| *** spatel has joined #openstack-placement | 00:27 | |
| *** spatel has quit IRC | 03:02 | |
| *** takashin has left #openstack-placement | 03:30 | |
| *** gryf has joined #openstack-placement | 04:25 | |
| *** gryf has quit IRC | 04:38 | |
| *** gryf has joined #openstack-placement | 04:39 | |
| *** gryf is now known as __gryf | 05:04 | |
| *** __gryf is now known as _gryf | 05:04 | |
| *** gryf has joined #openstack-placement | 05:05 | |
| *** tssurya has joined #openstack-placement | 07:16 | |
| *** helenafm has joined #openstack-placement | 07:47 | |
| *** cdent has joined #openstack-placement | 08:26 | |
| *** e0ne has joined #openstack-placement | 09:34 | |
| cdent | gibi: I've got a tiny fix to osc_placement_tree, for using with python3 | 11:04 |
|---|---|---|
| *** mriedem has joined #openstack-placement | 11:52 | |
| gibi | cdent: cool! just create a pull request on github | 11:59 |
| cdent | will do | 11:59 |
| *** artom has joined #openstack-placement | 12:04 | |
| cdent | efried, gibi : what's the expected behavior if you provide overlapping same_subtree params: same_subtree: _COMPUTE,_NIC&same_subtree=_NIC,_PORT1 ? | 13:07 |
| gibi | hm hm, can we simply combine them internally in placement into a single same_subtree param? | 13:16 |
| cdent | I would guess that that's what it is supposed to mean | 13:17 |
| *** artom has quit IRC | 13:21 | |
| efried | cdent: I'm in headless chicken mode, but did you look at the spec? | 14:05 |
| efried | this sounds familiar, like something we addressed there. | 14:05 |
| cdent | not yet, was dipping in here for extra brain | 14:05 |
| cdent | (and notekeeping) | 14:05 |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Add gabbits using a DeepNUMANetworkFixture https://review.opendev.org/676204 | 14:17 |
| efried | cdent: The spec says, "Each grouping is treated independently" | 14:21 |
| cdent | efried: yes, I just read that too, it's also in the microversion history | 14:22 |
| efried | It may be possible to optimize in the case of overlaps | 14:22 |
| efried | because the result should be the same (A,B+B,C == A,B,C) | 14:22 |
| cdent | but I wasn't sure (today, because you gave me your brain disease) what it means | 14:22 |
| cdent | but I've done some testing and it behaves like I'd like it to: it allows you to chain subtrees | 14:23 |
| cdent | sort of | 14:23 |
| cdent | or rather: it might do that, but I suspect my test of being wrong | 14:23 |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Add gabbits using a DeepNUMANetworkFixture https://review.opendev.org/676204 | 14:36 |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Avoid duplicate ProviderSummary in _merge_candidates https://review.opendev.org/675606 | 14:36 |
| *** tssurya has quit IRC | 15:22 | |
| *** artom has joined #openstack-placement | 15:23 | |
| *** helenafm has quit IRC | 15:36 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Add gabbits using a DeepNUMANetworkFixture https://review.opendev.org/676204 | 16:11 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Avoid duplicate ProviderSummary in _merge_candidates https://review.opendev.org/675606 | 16:11 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add rw_ctx.parent_uuid_by_rp_uuid, for clarity https://review.opendev.org/676233 | 16:11 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Use rp.id instead of uuid in _rp_rc_key https://review.opendev.org/676234 | 16:11 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add a rw_ctx.psum_res_by_rp_rc, for clarity https://review.opendev.org/676235 | 16:11 |
| cdent | with the latest batch of code and my standard benchmarking rig I can easily get to 120 concurrent connections. this was not true not all that long ago | 16:39 |
| cdent | efried: i know you're busy, but when you get a moment all those ^ patches above come out of the work to tidy up merge candidates. I tried to stick a useful test on the front, but what I think I ended up doing was making a fixutre that makes the test possible, but missed the test. | 16:40 |
| cdent | there's no rush on any of this. I'm done for the day | 16:40 |
| efried | ack | 16:40 |
| efried | on radar, but down the list vs broken sdk | 16:41 |
| cdent | ack ack | 16:41 |
| *** e0ne has quit IRC | 16:51 | |
| *** cdent has quit IRC | 16:53 | |
| *** e0ne has joined #openstack-placement | 18:23 | |
| *** mriedem has quit IRC | 19:08 | |
| *** mriedem has joined #openstack-placement | 19:12 | |
| *** e0ne has quit IRC | 20:38 | |
| *** artom has quit IRC | 21:32 | |
| *** mriedem has quit IRC | 23:08 | |
| *** takashin has joined #openstack-placement | 23:37 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!