| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-01-25 | |||
| 07:54:59 | jianghuaw | Huan created it for getting advices. | |
| 07:55:09 | jianghuaw | it's to fix a migration issue with xenapi. | |
| 07:55:31 | Spazmotic | sure will take a look | |
| 07:55:44 | jianghuaw | Spazmotic, thanks. | |
| 07:57:49 | Spazmotic | Interesting. We do volume based LM quite a bit but we get around this with --block-migrate for iSCSI | |
| 07:59:23 | jianghuaw | Spazmotic, iirc the issue happens only in the XS7.0 and forwarding releases. | |
| 08:00:01 | Spazmotic | That's definately going to be a problem i'll need to test out | |
| 08:00:47 | jianghuaw | or maybe 7.1 => . I put 7.1 in the log. | |
| 08:01:08 | jianghuaw | But anyway it will be issue for latest XS releases. | |
| 08:01:26 | jianghuaw | feel free to put comments in it. | |
| 08:01:54 | jianghuaw | If you interest, feel free to take it over:-) | |
| 08:02:06 | Spazmotic | I'll certainly look at it heavily, appeals to me greatly | |
| 08:02:55 | jianghuaw | Thanks. | |
| 08:08:20 | Spazmotic | Do you know if Huan has refactored it at all recently? Mentions they wanted to tweak the vmops method around, but i'm sure time was tight for you guys | |
| 08:14:57 | jianghuaw | Spazmotic, not yet I'm sure. | |
| 08:15:25 | Spazmotic | Alrighty cool.. I've got some 7.1s in an env, i'll run some tests over the next few days and get familiar with these maps, havn't had to use them before | |
| 08:15:34 | Spazmotic | Thanks for bringing to my attention is actually more important than you may know | |
| 08:16:03 | jianghuaw | cool. | |
| 08:29:29 | openstackgerrit | Hironori Shiina proposed openstack/nova master: virt/ironic: Implement rescue and unrescue https://review.openstack.org/416487 | |
| 08:35:32 | openstackgerrit | Maciej Jozefczyk proposed openstack/nova master: Do not normalize allocation_ratios https://review.openstack.org/532924 | |
| 08:38:29 | maciejjozefczyk | Hello | |
| 08:49:32 | gmann | mriedem: updated those. 2 are ready to go and 1 i need more work on test which i will tomorrow - https://review.openstack.org/#/q/topic:bp/api-extensions-merge-queens+status:open | |
| 08:50:30 | gmann | mriedem: there are other work also needed on this BP which we will continue on continued Rocky BP. if you want to move current patches to new one, it should be ok | |
| 08:52:47 | gmann | mriedem: or we defer the queens one to Rocky ? | |
| 09:05:03 | kashyap | It's middle of the night for Matt, so will take a few more hours for him to come online. | |
| 09:24:44 | Spazmotic | Aargh Zuul why you hate me | |
| 09:27:24 | Roamer` | Spazmotic, if it's any help, I doubt it's personal - it's in Zuul's nature to be very, very paranoid and suspicious of everyone, you just have to catch one of its rare good moods... :) | |
| 09:30:32 | openstackgerrit | Zengzhi Wang proposed openstack/nova stable/pike: Make eventlet hub use a monotonic clock https://review.openstack.org/537811 | |
| 09:57:45 | hrw | morning | |
| 09:58:40 | hrw | I am looking at nova's code. it adds graphics if it is needed. but I did not see where it adds usb host for usbtablet | |
| 09:58:55 | hrw | does it assume that usb host is already present in VM configuration? | |
| 09:59:54 | Roamer` | OK, so the StorPool volume attachment driver finally passed the Zuul check pipeline (after the Cinder backup bug was fixed)... I'm pretty sure mriedem intends to +W it once he wakes up and sees it's passed the tests, but the gate being what it is, I'd be really obliged if somebody else could +W it earlier :) Of course, "let's just wait for him" is a very reasonable option. | |
| 10:00:12 | Roamer` | oof, and that would be https://review.openstack.org/#/c/140733/19 of course | |
