| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-01-17 | |||
| 19:44:17 | mriedem | efried: b/c i've never taskflowed? | |
| 19:45:04 | mriedem | learning a new toolset is kind of real low priority when i have a mountain to climb first, but i know | |
| 19:45:12 | efried | I just mean, looking at TaskBase, it's essentially the same as a TaskFlow Task - execute and rollback methods - but we had to write the plumbing for things like doing the actual rollbacks or whatever. | |
| 19:45:39 | mriedem | task flow also has a lot of threading stuff in it though doesn't it? | |
| 19:45:41 | efried | more just wondering why it wasn't used in the first place. I get following momentum | |
| 19:45:50 | efried | you can thread, or not. | |
| 19:45:54 | mriedem | conductor tasks were written long long ago | |
| 19:46:26 | efried | I see TaskBase is from 2015 yeah | |
| 19:47:31 | mriedem | i believe there was also anecdotes (from years ago) when cinder refactored their volume create flow to use taskflow and it was kind of a mess | |
| 19:47:40 | mriedem | *were | |
| 19:48:49 | smcginnis | And then many debates were had on whether to do the work of removing taskflow by those that could no longer figure out what was going on. | |
| 19:48:59 | mriedem | glancing at the taskflow docs, conductor tasks in nova are essentially this https://docs.openstack.org/taskflow/latest/user/patterns.html#module-taskflow.patterns.linear_flow | |
| 19:49:29 | mriedem | yeah, it just seems more complicated than what i need | |
| 19:50:08 | mriedem | i'm sure it's very fancy and all, but fancy is kind of the last thing i need to deal with in the cross-cell resize series | |
| 19:50:40 | mriedem | jroll: ironic has a task engine thing right? does that use taskflow? | |
| 19:50:46 | efried | The fanciest part of it is making the task a class with an execute (and optional rollback) method. You're already doing that, so... | |
| 19:51:16 | mriedem | but but but i might need to raise a pinky when drinking that tea | |
| 19:51:37 | mriedem | anyway, | |
| 19:51:56 | mriedem | what i'd like to do is get an end to end working flow of this (again, since my initial poc with shelve had one) and then iterate on how to fancy it up | |
| 19:51:57 | efried | Yeah, I'm not pushing it, just something I noticed with your *Task patches. | |
| 19:52:00 | mriedem | sure | |
| 19:52:05 | mriedem | i knew you'd bring it up eventually | |
| 19:52:08 | efried | :) | |
| 19:52:09 | mriedem | but it's not what i'm losing sleep over :) | |
| 19:52:15 | efried | no doubt | |
| 19:52:32 | openstackgerrit | Jack Ding proposed openstack/nova master: Flavor extra spec and image properties validation https://review.openstack.org/620706 | |
| 19:52:52 | mriedem | if you're rechecking tempest ssh failures, | |
| 19:53:22 | mriedem | there is a known neutron bug | |
| 19:53:56 | mriedem | https://bugs.launchpad.net/neutron/+bug/1811515 | |
| 19:54:03 | openstack | Launchpad bug 1811515 in neutron "pyroute2.NetNS don't work properly with concurrency in oslo.privsep" [Critical,In progress] - Assigned to Slawek Kaplonski (slaweq) | |
| 19:54:31 | mriedem | luckily https://review.openstack.org/#/c/631275/ | |
| 20:50:16 | melwitt | nova meeting in 10 minutes | |
| 20:55:25 | efried | mriedem: thanks. Some of them are ssh failures, yes. | |
| 21:04:03 | mriedem | melwitt: os-vif was released with that change https://review.openstack.org/#/c/602384/ | |
| 21:04:34 | melwitt | mriedem: ack | |
| 21:05:16 | sean-k-mooney | yes there is a conflict with libvirt | |
| 21:05:29 | sean-k-mooney | os-vif creates teh port then libvirt deletes it and creates it again | |
| 21:05:41 | sean-k-mooney | tests that only check for port active race | |
