| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-06-29 | |||
| 09:58:43 | cdent | finucannot: see the backlog, there's some discussion of doing both | |
| 09:58:58 | cdent | s/doing/trying/ and seeing which one floats | |
| 10:02:42 | kaisers | Hey Novas, anyone familiar with Nova+QEMU please have a look: It seems to me the change https://review.openstack.org/#/c/553072/ introduced a dependency for QEMU v2.9+ while https://wiki.openstack.org/wiki/LibvirtDistroSupportMatrix shows QEMU v2.5 as minimum supported version. The patch only checks for a libvirt version but not for QEMU 2.9+ . Or maybe i'm overlooking something? (QEMU v2.9 adds the pass mtu to guest fe | |
| 10:03:57 | kaisers | I'm not absolutely sure my 2.9 ref is the required feature but it's the best lead i found so far. Our CIs are stumbling on setting the MTU as they're running QEMU 2.5 (because Xenial release) | |
| 10:11:57 | openstackgerrit | Yikun Jiang (Kero) proposed openstack/nova master: Refactor the policies to policy https://review.openstack.org/579113 | |
| 10:11:57 | openstackgerrit | Yikun Jiang (Kero) proposed openstack/nova master: Add InstanceGroupPolicy object https://review.openstack.org/573628 | |
| 10:39:23 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Define common variables for irrelevant-files https://review.openstack.org/578882 | |
| 10:50:08 | openstackgerrit | Stephen Finucane proposed openstack/nova master: manage: Remove dead code https://review.openstack.org/579118 | |
| 10:50:45 | finucannot | PapaOurs: Wanna send this through? https://review.openstack.org/#/c/530923/ | |
| 10:59:00 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Update admin/flavors document https://review.openstack.org/573063 | |
| 11:18:57 | openstackgerrit | Lee Yarwood proposed openstack/nova master: libvirt: Handle LM rollback error when detaching volumes from transient domain https://review.openstack.org/579125 | |
| 11:18:59 | lyaaaaaaaarwood | kashyap: ^ as discussed earlier, would you mind taking a look? | |
| 11:19:17 | kashyap | lyaaaaaaaarwood: _Just_ about to grab a quick bite | |
| 11:19:27 | lyaaaaaaaarwood | kashyap: ack no rush | |
| 11:20:01 | kashyap | Have it open in a tab. Will look. | |
| 11:20:10 | lyaaaaaaaarwood | thanks, enjoy lunch :) | |
| 12:19:48 | tssurya | mriedem, melwitt: good morning! I have replied on spec with comments on the confusions regarding delete stuff (sorry for the delay, I missed the notifications somehow). | |
| 12:20:17 | tssurya | if its not clear, whenever you are free later on, we could do an IRC chat | |
| 12:22:38 | openstackgerrit | Surya Seetharaman proposed openstack/nova stable/queens: Make nova service-list use scatter-gather routine https://review.openstack.org/579135 | |
| 12:31:10 | mriedem | tssurya: ok, i'm not adequately caffeinated yet | |
| 12:32:04 | tssurya | mriedem: np, we can do it later in your evening as well | |
| 12:33:06 | openstackgerrit | Surya Seetharaman proposed openstack/nova stable/pike: Make nova service-list use scatter-gather routine https://review.openstack.org/579136 | |
| 12:41:24 | PapaOurs | finucannot: re: https://review.openstack.org/#/c/579110/ we can't just bump the lower version easily | |
| 12:42:05 | PapaOurs | finucannot: because a lot of projects use webob, and in between 1.7 and 1.8, there are some not backwards compatibilty | |
| 12:44:00 | openstackgerrit | Matt Riedemann proposed openstack/osc-placement master: CLI allocation candidates (v1.10) https://review.openstack.org/514647 | |
| 12:44:01 | openstackgerrit | Matt Riedemann proposed openstack/osc-placement master: New dict format of allocations (v1.11, v1.12) https://review.openstack.org/542819 | |
| 12:44:02 | openstackgerrit | Matt Riedemann proposed openstack/osc-placement master: Transactionally update allocations (v1.13) https://review.openstack.org/546674 | |
