| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-03-06 | |||
| 14:46:55 | gibi | stephenfin: cool :) | |
| 14:48:54 | gibi | sahid: thanks for updating the rxt/tx spec. I left some questions inline https://review.openstack.org/#/c/539605/ | |
| 14:49:21 | stephenfin | gibi: Got the bugger (it was websockify). Ta :) | |
| 14:49:47 | gibi | stephenfin: :) | |
| 14:50:19 | openstackgerrit | Matt Riedemann proposed openstack/nova-specs master: Add extra-specs to the flavor show and detail API calls https://review.openstack.org/544158 | |
| 14:54:11 | mriedem | sahid: i wonder if you'd have any ideas about this bug https://bugs.launchpad.net/cinder/+bug/1753394 | |
| 14:54:12 | openstack | Launchpad bug 1753394 in OpenStack Compute (nova) "block_device_mapping_v2 cannot work with Pike" [Undecided,New] | |
| 14:54:26 | mriedem | lyarwood: or you | |
| 14:56:25 | odyssey4me | mriedem Not sure if you've seen this, but I expect that anyone using upgrade_levels/compute=auto will find that the upgrades fail. We have a patch up - I wonder if we could discuss whether that's appropriate or whether there's an alternative we should explore? https://review.openstack.org/#/c/549737/1 | |
| 14:57:39 | mriedem | odyssey4me: dansmith should take a look | |
| 14:58:07 | mriedem | sahid: my guess in that bug is something in https://github.com/openstack/nova/commit/c25629f85feb53b5be0347f68c43b3b55fb9f137 caused a regression | |
| 15:02:01 | sahid | mriedem: yes that seems to be related, i can't really figure that out right now, but i will try to look at it soon | |
| 15:02:09 | mriedem | odyssey4me: that's kind of weird because grenade sets upgrade_levels=auto https://github.com/openstack-dev/grenade/blob/master/projects/60_nova/upgrade.sh#L77 | |
| 15:02:17 | mriedem | sahid: ok, thanks | |
| 15:02:47 | odyssey4me | mriedem I think, in this case, we're seeing failures not on upgrade - but instead on a greenfield deploy | |
| 15:02:51 | mriedem | odyssey4me: were you upgrading from pike to queens? | |
| 15:02:59 | mriedem | odyssey4me: oh... | |
| 15:03:12 | odyssey4me | so this would be a rocky deploy, but with upgrade_levels=auto in the conf file | |
| 15:03:17 | mriedem | odyssey4me: ok so if i set that to auto in just devstack everything should explode | |
| 15:03:31 | odyssey4me | that's what we're seeing in our own tests, yes | |
| 15:03:51 | dansmith | odyssey4me: mriedem I'm commenting | |
| 15:03:53 | odyssey4me | lemme find you an example | |
| 15:04:53 | evrardjp | dansmith: mriedem thanks for your time there. | |
| 15:05:33 | odyssey4me | dansmith mriedem here's the conductor log where it explodes: http://logs.openstack.org/72/547072/6/check/openstack-ansible-functional-ubuntu-xenial/aa76783/logs/openstack/openstack1/nova/nova-conductor.log.txt.gz#_2018-03-05_18_07_21_754 | |
| 15:10:57 | odyssey4me | dansmith your comment makes sense I think, thanks - we may need some hand-holding to get that right, I'll defer to git-harry via review to ask for help if needed. | |
| 15:11:08 | dansmith | odyssey4me: ack | |
| 15:11:12 | odyssey4me | Thanks for the quick review! | |
| 15:11:15 | dansmith | np | |
| 15:13:21 | mriedem | dansmith: i left a question in there for clarification | |
| 15:14:28 | bauzas | disclaimer : I'm not full steam yet, in case people want to ping me some reviews | |
| 15:14:46 | bauzas | that said, if you'd like to see my comments, tell me it | |
| 15:22:28 | Spazmotic | Hey jaypipes, getting ready to head to bed, but just wanted to let you know I updated https://review.openstack.org/#/c/538415/ if you get some time this week. I'm sure you're pretty busy with work coming back from PTG :) | |
| 15:22:45 | Spazmotic | And with that, have a great night novaers.. i've been up way too long | |
| 15:22:49 | jaypipes | Spazmotic: yup, will review this afternoon. | |
| 15:22:53 | jaypipes | thx for th eping | |
| 15:23:01 | Spazmotic | No worries if you can't find the time sir. | |
| 15:23:04 | Spazmotic | Have a great day. | |
| 15:23:07 | jaypipes | ciao | |
| 15:23:55 | mriedem | Spazmotic: why isn't there a test for that bug fix? | |
| 15:24:53 | Spazmotic | I've put a fair amount of thought into that and was trying to think of how to test for a XenAPI fault being raised in such a specific situation. | |
| 15:26:12 | Spazmotic | Or in the case of testing that the assert did not run, ensureing it did not | |
| 15:26:21 | Spazmotic | ensuring* | |
| 15:27:02 | Spazmotic | Would love to hear your thoughts on it though. My mocking logic is not fantastic | |
| 15:27:12 | mriedem | maybe jianghuaw_ can help? | |
| 15:27:23 | Spazmotic | Yeah i've been wanting to sync back up with him since he got back. | |
| 15:33:13 | Spazmotic | The error thrown without the skip in place just comes from XenAPI and in a tight circumstance, not really a code error, so am still thinking of a good test condition I suppose. | |
| 15:34:36 | Spazmotic | And I suppose the real test should be to test the new feature of the skip in place, so not sure how to test that a method call that only asserts wasn't run hehe | |
| 15:37:21 | Spazmotic | Actually.. I think I Know.. will make a note | |
| 15:37:27 | mriedem | well, seems pretty clear to me, | |
| 15:37:35 | mriedem | test would raise an exception from _call_live_migrate_command, | |