| 10:03:54 | kevinz | alex_xu: Could you review https://review.openstack.org/#/c/494156/ if you have time? Stephen already give +2 :-) | |
| 10:05:37 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Transform instance.exists notification https://review.openstack.org/403660 | |
| 10:08:12 | alex_xu | kevinz: got it | |
| 10:11:12 | hrw | tracking how nova gets guest components suxx | |
| 10:13:20 | stephenfin | mriedem, melwitt: Yeah, sorry about https://review.openstack.org/#/c/523958/ I was the approver and that's my bad. Won't happen again | |
| 10:18:35 | hrw | ok. nova assumes that usb host is present | |
| 10:19:51 | kashyap | stephenfin: mriedem: As I've said in the past, I don't quite agree with that line of thinking. Happy to also elaborate why in Dublin. | |
| 10:19:55 | mdbooth | mriedem: melwitt stephenfin I wonder if that policy of old might be up for review, especially in a case like this which is uncontroversial. IIRC the understanding came about after RH was accused of railroading a particular controversial patch. That's an entirely valid concern, but I don't think anybody is claiming it happened here. | |
| 10:20:15 | kashyap | (It somehow seem to imply / assume "malice") | |
| 10:20:49 | kashyap | (But it _should_ be a concern, if it's a _pattern_ of two people from the same company ACKing changes in rapid succession without giving time for others to chime in.) | |
| 10:20:57 | kashyap | (Or for some goddamned deadline's sake.) | |
| 10:21:18 | stephenfin | mdbooth: Indeed, but the discussion should have happened _before_ /me did that :) EAFP probably doesn't apply where matters of trust are involved | |
| 10:21:33 | stephenfin | A good PTG topic though, for sure | |
| 10:22:00 | kashyap | Sure. But I do agree with Matt that, we shouldn't be too eager to push something for a deadline | |
| 10:22:36 | kashyap | Because, let's be honest --- most deadlines milestones don't matter. It's all artificial lines in the sand. | |
| 10:23:56 | mdbooth | stephenfin: There was a spec, many rounds of review, at least 3 ML posts that I recall over a period of a few weeks, and additional thorough review from 2 core reviewers, including yourself. I resent the implication, tbh. | |
| 10:25:03 | kashyap | Yeah, I recall the history / and even postmortem done by markmc. | |
| 10:25:10 | kashyap | http://lists.openstack.org/pipermail/openstack-dev/2014-August/044550.html -- [openstack-dev] [nova] libvirt version_cap, a postmortem | |
| 10:25:59 | kashyap | mdbooth: Ah, I thought you were talking about a previous case which caused this no two +2s from the same company thing.) | |
| 10:27:00 | mdbooth | The previous case was also a misunderstanding, tbh, although there was more substance to it. | |
| 10:27:14 | kashyap | Yeah, absolutely it was a case: "too eager to misunderstand" | |
| 10:27:17 | hrw | https://bugs.launchpad.net/nova/+bug/1745340 - what should I add to get it more complete? | |
| 10:27:18 | openstack | Launchpad bug 1745340 in OpenStack Compute (nova) "Nova assumes that USB Host is present" [Undecided,New] | |
| 10:28:52 | stephenfin | More complete? | |
| 10:29:43 | hrw | stephenfin: more useful to anyone who would look at issue | |
| 10:29:53 | hrw | if anyone at all ;D | |
| 10:30:07 | stephenfin | I think you've pretty much got it | |
| 10:30:39 | stephenfin | I assume the solution would be to check if we have a USB hub and either (a) add it if not or (b) refuse to start with a USB tablet? | |
| 10:31:04 | stephenfin | *refuse to start with a USB tablet if the platform doesn't support the USB hub | |
| 10:31:25 | stephenfin | hrw: I can point you to the code that needs modification, if you fancy the challenge? ;) | |
| 10:31:44 | hrw | stephenfin: please do. in a bug comment | |
