit IRC | 15:34 | |
| *** ttsiouts has quit IRC | 15:42 | |
| *** ttsiouts has joined #openstack-placement | 15:42 | |
| *** tssurya has quit IRC | 15:46 | |
| *** ttsiouts has quit IRC | 15:47 | |
| *** e0ne has quit IRC | 15:49 | |
| *** e0ne has joined #openstack-placement | 15:50 | |
| *** belmoreira has quit IRC | 15:50 | |
| *** efried is now known as efried_rollin | 15:57 | |
| openstackgerrit | Merged openstack/placement master: Blacklist sphinx 2.1.0 (autodoc bug) https://review.opendev.org/673760 | 16:45 |
| *** e0ne has quit IRC | 17:18 | |
| *** altlogbot_2 has quit IRC | 18:19 | |
| *** altlogbot_1 has joined #openstack-placement | 18:20 | |
| *** e0ne has joined #openstack-placement | 18:37 | |
| *** e0ne has quit IRC | 19:02 | |
| *** e0ne has joined #openstack-placement | 19:12 | |
| openstackgerrit | melanie witt proposed openstack/osc-placement master: Add --amend option to 'resource provider inventory set' https://review.opendev.org/640898 | 19:21 |
| *** e0ne has quit IRC | 19:24 | |
| *** efried_rollin is now known as efried | 19:47 | |
| *** e0ne has joined #openstack-placement | 19:55 | |
| *** e0ne has quit IRC | 20:12 | |
| *** mriedem has quit IRC | 20:44 | |
| *** mriedem has joined #openstack-placement | 20:45 | |
| *** mriedem has quit IRC | 20:54 | |
| *** mriedem has joined #openstack-placement | 20:55 | |
| *** e0ne has joined #openstack-placement | 21:37 | |
| *** mriedem has quit IRC | 22:43 | |
| *** e0ne has quit IRC | 23:33 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!