| 12:44:03 | openstackgerrit | Matt Riedemann proposed openstack/osc-placement master: Add nested resource providers (v1.14) https://review.openstack.org/546675 | |
| 12:44:04 | openstackgerrit | Matt Riedemann proposed openstack/osc-placement master: Limit allocation candidates (v1.15, v1.16) https://review.openstack.org/548043 | |
| 12:44:05 | openstackgerrit | Matt Riedemann proposed openstack/osc-placement master: Allocation candidates parameter: required (v1.17) https://review.openstack.org/548326 | |
| 12:45:24 | mriedem | we do have per-project lower constraints now for a reason | |
| 12:45:32 | mriedem | this doesn't need to be backported | |
| 12:45:48 | mriedem | i'm not sure if everyone assumes that all services on the same host are deployed in containers or venvs now though | |
| 12:46:10 | mriedem | we certainly don't in devstack.. | |
| 12:50:34 | PapaOurs | mriedem: so, what's your opinion ? | |
| 12:50:52 | mriedem | meh | |
| 12:51:24 | mriedem | would probably need to talk to the requirements or release team about how they think per-project lower-constraints is meant to be handled in this case, since we don't have a common shared global-requirements anymore | |
| 12:51:48 | mriedem | does that mean we, as the openstack community, assert that deployments must use containers/venvs if deploying multiple different services on the same host? | |
| 12:51:56 | mriedem | maybe that's a question for the ML... | |
| 12:54:36 | kashyap | Yeah, definitely sounds like an ML topic, that can end up being KM-long thread | |
| 12:58:07 | mriedem | wwad | |
| 12:58:12 | mriedem | what would airship do | |
| 12:58:57 | openstackgerrit | Matt Riedemann proposed openstack/osc-placement master: Limit allocation candidates (v1.15, v1.16) https://review.openstack.org/548043 | |
| 12:58:59 | openstackgerrit | Matt Riedemann proposed openstack/osc-placement master: Allocation candidates parameter: required (v1.17) https://review.openstack.org/548326 | |
| 13:01:06 | fried_rice | PapaOurs: I agree with finucannot; is there some reason we can't just bump requirements and l-c? | |
| 13:01:21 | giblet | finucannot: quick question in https://review.openstack.org/#/c/578825/1/tox.ini@3 before I approve | |
| 13:01:39 | PapaOurs | fried_rice: wait, lower-constraints is for all the projects | |
| 13:01:51 | finucannot | PapaOurs: It's per-project now | |
| 13:02:19 | finucannot | See the lower-constraints.txt file in root (and lower-constraints tox target) | |
| 13:03:06 | finucannot | giblet: Nope. That's a hangover from a previous rebase, I imagine. It's harmless but I can edit and switch it back if you'd like? | |
| 13:03:32 | giblet | finucannot: no need to change it back, I was just curious | |
| 13:03:35 | giblet | finucannot: approving | |
| 13:03:37 | PapaOurs | finucannot: so, we should bump it for nova ? | |
| 13:03:56 | fried_rice | giblet: I answered at least why it was reversed to begin with. | |
| 13:04:00 | fried_rice | PapaOurs: Yes | |
| 13:04:21 | finucannot | fried_rice: Ah, correct | |
| 13:04:24 | fried_rice | PapaOurs: You'll need to do it in both requirements.txt and lower-constraints.txt | |
| 13:04:31 | giblet | fried_rice: thanks | |
| 13:05:40 | finucannot | PapaOurs: Yeah, that was the idea behind moving those things in-tree, I thought | |
| 13:06:34 | giblet | fried_rice: I opened https://review.openstack.org/#/c/560317/17/nova/tests/functional/libvirt/test_report_cpu_traits.py but at a first glance I don't know the answer of the failure. | |
| 13:06:50 | giblet | fried_rice: I can try to dig on Monday | |
| 13:07:11 | fried_rice | giblet: What was the patch you got into recently where it was something to do with creating or deleting a file on the file system, and that raced among several tests? | |
