| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-07-10 | |||
| 14:31:48 | dansmith | jaypipes: can you legitimify my comment here? https://review.openstack.org/#/c/476631/19 | |
| 14:32:37 | mriedem | dansmith: we're talking about it in the scheduler meeting | |
| 14:32:42 | mriedem | #openstack-meeting-alt | |
| 14:32:43 | dansmith | oh | |
| 14:54:03 | openstackgerrit | Merged openstack/python-novaclient master: Fix cropping the endpoint url https://review.openstack.org/479818 | |
| 14:59:10 | artom | mriedem, could you remove the procedural -2 on https://review.openstack.org/#/c/467321/? | |
| 15:00:11 | mriedem | done, but i haven't really looked at everything else going on there | |
| 15:00:31 | artom | mriedem, cheers! | |
| 15:00:41 | artom | mriedem, bah, that can be done in its own time | |
| 15:00:49 | artom | Getting the -2 off is a good first step ;) | |
| 15:03:29 | mriedem | i was a -W but who's counting | |
| 15:03:53 | mriedem | artom: not sure if you saw before going on PTO, but i opened a docs bug for putting something in the feature support matrix for device tags, | |
| 15:03:57 | mriedem | is that something you can work on? | |
| 15:04:13 | artom | mriedem, I saw that, yeah | |
| 15:04:26 | artom | mriedem, didn't someone already submit a patch? | |
| 15:05:00 | mriedem | ah yes https://review.openstack.org/#/c/481478/ | |
| 15:05:19 | artom | mriedem, err, twice, even | |
| 15:05:24 | artom | https://review.openstack.org/#/c/481479/ as well | |
| 15:05:34 | artom | Or are those different? | |
| 15:05:43 | mriedem | once is for the server create, the other is attach | |
| 15:06:02 | artom | Aha! I'll put those on my to-review list, in a call right this instant | |
| 15:08:34 | openstackgerrit | KaiLi proposed openstack/nova master: Add "default" rule when generate policy.yaml.sample https://review.openstack.org/481443 | |
| 15:14:47 | mkucia | Hi, I have question for qeens as I believe this is beyond pike. Is there some work planned that is related to extracting ResourceProvider handling from nova's driver code? For example when libvirt xml is being assembled there are multiple interations over different kinds of resources. In the RPs perspective it seems for me that those resources can be generalized. | |
| 15:15:10 | openstackgerrit | sean mooney proposed openstack/os-vif master: add host port profile info class https://review.openstack.org/441590 | |
| 15:18:30 | mriedem | mkucia: there are no plans for that | |
| 15:18:55 | mriedem | mkucia: there are only plans to extract the placement service/api/data model into a separate project | |
| 15:19:27 | mkucia | Ok. Thanks! | |
| 15:27:05 | openstackgerrit | sean mooney proposed openstack/os-vif master: set mtu on all code paths https://review.openstack.org/469951 | |
| 15:27:06 | bauzas | dumb question, but can someone point me if/where we have privsep commands in our codebase? | |
| 15:29:28 | mriedem | we don't | |
| 15:29:41 | mriedem | os-brick and os-vif use privsep | |
| 15:29:41 | bauzas | nevermind, got the fully monty https://blueprints.launchpad.net/nova/+spec/privsep | |
| 15:29:50 | bauzas | full* | |
| 15:29:59 | bauzas | mriedem: that's what I discovered | |
| 15:30:03 | mriedem | you can also check out https://review.openstack.org/#/c/459166/ | |
| 15:30:05 | bauzas | the whiteboard is pretty explicit | |
| 15:30:07 | bauzas | yeah | |
| 15:30:20 | bauzas | finding https://blueprints.launchpad.net/nova/+spec/privsep was helpful | |
| 15:45:16 | mriedem | sdague: are you back around this week? | |
| 15:46:02 | openstackgerrit | sean mooney proposed openstack/os-vif master: set mtu on all code paths https://review.openstack.org/469951 | |
| 15:49:48 | mgagne | mriedem: sorry about that (overriding your changes) | |
| 15:50:20 | mriedem | mgagne: no problem, got everything updated last night, should be ready for core review | |
| 15:51:18 | mriedem | oomichi: can you take a look at this new microversion change when you get a chance? https://review.openstack.org/#/c/454322/ | |
| 15:51:29 | mriedem | i've gone through it and would +2 but i added the os-instance-actions changes | |
| 16:04:45 | openstackgerrit | dane-fichter proposed openstack/nova master: Mask password when logging connection_info during swap volume https://review.openstack.org/466533 | |
| 16:09:34 | openstackgerrit | Alex Szarka proposed openstack/nova master: Raise a real Exception instead of direct Exception call https://review.openstack.org/482200 | |
| 16:10:11 | sean-k-mooney | jaypipes: 30 second rereivew https://review.openstack.org/#/c/469951/ i move some comments that did not make sense after the refactor but otherwise identical to the version you just reviewed. sorry for the churn | |
| 16:10:57 | openstackgerrit | Alex Szarka proposed openstack/nova master: Raise a real Exception instead of direct Exception call https://review.openstack.org/482200 | |
| 16:11:48 | openstackgerrit | Alex Szarka proposed openstack/nova master: Raise a real Exception instead of direct Exception call https://review.openstack.org/482200 | |
| 16:13:15 | jaypipes | sean-k-mooney: cool +2 | |
