Friday, 2019-05-03

*** mriedem has quit IRC00:03
*** e0ne has quit IRC00:22
*** dklyle has joined #openstack-placement02:30
*** dklyle has quit IRC02:36
*** efried has joined #openstack-placement03:25
*** e0ne has joined #openstack-placement04:06
*** e0ne has quit IRC05:15
*** altlogbot_3 has quit IRC08:03
*** altlogbot_0 has joined #openstack-placement08:04
*** e0ne has joined #openstack-placement11:55
*** e0ne has quit IRC13:22
*** e0ne has joined #openstack-placement14:05
*** efried has quit IRC14:12
*** efried has joined #openstack-placement14:31
*** mriedem has joined #openstack-placement14:34
*** e0ne has quit IRC14:41
*** cdent has joined #openstack-placement14:44
*** gryf has joined #openstack-placement14:52
*** dklyle has joined #openstack-placement14:56
*** e0ne has joined #openstack-placement14:58
*** dklyle has quit IRC15:01
*** tssurya has joined #openstack-placement15:08
*** sean-k-mooney has joined #openstack-placement15:14
*** sean-k-mooney has quit IRC15:24
*** amodi has joined #openstack-placement15:33
*** takashin has joined #openstack-placement15:43
*** takashin has quit IRC16:13
*** efried has quit IRC16:13
*** amodi has quit IRC16:15
gibi_capecdent: when do you plan to start discussing numa modelling stuff in the placement room?16:16
*** takashin has joined #openstack-placement16:16
*** efried has joined #openstack-placement16:16
*** dklyle has joined #openstack-placement16:20
cdentgibi_cape: when would you like to. We should probably pick a time as getting you and tetsuro in the same room at the same time would be good16:24
cdentgetting efried in the same conversation would proabably also be good so we might want to try to coordinate with the nova timetable16:25
gibi_capecdent: I'm free now until the nova team photo (11:50)16:26
efriedcdent: What's this about? I expect tomorrow afternoon to open up, for what that's worth.16:26
gibi_capeto be clear when I said numa I should said nested magic16:27
cdentefried: tetsuro's message (and your response) on how to do the same_tree implementation16:27
efriedack16:27
cdentif we can nail down a slot tomorrow, that's probably best as on the fly doesn't seem to be working well16:27
cdentgibi_cape and tetsuro are you both around tomorrow?16:32
gibi_capecdent: for me tomorrow 11:00 - 12:00 seems good16:36
gibi_capeto me16:36
cdentgibi_cape: any option to do it after 13:30, as efried might be able to be available then. If not, then that slot could be be made to work16:38
gibi_capecdent: I have to leave around 15:00 tomorrow to catch my plane16:39
gibi_capecdent: so 13:30 - 15:00  works for me16:40
cdentcool, let's do that, thank you16:40
*** klindgren has joined #openstack-placement16:40
*** e0ne has quit IRC16:41
klindgrenHello placement folks.  I have a question about running stein placement with pike level nova && neutron.  What is my likelyhood for success here?16:41
cdentklindgren: one sec16:41
cdentgibi_cape: line 48 on https://etherpad.openstack.org/p/placement-ptg-train16:42
cdentefried, tetsuro, you too ^16:42
cdentklindgren: theoretically it should "just work" because of the way microversions are used in placement16:43
cdentand because of the way placement is found via the service catalog in both nova and neutron16:44
cdentyou'll need to do some tricks in nova to make sure you don't run the nova placement, but mostly that's "don't turn it on"16:44
cdentout of curiousity what's driving you wanting to do that?16:44
klindgrenWe saw some notes in the api-ref about not using micro versions > 1.18 and ones at or greater together on the same placement.16:45
gibi_capecdent: ack thanks16:45
klindgrenSo we use kolla-ansible on the master branch to deploy.  And the merged in a bunch of changes split placement out to its own service for stein16:45
klindgrenIs the main driver.  We could patch it back in with some local patches16:46
klindgrenBut just trying to keep the patch levels down if possible.16:46
klindgrenAnyway just wanted a 'swag' at if it would probably work.  Or if we were signing ourselves up for much misery.16:49
*** amodi has joined #openstack-placement16:51
*** wwriverrat has joined #openstack-placement16:53
efriedcdent: ack, I'll trundle over once we close out deferred topics (I know of at least one)16:53
cdentefried: cool16:53
*** dklyle has quit IRC17:02
*** e0ne has joined #openstack-placement17:02
*** dklyle has joined #openstack-placement17:23
*** dklyle has quit IRC17:29
*** takashin has quit IRC17:51
*** efried has quit IRC17:52
*** amodi has quit IRC17:54
*** efried has joined #openstack-placement17:57
*** takashin has joined #openstack-placement17:58
*** e0ne has quit IRC18:12
*** wwriverrat has quit IRC18:14
*** dklyle has joined #openstack-placement18:28
*** takashin has quit IRC18:28
*** openstackgerrit has joined #openstack-placement18:31
openstackgerritMerged openstack/os-traits master: Dropping the py35 testing  https://review.opendev.org/65465118:31
