| *** harlowja is now known as harlowja_away | 00:04 | |
| Haomeng | morning ironic:) | 00:06 |
|---|---|---|
| NobodyCam | morning Haomeng :) | 00:10 |
| Haomeng | NobodyCam: :) | 00:12 |
| *** harlowja_away is now known as harlowja | 00:15 | |
| *** matsuhashi has joined #openstack-ironic | 00:19 | |
| *** anniec has joined #openstack-ironic | 00:22 | |
| *** mrda is now known as mrda_away | 00:23 | |
| * NobodyCam wanders afk | 00:32 | |
| *** datajerk has joined #openstack-ironic | 00:43 | |
| *** rloo_ has quit IRC | 00:54 | |
| *** anniec has quit IRC | 01:04 | |
| *** anniec has joined #openstack-ironic | 01:10 | |
| *** epim has quit IRC | 01:12 | |
| *** nosnos has joined #openstack-ironic | 01:28 | |
| *** hemna has quit IRC | 01:31 | |
| *** mrda_away is now known as mrda | 01:33 | |
| *** igor__ has joined #openstack-ironic | 01:42 | |
| *** igor has quit IRC | 01:42 | |
| *** wusWWW has quit IRC | 02:14 | |
| *** wusWWW has joined #openstack-ironic | 02:15 | |
| *** wusWWW has quit IRC | 02:26 | |
| *** anniec has quit IRC | 02:47 | |
| *** anniec has joined #openstack-ironic | 02:55 | |
| *** fandi has joined #openstack-ironic | 03:08 | |
| *** anniec has quit IRC | 03:14 | |
| *** anniec has joined #openstack-ironic | 03:39 | |
| *** harlowja is now known as harlowja_away | 03:53 | |
| *** anniec has quit IRC | 04:22 | |
| *** anniec has joined #openstack-ironic | 04:30 | |
| *** killer_prince has joined #openstack-ironic | 05:06 | |
| *** saju_m has joined #openstack-ironic | 05:08 | |
| *** pradipta` is now known as pradipta | 05:14 | |
| *** jcooley_ has quit IRC | 05:14 | |
| *** jcooley_ has joined #openstack-ironic | 05:15 | |
| *** shausy has joined #openstack-ironic | 05:30 | |
| openstackgerrit | Jenkins proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/71192 | 06:05 |
| *** mrda is now known as mrda_away | 06:17 | |
| *** saju_m has quit IRC | 06:18 | |
| *** lazy_prince has joined #openstack-ironic | 06:28 | |
| *** anniec has quit IRC | 06:48 | |
| *** jcooley_ has quit IRC | 07:05 | |
| *** matsuhashi has quit IRC | 07:05 | |
| *** jcooley_ has joined #openstack-ironic | 07:06 | |
| *** jcooley_ has quit IRC | 07:08 | |
| openstackgerrit | Shane Wang proposed a change to openstack/ironic: Fix misspellings in ironic https://review.openstack.org/71777 | 07:14 |
| *** matsuhashi has joined #openstack-ironic | 07:21 | |
| *** saju_m has joined #openstack-ironic | 07:32 | |
| *** jcooley_ has joined #openstack-ironic | 07:34 | |
| *** athomas has quit IRC | 07:43 | |
| *** nosnos_ has joined #openstack-ironic | 07:51 | |
| *** nosnos has quit IRC | 07:54 | |
| *** matsuhashi has quit IRC | 08:10 | |
| *** mdurnosvistov_ has joined #openstack-ironic | 08:20 | |
| *** martyntaylor has joined #openstack-ironic | 08:25 | |
| *** yuriyz has joined #openstack-ironic | 08:29 | |
| *** aignatov_ is now known as aignatov | 08:31 | |
| *** ndipanov_gone is now known as ndipanov | 08:41 | |
| *** mdurnosvistov_ has quit IRC | 08:57 | |
| *** Haomeng|2 has joined #openstack-ironic | 09:22 | |
| *** derekh has joined #openstack-ironic | 09:22 | |
| *** Haomeng has quit IRC | 09:25 | |
| *** lucasagomes has joined #openstack-ironic | 09:30 | |
| yuriyz | Morning Ironic | 09:36 |
| yuriyz | ping lucasagomes | 09:38 |
| max_lobur_afk | morning folks | 09:41 |
| *** nosnos_ has quit IRC | 09:43 | |
| mdurnosvistov | Morning yuriyz, max_lobur_afk =) | 09:44 |
| *** max_lobur_afk is now known as max_lobur | 09:46 | |
| lucasagomes | yuriyz, morning :) | 09:46 |
| lucasagomes | max_lobur, mdurnosvistov morning | 09:46 |
| yuriyz | Lucas, there are not tests for get_one() in tests.api, is this ok? | 09:46 |
| lucasagomes | yuriyz, there's no? on the nodes? | 09:52 |
| lucasagomes | well heh we should have tests for everything | 09:52 |
| lucasagomes | so it's not ok :( | 09:53 |
| yuriyz | for all resources | 09:53 |
| romcheg | Morning lucasagomes, yuriyz! | 09:54 |
| yuriyz | lucasagomes, I will put this task for Ironic newcomers, ok? | 09:55 |
| lucasagomes | yuriyz, oh good catch I can see we have tests for /nodes, /nodes/detail | 09:56 |
| lucasagomes | yuriyz, sure, thank u very much for that! | 09:56 |
| lucasagomes | romcheg, morning | 09:56 |
| Haomeng|2 | morning all:) | 09:56 |
| lucasagomes | Haomeng|2, morning | 09:57 |
| romcheg | I will poke my intern who's now working on expanding tempest API tests to add test for get_one | 09:57 |
| Haomeng|2 | lucasagomes: :) | 09:57 |
| romcheg | Morning Haomeng|2 | 09:57 |
| Haomeng|2 | romcheg: morning:) | 09:57 |
| *** Haomeng|2 has quit IRC | 09:58 | |
| mdurnosvistov | Morning all!))) | 09:58 |
| romcheg | mdurnosvistov: Ping, are you around? | 09:58 |