| 10:32:08 | hrw | stephenfin: it may end on my or kevinz's plate or maybe we assign it to someone else at Linaro | |
| 10:32:36 | hrw | stephenfin: imho the proper way would be to add usb-host device if not present | |
| 10:33:02 | hrw | as the goal is to have graphical console in Rocky (as it is too late for Queens) | |
| 10:34:21 | hrw | during ptg I have to track nova sticker for my laptop ;D to put it next to kolla one | |
| 10:36:07 | stephenfin | hrw: https://bugs.launchpad.net/nova/+bug/1745340/comments/3 | |
| 10:36:08 | openstack | Launchpad bug 1745340 in OpenStack Compute (nova) "Nova assumes that USB Host is present" [Undecided,New] | |
| 10:37:26 | hrw | stephenfin: exactly where I was looking at | |
| 10:37:40 | hrw | stephenfin: probably some code in nova.virt.libvirt.* too | |
| 10:37:52 | hrw | to 'plug' host | |
| 10:38:14 | hrw | will later look into | |
| 10:38:56 | Spazmotic | Come on tempest... you running slow but i have faith in you | |
| 11:46:01 | stephenfin | sean-k-mooney: Think I've reviewed all of bp/enable-sriov-nic-features now :) | |
| 11:46:13 | stephenfin | Lemme know if you have questions | |
| 11:55:56 | Spazmotic | hey huanxie you still around | |
| 11:56:51 | Spazmotic | or jianghuaw can you explain to me what XS expects of these VDI MAPS on a live migration? | |
| 12:06:12 | jianghuaw | Spazmotic, I've talked with Huan. You can chat with her on tomorrow. | |
| 12:06:29 | Spazmotic | Sounds good, thanks, i'll stop by tomorrow :) | |
| 12:06:48 | lyarwood | mriedem: iirc live snapshot was only for local images and not volumes right? Even then we skip it for rbd? | |
| 12:12:55 | Spazmotic | IBM PowerKVM CI failing constantly | |
| 12:17:10 | Spazmotic | Looks like past 4 or 5 hours all failures on its tempest-dsvm-full-xenialchecks | |
| 12:17:37 | sean-k-mooney | stephenfin: just in so ill take a look at them now | |
| 12:27:02 | owalsh | stephenfin: is vencrypt_ca_certs required when vencrypt is enabled? doesn't looks like it default to the system ca certs so I'm guessing yea | |
| 12:38:49 | openstackgerrit | Brianna Poulos proposed openstack/nova master: Reduce complexity of _from_db_object https://review.openstack.org/537728 | |
| 12:39:02 | openstackgerrit | Gaudenz Steinlin proposed openstack/nova master: Ignore some updates from virt driver https://review.openstack.org/523006 | |
| 13:20:48 | ameeda | jaypipes: Sir are you around ? | |
| 13:21:51 | openstackgerrit | Brianna Poulos proposed openstack/nova master: Implement certificate_utils https://review.openstack.org/479949 | |
| 13:21:51 | openstackgerrit | Brianna Poulos proposed openstack/nova master: Add trusted_certs object https://review.openstack.org/489408 | |
| 13:21:52 | openstackgerrit | Brianna Poulos proposed openstack/nova master: Add trusted_certs to instance_extra https://review.openstack.org/537897 | |
| 13:21:52 | openstackgerrit | Brianna Poulos proposed openstack/nova master: Add trusted_image_certificates to REST API https://review.openstack.org/486204 | |
| 13:41:11 | stephenfin | owalsh: I think so, yup. I should probably encode that | |
| 13:51:39 | Spazmotic | IBM CI what is wrong with you :( http://dal05.objectstorage.softlayer.net/v1/AUTH_3d8e6ecb-f597-448c-8ec2-164e9f710dd6/pkvmci/index.html | |
| 13:52:06 | mmedvede | Spazmotic: it should be fixed now, at least the original problem that was causing earlier failures | |
| 13:52:16 | Spazmotic | Ahhh thanks mmedvede | |
| 13:52:19 | Spazmotic | I'll give it a new recheck hehe | |
| 13:53:13 | mriedem | Roamer`: i won't be +Wing your change because we require 2 +2s | |
| 13:58:26 | mriedem | gibi_: we should probably consider bumping the nova functional job timeouts as a change on top of your patch to move them into the nova tree | |