| 16:24:53 | kashyap | Any Ceph people, the RBD cache is enabled by default - isn't it? | |
| 16:25:24 | kashyap | Yep, it is - http://docs.ceph.com/docs/kraken/rbd/rbd-config-ref/ | |
| 16:29:24 | sdague | mriedem: very ish... I'm traveling out to openwest tomorrow and there for the rest of the week | |
| 16:29:54 | sdague | mriedem: if you have a few priority things for me to look at I can, but won't have a ton of bw this week | |
| 16:30:52 | mriedem | sdague: ok, https://review.openstack.org/#/c/454322/ would be good | |
| 16:31:15 | mriedem | there is a cinder change dependent on it, and i've pushed devstack and tempest changes to test it all out - flushed a few bugs last week, test is green now | |
| 16:38:49 | openstackgerrit | Jay Pipes proposed openstack/nova master: placement: claim resources in scheduler https://review.openstack.org/476632 | |
| 16:38:51 | jaypipes | mriedem, dansmith: ^^ | |
| 16:44:08 | mriedem | yar matey | |
| 16:59:05 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Mark LXC as missing for swap volume support https://review.openstack.org/482216 | |
| 17:14:08 | dansmith | jaypipes: remind me.. we create allocations but the computes won't delete them yet because they're not instance.host=$me, right? | |
| 17:14:31 | dansmith | jaypipes: you're not very defensive about your allocation delete, and I was wondering if we could race to delete those with anything making that risky | |
| 17:15:59 | jaypipes | dansmith: in answer to your first question, yes, we don't delete the allocations because of this code: https://github.com/openstack/nova/blob/master/nova/compute/resource_tracker.py#L1055-L1061 | |
| 17:16:45 | jaypipes | dansmith: w.r.t. not being very defensive about allocation delete, I'm not following you. I'm deleting allocations that were successfully claimed in a multi-instance launch request where one of the claims failed but others succeeded. | |
| 17:17:13 | dansmith | jaypipes: right, but if one delete fails we won't continue to delete the rest (or try at least) | |
| 17:18:03 | artom | What's our policy on English mistakes in things like commit messages and comments? | |
| 17:18:07 | jaypipes | dansmith: if one delete fails, that's correct, we just ignore the failure and attempt deleting the others that were successfully allocated. | |
| 17:18:35 | dansmith | jaypipes: I was asserting that you wouldn't, but it looks like delete_allocation_for_instance() won't ever raise/fail | |
| 17:18:39 | jaypipes | artom: I always point out the mistakes | |
| 17:18:42 | dansmith | artom: +1 for correcting | |
| 17:18:44 | dansmith | me too | |
| 17:18:47 | jaypipes | dansmith: right. | |
| 17:19:00 | artom | dansmith, jaypipes, cheers! | |
| 17:19:12 | jaypipes | artom: I speak good. | |
| 17:19:26 | artom | jaypipes, only because you not fail english. | |
| 17:19:37 | jaypipes | heh :) | |
| 17:20:54 | artom | Aww, I was hoping someone would come in with the unpossible reference :( | |
| 17:22:47 | kashyap | artom: It's the way you phrase it, right? | |
| 17:22:55 | kashyap | artom: About correcting non-native English | |
| 17:23:02 | kashyap | If one talks like a moron, it'll come off bad | |
| 17:23:18 | kashyap | A good community example for this is - Eric Blake from QEMU | |
| 17:23:38 | artom | kashyap, you mean me correcting English mistakes and coming off as a pedantic twat? | |
| 17:23:39 | kashyap | He has this amazing knack to give feedback without sounding like a ****. And he's a stickler for English :-) | |
| 17:23:50 | kashyap | (So am I!) | |
| 17:23:52 | kashyap | aries: Yeah | |
| 17:23:59 | kashyap | Err, it was for artom | |
| 17:24:31 | artom | kashyap, thanks, I'll try to see if I can distill his technique into usable dos and donts | |
| 17:24:52 | cdent | dansmith: is this functional test still something you hope to make happen, or has it been superseded ? https://review.openstack.org/#/c/452006/ | |
| 17:24:58 | artom | kashyap, it's still over text though - I can have the best intentions in the world, it's hard to know how your comments will be received | |
| 17:25:39 | kashyap | artom: Meanwhile, look at this feedback I got about how I over-use "comma", a developer (I know) labelled me: "You've gone comma-wild"! | |
| 17:26:01 | dansmith | cdent: yeah I guess | |
| 17:26:07 | kashyap | artom: `grep` for the terms: "conversation" and "wild" here :-) https://lists.nongnu.org/archive/html/qemu-devel/2017-07/msg01455.html | |
| 17:26:11 | artom | kashyap, I am disappoint | |
| 17:26:21 | artom | They completely missed the "you've made me comma-tose" pun | |
| 17:26:45 | cdent | dansmith: yeah, you’re still working on it, or yeah, it’s dead, or maybe “yeah, it should exist but I’ve not got the time”? | |
| 17:26:48 | kashyap | artom: Ah, good one :-) | |
| 17:26:58 | dansmith | cdent: mostly the latter | |
| 17:27:21 | dansmith | cdent: why, you want to work on it? | |
| 17:27:36 | cdent | dansmith: cool, thanks | |
| 17:28:20 | cdent | dansmith: well, perhaps, but the immediate issue was that it’s in my to do list as a recurring thing to review, and I’ve been defering it over increasing long periods | |