*** cdent has quit IRC18:32
*** dklyle has quit IRC18:35
*** dklyle has joined #openstack-placement18:53
*** takashin has joined #openstack-placement18:55
*** sean-k-mooney has joined #openstack-placement18:57
*** sean-k-mooney has quit IRC18:58
*** sean-k-mooney has joined #openstack-placement18:58
*** efried has quit IRC19:02
*** efried has joined #openstack-placement19:02
*** klindgren has quit IRC19:10
*** klindgren has joined #openstack-placement19:11
*** dklyle has quit IRC19:20
*** takashin has quit IRC19:21
*** takashin has joined #openstack-placement19:23
*** amodi has joined #openstack-placement19:33
*** cdent has joined #openstack-placement19:33
efriedcdent: Thanks for the summary on nested topics19:33
cdent'nother one coming soon19:33
*** e0ne has joined #openstack-placement19:34
*** Sundar has joined #openstack-placement19:40
*** dklyle has joined #openstack-placement19:52
*** e0ne has quit IRC20:00
*** e0ne has joined #openstack-placement20:06
*** takashin has quit IRC20:28
*** takashin has joined #openstack-placement20:33
efriedmnaser: btw, patch to delete placement from nova is +W, making its way through the gate.20:38
mnaserefried: cool, thanks for heads up.  the work is almost done in OSA world :)20:38
efriedgood deal20:38
openstackgerritChris Dent proposed openstack/osc-placement master: Use PlacementFixture in functional tests  https://review.opendev.org/65193921:05
*** e0ne has quit IRC21:18
*** e0ne has joined #openstack-placement21:25
mriedemcdent: i think klindgren is asking about stein placement with pike nova because they want https://review.opendev.org/#/c/656885/21:30
mriedemklindgren: pretty sure ^ won't work b/c you need at least rocky placement https://docs.openstack.org/placement/latest/placement-api-microversion-history.html#support-multiple-member-of-queryparams21:31
mriedemoh duh,21:31
mriedemso yeah stein placement,21:31
mriedembut pike nova won't know about that version and stuff21:32
klindgrenWell that's one reason.  But the real driver is kolla-ansible recent master merge which removed placement from nova.21:32
mriedemklindgren: i thought kolla-ansible said they supported the upgrade to extracted placement as well?21:32
mriedemanyway, as cdent said, nova talks to placement over microversions so pike nova shouldn't care if placement is newer21:33
mriedemthe other way around would be a problem21:33
klindgrenThey do have the migration, but we are using kolla-ansible on master, with pike for nova/neutron21:33
mriedemdo they stop/kill the placement-in-nova service during the migration?21:34
mriedemso you don't have placement running in two places?21:34
klindgrenThey do.  But once the stein release got cut that got removed, so we are going to have to do that part ourselves.21:35
klindgrenWhich is fine, I just wanted to see if people though stein placement with pike nova/neutron would "just work" or would blow up in my face horribly :-D21:35
klindgrenThe answer appears to be "should just work, probably" :-D21:36
mriedemyup21:36
*** mgagne has quit IRC21:39
*** mugsie has quit IRC21:39
efriedcdent: I just abandoned a bunch of nova patches that touched nova-in-placement. I could probably have been more diligent and tracked down whether any of them were still relevant to the extracted placement code, but I didn't. If you have the time/inclination, you may wish to have a peek.21:42
*** jroll has quit IRC21:43
cdentefried: i'll look but I suspect it's fine21:44
efriedcdent:21:44
efriedhttps://review.opendev.org/47237821:44
efriedhttps://review.opendev.org/55801421:44
efriedhttps://review.opendev.org/56487621:44
efriedhttps://review.opendev.org/57916321:44
efriedhttps://review.opendev.org/59808821:44
cdentthanks21:45
efriedthanks for the followup cdent, appreciate it.21:52
cdentefried: yeah, those are all fine, left notes confirming a couple but otherwise didn't comment if there was nothing to say21:52
efriedsince we don't have +1 on abandoning21:52
*** Sundar has quit IRC22:07
*** jroll has joined #openstack-placement22:08
*** mugsie has joined #openstack-placement22:10
*** mgagne has joined #openstack-placement22:10
openstackgerritChris Dent proposed openstack/placement master: WIP: Run nova-tox-functional-py36 in the placement gate  https://review.opendev.org/65707722:29
*** openstackstatus has quit IRC22:47
*** e0ne has quit IRC22:47
*** egonzalez has quit IRC22:50
*** egonzalez has joined #openstack-placement22:50
*** e0ne has joined #openstack-placement22:50
*** e0ne has quit IRC22:50
*** e0ne has joined #openstack-placement22:51
*** dklyle has quit IRC22:51
*** e0ne has quit IRC22:51
*** e0ne has joined #openstack-placement22:52
*** e0ne has quit IRC22:57
*** e0ne has joined #openstack-placement23:03
*** e0ne has quit IRC23:05
*** e0ne has joined #openstack-placement23:05
*** mriedem has quit IRC23:07
*** takashin has left #openstack-placement23:08
*** e0ne has quit IRC23:08
*** sean-k-mooney has quit IRC23:11
*** amodi has quit IRC23:14
openstackgerritChris Dent proposed openstack/placement master: Run nova-tox-functional-py36 in the placement gate  https://review.opendev.org/65707723:19
*** cdent has quit IRC23:23
*** mriedem has joined #openstack-placement23:29
*** efried has quit IRC23:38

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!