Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-10-05
15:28:21 johnthetubaguy mriedem: its just a public flavor and a private flavor, mapping to different groups of hosts
15:28:36 mriedem johnthetubaguy: so flavor-access with host aggregates?
15:30:01 dansmith mriedem: I'm not sure what your point is? that you can accomplish the same thing with aggregates and a private flavor?
15:30:17 mriedem yeah
15:30:29 dansmith but you can't right? because aggregates are around services not compute nodes
15:30:31 dansmith and
15:30:35 johnthetubaguy mriedem: well, ironic and host aggregates doesn't really do it
15:30:43 dansmith that's just one use of the general traits stuff being described here,
15:30:56 dansmith which also helps with external services using the same ironic
15:31:03 dansmith like some dedicated to nova and some for direct use
15:31:08 dansmith and nova aggregates wouldn't help with that
15:31:57 johnthetubaguy ++
15:32:58 mriedem ok i figured the host aggregates thing screwed it up
15:33:17 mriedem btw, at some point, can someone that understands the issues with host aggregates and ironic put some words about that in a section in https://docs.openstack.org/nova/pike/user/aggregates.html ?
15:33:24 mriedem because i think it's purely tribal knowledge
15:33:24 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/nova-specs master: Network bandwitdh resource provider https://review.openstack.org/502306
15:38:26 jaypipes ralonsoh: I'm going to have sean-k-mooney reach over and slap you if you don't fix that "bandwitdh" in your commit message :P
15:38:45 ralonsoh jaypipes: sorry!!!
15:38:48 ralonsoh one sec...
15:38:53 jaypipes ralonsoh: lol, it's ok :)
15:40:27 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/nova-specs master: Network bandwidth resource provider https://review.openstack.org/502306
15:40:33 jaypipes \o/
15:43:18 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/nova-specs master: Enable SR-IOV NIC offload feature discovery https://review.openstack.org/504895
15:43:40 mriedem easy functional test race fix if someone wants to +W https://review.openstack.org/#/c/509759/
15:44:05 dansmith oh this is the one I was just asking edleafe about
15:44:15 mriedem shelve tests?
15:44:21 efried mriedem Would a rename followup to https://review.openstack.org/#/c/507052/ resolve the mismatched bp/spec name issue, or does the old name stick around somehow?
15:44:30 mriedem efried: see the patch on top
15:44:45 efried mriedem Nyak.
15:51:25 openstackgerrit Balazs Gibizer proposed openstack/nova stable/pike: fix unstable shelve offload functional tests https://review.openstack.org/509818
15:51:51 dansmith cdent: fwiw, I agree with you on this: https://review.openstack.org/#/c/495854/4/nova/scheduler/utils.py
15:51:52 dansmith I would have made it a classmethod
15:51:55 dansmith not a big deal of course, but that's what I would have done
15:52:43 cdent <british>cheers
15:54:57 edleafe dansmith: heh, I had a 'from_host_state()' method on the orginal version. Should have left it there. :)
15:55:19 dansmith just my preference.. doesn't matter
15:56:04 dansmith edleafe: out of curiosity, why is this three patches instead of two with the object definition first?
15:56:16 openstackgerrit Matt Riedemann proposed openstack/nova master: Remove metadata/system_metadata filter handling from get_all https://review.openstack.org/509819
15:56:28 dansmith looking now, it seems like if you did the object patch first, you could have squashed the first and third together, which would be less back and forth
16:39:29 openstackgerrit Chris Friesen proposed openstack/nova master: Add validation for cpu_realtime_mask https://review.openstack.org/468203
16:44:57 openstackgerrit Dan Smith proposed openstack/nova master: Pre-create migration object https://review.openstack.org/498950
16:44:58 openstackgerrit Dan Smith proposed openstack/nova master: Refactor resource tracker to account for migration allocations https://review.openstack.org/506419
16:44:58 openstackgerrit Dan Smith proposed openstack/nova master: Revert allocations by migration uuid https://review.openstack.org/498949
16:44:59 openstackgerrit Dan Smith proposed openstack/nova master: Make live migration hold resources with a migration allocation https://review.openstack.org/507638
16:44:59 openstackgerrit Dan Smith proposed openstack/nova master: Make migration uuid hold allocations for migrating instances https://review.openstack.org/506420
16:50:33 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: fix unstable shelve offload functional tests https://review.openstack.org/509818
16:54:46 openstackgerrit Merged openstack/nova master: Add attachment_get call to volume/cinder_api https://review.openstack.org/506805
17:01:21 mriedem melwitt: you should probably look at this one https://review.openstack.org/#/c/506822/
17:01:28 mriedem i think it's ok, but don't want to mess it up
17:04:18 dansmith mriedem: https://review.openstack.org/#/c/498950/16/nova/tests/unit/conductor/tasks/test_migrate.py line 119, I asserted that it chose the migration type correctly
17:04:26 dansmith mriedem: did you mean I didn't check that it detects resizes?