| 21:05:53 | sean-k-mooney | tetsts taht check the interface is pingable would be fien | |
| 21:06:19 | sean-k-mooney | e.g. if they wait for the vm to be accessabel before sshing then it wont break | |
| 21:07:09 | sean-k-mooney | im looking in to this week | |
| 21:08:59 | mriedem | dansmith: https://bugs.launchpad.net/nova/+bug/1812196 | |
| 21:09:04 | openstack | Launchpad bug 1812196 in OpenStack Compute (nova) "when using templated cells urls, validation fails with create_cell when using {scheme} and {port}" [Undecided,New] | |
| 21:10:12 | dansmith | um | |
| 21:10:28 | dansmith | those shouldn't be parsed by nova-manage | |
| 21:12:18 | dansmith | hrm | |
| 21:12:35 | dansmith | well, that's kindof the problem I guess, but.. hmm | |
| 21:14:27 | dansmith | cripes can this always have been broken with nova-manage? | |
| 21:23:34 | mriedem | i haven't dug into the bug, just noticed it in triage | |
| 21:27:00 | dansmith | yeah, this is a pretty dumb ass mistake.. mah bad | |
| 21:29:47 | openstackgerrit | Dan Smith proposed openstack/nova master: Fix using template cell urls with nova-manage https://review.openstack.org/631649 | |
| 21:29:51 | dansmith | mriedem: ^ we'll want that backported to .. at least rocky, IIRC | |
| 21:41:40 | melwitt | thanks tonyb | |
| 21:43:01 | tonyb | first nova review in $ages | |
| 21:43:26 | melwitt | :) | |
| 21:58:46 | openstackgerrit | Dakshina Ilangovan proposed openstack/nova-specs master: Nova changes to support Local Resource Management that uses Resource Management Daemon https://review.openstack.org/630817 | |
| 22:01:19 | mriedem | dansmith: might have just been a result of https://github.com/openstack/nova/commit/bc6ca87a6a01ecbda926a78d07cadbd4e3ab97e0 not taking into account templated urls | |
| 22:01:27 | mriedem | i'm not sure which came first | |
| 22:01:55 | dansmith | mriedem: nah, mine was after | |
| 22:18:24 | openstackgerrit | Dan Smith proposed openstack/nova master: Fix using template cell urls with nova-manage https://review.openstack.org/631649 | |
| 22:20:06 | mriedem | e-r uncategorized should be fixed, will take time to regen the page | |
| 22:20:09 | mriedem | https://review.openstack.org/#/c/631653/1 | |
| 22:41:23 | mriedem | dansmith: +2 | |
| 22:42:20 | openstackgerrit | Matt Riedemann proposed openstack/nova stable/rocky: Send context.global_id on neutron calls https://review.openstack.org/627603 | |
| 23:10:14 | melwitt | mriedem: looks like we need to be the T specs directory created https://review.openstack.org/630037 (based on seeing someone propose https://review.openstack.org/630817 earlier) | |
| 23:11:32 | mriedem | done | |
| 23:11:58 | melwitt | ty | |
| 23:23:42 | openstackgerrit | Merged openstack/nova-specs master: Create specs directory for Train https://review.openstack.org/630037 | |
| #openstack-nova - 2019-01-18 | |||
| 00:46:14 | openstackgerrit | Merged openstack/nova master: Fix incompatible version handling in BuildRequest https://review.openstack.org/631179 | |
| 01:00:49 | openstackgerrit | Hang Yang proposed openstack/nova master: Fix port dns_name reset https://review.openstack.org/631684 | |
| 01:19:05 | takashin | oomichi: Would you check my comment in https://review.openstack.org/#/c/627282/ ? | |
| 01:39:20 | openstackgerrit | Jack Ding proposed openstack/nova master: Flavor extra spec and image properties validation https://review.openstack.org/620706 | |
| 01:40:12 | openstackgerrit | Jack Ding proposed openstack/nova master: Correct instance port binding for rebuilds https://review.openstack.org/603844 | |
| 01:40:35 | openstackgerrit | Jack Ding proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.openstack.org/621646 | |
| 01:41:01 | openstackgerrit | Jack Ding proposed openstack/nova master: Improve libvirt image and snapshot handling https://review.openstack.org/616692 | |