| 15:37:47 | mriedem | the new conditional would eval to true, | |
| 15:38:00 | mriedem | and the test would assert that the exception wasn't re-raised | |
| 15:38:42 | Spazmotic | That was my fault as well just now.. not sure why it didn't appear to me when I was thinking.. just side_effect the method and ensure the XCP versions are passed so it's skiped and not re-raised. | |
| 15:38:47 | Spazmotic | thought*.. ugh.. tired. | |
| 15:39:14 | Spazmotic | I'll write it up after some sleep mriedem, thanks. | |
| 15:39:39 | mriedem | right, i left a comment as a reminder | |
| 15:39:41 | mriedem | get some sleep | |
| 15:39:45 | Spazmotic | Thanks much | |
| 15:40:41 | dansmith | jaypipes: around? | |
| 15:45:56 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Drop API compat handling for old compute error cases https://review.openstack.org/546256 | |
| 15:46:04 | cdent | biab | |
| 15:50:23 | openstackgerrit | Merged openstack/nova stable/queens: Handle spawning error on unshelving https://review.openstack.org/548415 | |
| 15:50:41 | openstackgerrit | Eric Berglund proposed openstack/nova master: PowerVM Driver: Snapshot https://review.openstack.org/543023 | |
| 15:51:45 | jaypipes | dansmith: yessir | |
| 15:52:45 | dansmith | jaypipes: is there an inexpensive way I can count all rows in a big table? like, if I do count(indexed_column), is that heavy or light? | |
| 15:53:20 | openstackgerrit | Stephen Finucane proposed openstack/nova master: conf: Deprecate 'keymap' options https://review.openstack.org/483994 | |
| 15:53:36 | stephenfin | mriedem: I've validated that noVNC 1.0 works and updated the release note/commit message ^ | |
| 15:54:23 | jaypipes | dansmith: count(indexed_col) is the fastest way for innodb tables, yes. | |
| 15:54:23 | stephenfin | mriedem: Only thing that needed changing was a 'nova.conf' option (notes on the review). Thought that was too specific to note but let me know if you think otherwise. | |
| 15:54:39 | dansmith | jaypipes: and count(unindexed_column) is expensive? | |
| 15:55:02 | dansmith | and count(*) ? | |
| 15:58:18 | openstackgerrit | Merged openstack/nova stable/queens: [placement] Add sending global request ID in put (3) https://review.openstack.org/543113 | |
| 15:58:25 | openstackgerrit | Merged openstack/nova stable/queens: Update noVNC deployment docs to mention non-US keymap fix in 1.0.0 https://review.openstack.org/548057 | |
| 15:58:34 | openstackgerrit | Merged openstack/nova stable/queens: [placement] Add functional tests for traits API https://review.openstack.org/547843 | |
| 15:58:50 | jaypipes | dansmith: yes sir | |
| 15:58:54 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: Usages per project and user (v1.8, v1.9) https://review.openstack.org/514646 | |
| 15:58:54 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: Resource class set (v1.7) https://review.openstack.org/514644 | |
| 15:58:55 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: New dict format of allocations (v1.11, v1.12) https://review.openstack.org/542819 | |
| 15:58:55 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: CLI allocation candidates (v1.10) https://review.openstack.org/514647 | |
| 15:58:56 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: Add nested resource providers (v1.14) https://review.openstack.org/546675 | |
| 15:58:56 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: Transactionally update allocations (v1.13) https://review.openstack.org/546674 | |
| 15:58:57 | jaypipes | dansmith: (sorry, on a webex concurrently...) | |
| 15:58:57 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: Allocation candidates parameter: required (v1.17) https://review.openstack.org/548326 | |
| 15:58:57 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: Limit allocation candidates (v1.15, v1.16) https://review.openstack.org/548043 | |
| 15:59:19 | dansmith | jaypipes: okay, I guess it seems like mysql must keep a row count somewhere that we can get at for free, but it's not critical | |
| 16:01:19 | mriedem | stephenfin: ack | |
| 16:02:30 | stephenfin | dansmith: Given that vnc/spice options are only used by nova-compute, would it make sense to only set them in 'nova-cpu.conf' in DevStack? | |
| 16:02:48 | dansmith | stephenfin: yeah | |
| 16:02:58 | stephenfin | dansmith: Cool. I'll do that | |
| 16:03:59 | mriedem | stephenfin: so novnc 1.0 has a breaking change huh | |
| 16:04:01 | mriedem | on the base url | |
| 16:04:24 | stephenfin | mriedem: Yup. That seems to be the only one I could identify though | |
| 16:05:15 | stephenfin | i.e. it just worked once I changed that | |
| 16:06:08 | openstackgerrit | Merged openstack/nova stable/queens: libvirt: disconnect volume from host during detach https://review.openstack.org/546655 | |
| 16:08:13 | jaypipes | dansmith: innodb keeps a count of records in an index, yes. not for the total table, though. | |
| 16:08:22 | dansmith | jaypipes: okay | |
| 16:10:50 | artom | stephenfin, does xenapi not use VNC? | |
| 16:11:16 | artom | Err, why did I parse "nova-compute" as "libvirt"? | |
| 16:11:25 | artom | Ignore the noise :/ | |
| 16:11:36 | stephenfin | artom: Yeah, unrelated. However, I do think they mostly use XVP instead | |