17:09:01 mriedem dansmith: yeah
17:09:12 dansmith okay
17:09:26 mriedem but a follow up is fine
17:41:52 openstackgerrit Dan Smith proposed openstack/nova master: Add test for resize migration type detection https://review.openstack.org/509852
18:13:36 openstackgerrit Merged openstack/nova master: Update RT aggregate map less frequently https://review.openstack.org/489633
18:14:11 openstackgerrit Merged openstack/nova master: Don't shell out to mkdir, use ensure_tree() https://review.openstack.org/492326
18:14:14 dansmith man, having only 80% of nodepool available has a bigger impact on throughput than I would have expected
18:16:24 mriedem hey, they can surely flip the switch back to 0% throughput again
18:16:29 mriedem smith
18:16:29 mriedem if you don't pipe down
18:16:40 dansmith heh
18:17:48 melwitt mriedem: sure, will do
18:40:39 cdent mriedem: is this back in business? https://review.openstack.org/#/c/267587/
18:46:42 mriedem cdent: the spec isn't approved yet
18:46:53 mriedem cdent: the multiattach stuff is all dependent on the new style volume attachment stuff
18:47:40 cdent ah. I got so excited by seeing it laying around.
18:48:03 cdent the feature that failed to live but won’t die
18:49:05 dansmith too fast to live, too young to die
18:49:09 dansmith james dean.
19:02:59 dansmith mriedem: I was just looking at my live migration patch wondering why it wasn't failing a test because it doesn't clean up by migration uuid on rollback
19:03:09 dansmith and then ran tests again to check, which it failed
19:03:15 dansmith confused because I thought this was passing
19:03:24 dansmith found the test that tests this, which I fail
19:03:33 dansmith dug through the path to figure out where we're cleaning up now
19:03:51 dansmith didn't find it, almost filed a bug
19:03:52 dansmith re-read test, found bug reference, found patch that addeded
19:03:58 dansmith remembered I reviewed and approved that
19:04:05 dansmith and that you have a fix on top of it waiting for +W
19:04:29 dansmith so, altogether a pretty productive hour.. total work product output: one +W vote
19:12:59 openstackgerrit OpenStack Proposal Bot proposed openstack/os-vif stable/newton: Updated from global requirements https://review.openstack.org/373293
19:13:12 openstackgerrit Merged openstack/nova master: fix unstable shelve offload functional tests https://review.openstack.org/509759
19:15:16 dansmith jaypipes: this is needed for the migration uuid stuff, if you wanna slap yer +W on it: https://review.openstack.org/#/c/508699/
19:15:24 openstackgerrit Dan Smith proposed openstack/nova master: Revert allocations by migration uuid https://review.openstack.org/498949
19:15:24 openstackgerrit Dan Smith proposed openstack/nova master: Pre-create migration object https://review.openstack.org/498950
19:15:25 openstackgerrit Dan Smith proposed openstack/nova master: Make migration uuid hold allocations for migrating instances https://review.openstack.org/506420
19:15:25 openstackgerrit Dan Smith proposed openstack/nova master: Refactor resource tracker to account for migration allocations https://review.openstack.org/506419
19:15:26 openstackgerrit Dan Smith proposed openstack/nova master: Make live migration hold resources with a migration allocation https://review.openstack.org/507638
19:15:34 jaypipes dansmith: on it.
19:16:02 jaypipes dansmith: the whole series I presume?
19:16:19 dansmith jaypipes: just that linked one is what I want right now
19:16:33 dansmith jaypipes: I was just pushing the series in parallel to begging for your +W
19:16:53 dansmith jaypipes: i.e. https://review.openstack.org/#/c/508699/
19:18:41 mriedem dansmith: heh, well, i did beg for reviews a few times on that one
19:18:54 dansmith mriedem: I know, I had a lot of stuff in my head at the time
19:18:57 mriedem oh wait not that one
19:19:27 mriedem was thinking this one https://review.openstack.org/#/c/507687/
19:19:31 mriedem is that what you meant?
19:19:52 mriedem yeah it is, nvm
19:20:24 mriedem just about done with this epic internal email on scheduler configuration

Earlier   Later