| 02:32:35 | fanzhang | Hi edleafe, sorry to borther you, I notice that you have some commits to keystoneauth, I'm wondering that do you have any idea to config 'connect_retries' of Adapter in Nova or in Placement? We've met some ('Connection aborted.', BadStatusLine("''",) errors while trying to get allocation candidates, I am not sure adding retries would help. But it seems https://github.com/openstack/keystoneauth/blob/master/keystoneauth1/adapter.py#L46 here | |
| 02:32:35 | fanzhang | we can pass the connect_reties args, but in https://github.com/openstack/keystoneauth/blob/master/keystoneauth1/loading/adapter.py#L205 there are only few args. | |
| 09:00:13 | openstackgerrit | Merged openstack/nova master: Fix overcommit for NUMA-based instances https://review.openstack.org/629281 | |
| 09:00:54 | jangutter | :-O finally! | |
| 09:08:22 | openstackgerrit | Bhagyashri Shewale proposed openstack/os-resource-classes master: Add PCPU standard resource class https://review.openstack.org/631711 | |
| 09:10:51 | openstackgerrit | Lee Yarwood proposed openstack/nova stable/queens: libvirt: Add workaround to cleanup instance dir when using rbd https://review.openstack.org/628726 | |
| 09:12:10 | openstackgerrit | Elod Illes proposed openstack/nova stable/rocky: Fix incompatible version handling in BuildRequest https://review.openstack.org/631731 | |
| 11:04:15 | openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/os-vif master: Import IP implementation modules outside privsep context https://review.openstack.org/629025 | |
| 11:49:52 | openstackgerrit | Daniel Alvarez proposed openstack/os-vif master: Don't set iface-id when creating an OVS port https://review.openstack.org/631782 | |
| 11:52:08 | openstackgerrit | Daniel Alvarez proposed openstack/os-vif master: Don't set iface-id when creating an OVS port https://review.openstack.org/631782 | |
| 12:15:27 | openstackgerrit | Daniel Alvarez proposed openstack/os-vif master: Don't set iface-id when creating an OVS port https://review.openstack.org/631782 | |
| 14:15:55 | edleafe | fanzhang: Sorry, but I haven't worked with keystone in many years. Perhaps someone else here might have more insight into how to debug the BadStatusLine errors. | |
| 14:18:14 | openstackgerrit | Matt Riedemann proposed openstack/nova stable/queens: Send context.global_id on neutron calls https://review.openstack.org/627604 | |
| 14:34:55 | mriedem | i've been seeing this traceback in n-cpu logs in the gate, it seems harmless but is ugly, and i haven't dug into what introduced it http://logs.openstack.org/80/630980/5/check/grenade-py3/49818d8/logs/screen-n-cpu.txt.gz#_Jan_17_08_04_31_907868 | |
| 14:35:01 | mriedem | something trying to hit the db from nova-compute | |
| 14:35:10 | mriedem | w/o going through conductor | |
| 14:36:20 | cdent | ah, interesting, I saw that in the vmware ci logs and assumed it was the vmware ci being set up incorrectly (it is, but apparently not because of that) | |
| 14:39:30 | mriedem | kashyap: https://review.openstack.org/#/q/topic:bp/support-qemu-native-tls-for-live-migration+status:open would be OK except for you have libvirt 4.0.0 as a minimum requirement in a couple of places | |
| 14:39:34 | mriedem | which i'm guessing is a typo | |
| 14:40:15 | kashyap | Meeting hell, will check back shortly here; sorry. | |
| 14:41:11 | mriedem | cdent: looks like it's actually in rocky too... | |
| 14:41:33 | mriedem | i think i know what it is | |
| 14:42:46 | kashyap | mriedem: Damn, indeed :-( it is 4.4.0 | |
| 14:42:58 | mriedem | https://review.openstack.org/#/q/Icddbe4760eaff30e4e13c1e8d3d5d3f489dac3c4 | |
| 14:43:10 | mriedem | dansmith: i think ^ is causing the tracebacks in the n-cpu logs on startup | |