| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-03-30 | |||
| 11:02:59 | dims | hey cdent, bauzas : is this the only method where we sanitize a end-user provider name to a host name? http://git.openstack.org/cgit/openstack/nova/tree/nova/utils.py#n543 | |
| 11:03:12 | dims | cdent : bauzas : has it changed a lot over time? | |
| 11:04:31 | cdent | dims: sorry, I don't know much about that file nor how/where it is used. Looks like that method last changed in 2015 | |
| 11:04:35 | dims | looks like no one has touched it for at least 2 years ... https://github.com/openstack/nova/blame/master/nova/utils.py | |
| 11:04:37 | dims | right | |
| 11:04:45 | dims | good thing! | |
| 11:04:59 | dims | for context, am looking at https://github.com/kubernetes/kubernetes/pull/61890#issuecomment-377282182 | |
| 11:07:16 | cdent | is that instance hostnames? | |
| 11:07:48 | cdent | because there's no uniqueness guarantee on those | |
| 11:16:12 | openstackgerrit | Chris Dent proposed openstack/nova master: [placement] Filter resource providers by forbidden traits in db https://review.openstack.org/556472 | |
| 11:16:13 | openstackgerrit | Chris Dent proposed openstack/nova master: [placement] Parse forbidden traits in query strings https://review.openstack.org/556819 | |
| 11:16:13 | openstackgerrit | Chris Dent proposed openstack/nova master: [placement] Filter allocation candidates by forbidden traits in db https://review.openstack.org/556660 | |
| 11:16:14 | openstackgerrit | Chris Dent proposed openstack/nova master: [placement] Support forbidden traits in API https://review.openstack.org/556820 | |
| 11:16:21 | dims | cdent : y just looking for possibilities | |
| 11:16:51 | dims | cdent : oops - found a bug under python3 - see output at bottom of http://paste.openstack.org/show/718009/ | |
| 11:17:05 | dims | extra ü | |
| 11:18:24 | dims | but that's a digression, basically i am looking for is "given a sanitized host name, is there a way for me to query nova api to get the real vm (or list of vms since no guarantee)" | |
| 11:20:45 | cdent | I certainy hope there's a way | |
| 11:21:06 | dims | ack thanks. will look deeper :) | |
| 11:22:45 | cdent | If we assume that hostname and instance name are the same thing, listing servers can be filtered by name: https://developer.openstack.org/api-ref/compute/#list-servers (with regex too apparently) | |
| 11:23:16 | cdent | but whether that corresponds to this sanitized thing, dunno dims | |
| 11:24:24 | dims | i have a feeling we don't save that sanitized host name anywhere ... mriedem would know, will bug him | |
| 11:25:35 | cdent | it appears like the santiization is done during server create and applies to the eventual hostname on the instance objects | |
| 11:26:27 | cdent | (I'm in nova/compute/api.py ) | |
| 11:27:08 | openstackgerrit | Bhagyashri Shewale proposed openstack/nova-specs master: Disallow rotation parameter 0 for 'createBackup' API https://review.openstack.org/511825 | |
| 11:35:00 | openstackgerrit | Kashyap Chamarthy proposed openstack/nova master: libvirt: Allow to specify granular CPU feature flags https://review.openstack.org/534384 | |
| 11:44:22 | openstackgerrit | Merged openstack/nova master: libvirt: slow live-migration to ensure network is ready https://review.openstack.org/497457 | |
| 12:28:54 | bhagyashris | jaypipes: Hi, | |
| 12:50:36 | openstackgerrit | Claudiu Belu proposed openstack/nova master: tests: autospecs all the mock.patch usages https://review.openstack.org/470775 | |
| 12:51:48 | openstackgerrit | Claudiu Belu proposed openstack/nova master: WIP: replace spec with autospec https://review.openstack.org/557299 | |
| 12:52:03 | openstackgerrit | Claudiu Belu proposed openstack/nova master: hyper-v: autospec classes before they are instantiated https://review.openstack.org/342211 | |
| 12:55:19 | bauzas | dims: sorry, was afk for sports thingy | |
| 12:55:29 | bauzas | has cdent answered your question ? | |
| 13:05:10 | dims | bauzas : yes, he nudged me forward. so i am at the point where i have an instance name = | |
| 13:05:20 | dims | "Ümlauts rule!" | |
| 13:05:21 | jaypipes | bhagyashris: good morning | |
| 13:06:07 | dims | curl http://169.254.169.254/latest/meta-data/hostname gives me "mlauts-rule.novalocal" | |
| 13:06:19 | bhagyashris | jaypipes: Good morning :) for it's evening ;) | |
| 13:06:43 | dims | bauzas : and the "openstack server show" does not have the sanitized host name "mlauts-rule.novalocal" | |
| 13:06:55 | dims | bauzas : do we store that somewhere? | |
| 13:06:59 | bhagyashris | jaypipes: I want to discuss regarding the shared resource provider and nested resource provider thing. | |
| 13:07:19 | bhagyashris | jaypipes: I saw the blueprint #link: https://blueprints.launchpad.net/nova/+spec/nested-resource-provider and that is marked as implemented. | |
| 13:07:25 | bauzas | dims: are you talking of a i18n'd hostname ? | |