| 13:08:14 | fried_rice | If you can just point me at that, I can look into it. Cause I suspect the failure is the same or similar. Or at least worth eliminating as a suspect. | |
| 13:08:33 | giblet | fried_rice: that wasn't my patch but I try to find it as now I remember what you are referring to | |
| 13:08:51 | fried_rice | giblet: Right, I remember you swooped in and saved the day | |
| 13:09:40 | cdent | he has a habit of doing that | |
| 13:10:05 | fried_rice | Heh, I misread, at first I thought you were saying *you* have a habit of doing that. | |
| 13:10:26 | fried_rice | i.e. a habit of picturing giblet in a cape. | |
| 13:10:44 | giblet | :) | |
| 13:10:47 | cdent | don't share my secrets in public, please | |
| 13:16:00 | giblet | fried_rice: is the one? https://review.openstack.org/#/c/560459/16 | |
| 13:16:12 | openstackgerrit | Merged openstack/nova master: [placement] Extract create_allocation_list https://review.openstack.org/577200 | |
| 13:16:21 | openstackgerrit | Merged openstack/nova master: [placement] Fix capacity tracking in POST /allocations https://review.openstack.org/578133 | |
| 13:16:22 | giblet | fried_rice: here the CONF.instance_path was the root cause | |
| 13:17:06 | openstackgerrit | Merged openstack/nova master: libvirt: Add missing encryption_secret_uuid tests for pre_live_migration https://review.openstack.org/540679 | |
| 13:17:24 | giblet | http://logs.openstack.org/59/560459/16/gate/nova-tox-functional/c04cbeb/testr_results.html.gz | |
| 13:19:40 | fried_rice | giblet: Yeah, yeah, doesn't that look pretty much identical? | |
| 13:19:50 | giblet | fried_rice: yeah it does | |
| 13:20:19 | giblet | fried_rice: so something is missing from the filesystem the test depends on | |
| 13:22:02 | mriedem | tssurya: replied inline on the spec, i think we're on the same page | |
| 13:22:16 | fried_rice | fixtures.MockPatch('nova.virt.libvirt.utils.get_fs_info')) | |
| 13:22:16 | fried_rice | self.useFixture( | |
| 13:22:16 | fried_rice | giblet: In the other patch, the solution was | |
| 13:23:08 | tssurya | mriedem: FYI regarding backporting https://review.openstack.org/#/c/578152/ to Pike, just realized that Pike is way behind in listing , it doesn't have https://github.com/openstack/nova/commit/bac8f5b046283fbc641d418443e5b899f2989fdf#diff-41e6d0ef0ef79d44593ee0f1b241f9f4 | |
| 13:23:31 | giblet | fried_rice: yeah, the current patch hits that function as well | |
| 13:23:35 | tssurya | so will stop at queens which is already merged. | |
| 13:23:57 | giblet | fried_rice: either the same mock is needed here or we should globally mock the get_fs_info in a fixture | |
| 13:24:20 | giblet | giblet: I would suggest the later as we hit this issue twice in a short time | |
| 13:24:42 | fried_rice | giblet: I agree, but maybe not in this patch? | |
| 13:24:57 | giblet | fried_rice: doing the fixture work in a followup is fine by me | |
| 13:25:18 | giblet | fried_rice: just add a the mock to make the test stable now | |
| 13:26:28 | fried_rice | ++ Thanks for the help, giblet | |
| 13:27:06 | giblet | fried_rice: I'm happy that I could help | |
| 13:30:02 | PapaOurs | cdent: which webob version should be lower for nova ? 1.8.2 ? | |
| 13:30:09 | PapaOurs | lowest* | |
| 13:32:09 | cdent | PapaOurs: I can't guess. The new functionality came in with 1.8.0 but appears it was somewhat broken and fixed in 1.8.1. So I'd say either 1.8.1. or 1.8.2. May as well 1.8.2 unless there's some reason not to? | |
| 13:35:38 | fried_rice | ^ wfm | |
| 13:40:19 | mriedem | tssurya: oh that's right | |
| 13:43:47 | openstackgerrit | Matt Riedemann proposed openstack/osc-placement master: Allocation candidates parameter: required (v1.17) https://review.openstack.org/548326 | |
| 13:47:55 | PapaOurs | cdent: yup, I think 1.8.2 is good | |