| 13:07:34 | bauzas | for the guest ? | |
| 13:07:38 | dims | yes | |
| 13:07:52 | bauzas | leakypipes: isn't Good Friday for you? | |
| 13:07:54 | dims | i believe we use sanitized_hostname in utils.py | |
| 13:07:57 | bauzas | and dims as well ? | |
| 13:08:11 | dims | bauzas : yep :) | |
| 13:08:19 | bauzas | so AFAIR, there are two different things | |
| 13:08:26 | bauzas | the display name and the instance hostname | |
| 13:08:26 | leakypipes | dims: morning :) | |
| 13:08:43 | bauzas | dims: leakypipes: isn't that day off for most of the folks in the US ? | |
| 13:08:44 | leakypipes | bauzas: is it? not sure... I don't religion | |
| 13:08:52 | bhagyashris | jaypipes: I have tested one scenario which is mentioned in #link: http://paste.openstack.org/show/718016/ and observed that the shared resource provider DISK_GB is not consumed when I boot the instance. (Ideally it should take DISK_GB from shared resource provider and VCPU and MEMORY_MB from compute node resource provider). | |
| 13:09:00 | bauzas | or, maybe just a company holiday | |
| 13:09:00 | openstackgerrit | Matt Riedemann proposed openstack/nova stable/queens: libvirt: slow live-migration to ensure network is ready https://review.openstack.org/557930 | |
| 13:09:05 | bauzas | snaaaaaap | |
| 13:09:10 | bauzas | now I see mriedem too | |
| 13:09:12 | bhagyashris | jaypipes: I have followed the steps which are mentioned in provided link if I am wrong at any place please correct me. | |
| 13:09:25 | bauzas | me who was thinking the nova channel be quiet today afternoon :p | |
| 13:09:47 | mriedem | is this the one day of the year that red hat people in the US are on mandatory vacation but france isn't? | |
| 13:10:09 | bauzas | leakypipes: and I don't do religion either, but in my country, our religious holidays are off | |
| 13:10:19 | mriedem | it's black friday right? | |
| 13:10:20 | bauzas | so maybe it's the source of my confusion | |
| 13:10:21 | leakypipes | bhagyashris: k, I will look into it later, ok? I need to get some coffee and food (sorry) | |
| 13:10:28 | mriedem | oh no... | |
| 13:10:29 | dims | bauzas : y, i can use the metadata url to get the hostname and i want to find the correct vm corresponding to that hostname | |
| 13:10:30 | mriedem | heh | |
| 13:10:31 | mriedem | good friday? | |
| 13:10:36 | mriedem | something something christian friday | |
| 13:10:36 | bauzas | mriedem: yeah, indeed, most of the Red Hat world is shutdown except France | |
| 13:10:44 | efried | oh yeah... | |
| 13:10:48 | bauzas | take it as an evidence that France is ruling the company | |
| 13:11:00 | bhagyashris | leakypipes: ok | |
| 13:11:30 | bauzas | dims: I see your point | |
| 13:11:51 | dims | bauzas : kubernetes cloud provider scenario | |
| 13:11:58 | bauzas | mostly for cloud-init reasons where you wanna set the correct hostname on the guest itself | |
| 13:12:07 | bauzas | gotcha | |
| 13:13:24 | openstackgerrit | sahid proposed openstack/nova stable/queens: libvirt: slow live-migration to ensure network is ready https://review.openstack.org/557930 | |
| 13:13:43 | kashyap | alex_xu_: mriedem: When you get a moment, I think, this is ready: https://review.openstack.org/#/c/534384/21 | |
| 13:13:56 | bauzas | dims: so, back to your concern | |
| 13:14:33 | kashyap | But it's a holiday for Matt, so feel free to disregard it until you get back | |
| 13:14:33 | bauzas | dims: you wanna make sure that we sanitize hostnames ? | |
| 13:15:03 | dims | bauzas : we are doing that already, so given a sanitized hostname i can't find the vm by instance name | |
| 13:15:09 | sahid | mriedem: hum.. you have already cherry-picked https://review.openstack.org/#/c/497457/ | |
| 13:15:15 | bauzas | dims: ah I see | |
| 13:15:24 | mriedem | sahid: not supposed to go to stable? | |
| 13:15:29 | sahid | i did not noticed that, and cherry-picked it again | |
| 13:15:43 | dims | bauzas : "Ümlauts rule!" becomes "mlauts-rule.novalocal" | |
| 13:15:44 | sahid | mriedem: it is | |
| 13:15:45 | mriedem | sahid: oh, that's fine | |
| 13:15:49 | sahid | ok | |
| 13:15:58 | mriedem | kashyap: at this point i think i want to wait for dan to review that again | |
| 13:16:12 | kashyap | Sure, I'm not hurrying | |
| 13:16:45 | kashyap | But it is still important for upstream backports; just that the sooner we ship this, the better it is for users. | |
| 13:17:23 | kashyap | If it could wait this long; it can wait until dansmith gets back on Tue or whenever. | |
| 13:17:24 | mriedem | bauzas: can you go through these other pike backports that have a +2? https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/pike+label:Code-Review=2 melwitt wants to do a release | |
| 13:17:39 | bauzas | mriedem: it's doable for a Friday task | |
| 13:18:23 | bauzas | dims: okay, and then per that hostname, you'd like to find the instance ? | |