| *** Haomeng has joined #openstack-ironic | 09:58 | |
| romcheg | Have you seen the problem with the missing test ^ | 09:58 |
| *** saju_m has quit IRC | 09:58 | |
| *** killer_prince has quit IRC | 10:05 | |
| *** Haomeng|2 has joined #openstack-ironic | 10:10 | |
| *** Haomeng has quit IRC | 10:11 | |
| mdurnosvistov | romcheg, pong)) | 10:13 |
| mdurnosvistov | romcheg, where? | 10:13 |
| romcheg | mdurnosvistov: Since you're already working on improving API tests in tempest, could you please take care about testing get_one for nodes? | 10:14 |
| romcheg | mdurnosvistov: The problem is that currently it's not tested in any way | 10:14 |
| *** Haomeng|2 has quit IRC | 10:15 | |
| lucasagomes | rloo, thanks for reviewing the driver :) | 10:18 |
| mdurnosvistov | romcheg, I've added you to patch review | 10:18 |
| *** aignatov is now known as aignatov_ | 10:25 | |
| max_lobur | folks, how do you think, does change to RPC method (CAST to CALL) requires updating RPC_API_VERSION ? | 10:45 |
| max_lobur | the interface stays the same | 10:46 |
| lucasagomes | max_lobur, is call() returning something? | 10:46 |
| lucasagomes | if so I would say we need to change the version | 10:46 |
| lucasagomes | ohhh | 10:46 |
| lucasagomes | cast to call | 10:46 |
| lucasagomes | ignore my comment :P | 10:46 |
| lucasagomes | well not really... after changing to call() ur going to return something for the user, ack? | 10:47 |
| max_lobur | lucasagomes, nope | 10:47 |
| lucasagomes | if so we might want to change it yes, cause the return value is going to change | 10:47 |
| max_lobur | this is just to transferring exception back to the API | 10:48 |
| max_lobur | it always returns None in normal case | 10:48 |
| lucasagomes | right | 10:48 |
| lucasagomes | but it's going to raise a different exception or something like that? | 10:48 |
| max_lobur | yep | 10:48 |
| max_lobur | so it was asyng - returned nothing | 10:48 |
| lucasagomes | if nothing changes in the API POV, I think it's grand to keep the same version then | 10:49 |
| max_lobur | and became sync - returns none | 10:49 |
| lucasagomes | ic | 10:49 |
| lucasagomes | well so I think we should keep the same | 10:49 |
| max_lobur | well, async returned None as well | 10:49 |
| max_lobur | the only difference in exceptions | 10:49 |
| max_lobur | k, I agree | 10:49 |
| *** jcooley_ has quit IRC | 10:52 | |
| *** jcooley_ has joined #openstack-ironic | 10:52 | |
| *** jcooley_ has quit IRC | 10:57 | |
| *** aignatov_ is now known as aignatov | 11:04 | |
| *** shausy has quit IRC | 11:12 | |
| *** shausy has joined #openstack-ironic | 11:13 | |
| *** romcheg is now known as romcheg_ltp | 11:24 | |
| *** romcheg_ltp is now known as romcheg | 11:24 | |
| openstackgerrit | Sascha Peilicke proposed a change to stackforge/pyghmi: Support building wheels (PEP-427) https://review.openstack.org/57177 | 11:25 |
| *** shausy has quit IRC | 11:35 | |
| *** shausy has joined #openstack-ironic | 11:36 | |
| *** viktors has joined #openstack-ironic | 11:42 | |
| *** martyntaylor has left #openstack-ironic | 11:51 | |
| *** lynxman has quit IRC | 12:12 | |
| *** lynxman has joined #openstack-ironic | 12:13 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: PXE clean_up() to remove the pxe_deploy_key parameter https://review.openstack.org/71879 | 12:13 |
| *** jcooley_ has joined #openstack-ironic | 12:24 | |
| *** jcooley_ has quit IRC | 12:30 | |
| *** lucasagomes is now known as lucas-hungry | 12:40 | |
| *** fandi has quit IRC | 12:42 | |
| *** athomas has joined #openstack-ironic | 13:16 | |
| *** jcooley_ has joined #openstack-ironic | 13:19 | |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Fix race condition when changing node states https://review.openstack.org/69135 | 13:22 |
| *** jdob has joined #openstack-ironic | 13:25 | |
| *** jcooley_ has quit IRC | 13:26 | |
| *** russellb is now known as rustlebee | 13:30 | |
| *** ndipanov is now known as ndipanoff | 13:31 | |
| *** lucas-hungry is now known as lucasagomes | 13:58 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: PXE clean_up() to remove the pxe_deploy_key parameter https://review.openstack.org/71879 | 14:06 |
| *** jbjohnso has joined #openstack-ironic | 14:09 | |
| *** jcooley_ has joined #openstack-ironic | 14:14 | |
| *** igor_ has joined #openstack-ironic | 14:15 | |
| *** igor__ has quit IRC | 14:15 | |
| *** jdob has quit IRC | 14:17 | |
| *** jdob has joined #openstack-ironic | 14:17 | |
| *** jcooley_ has quit IRC | 14:19 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: PXE clean_up() to remove the pxe_deploy_key parameter https://review.openstack.org/71879 | 14:19 |
| *** matty_dubs|gone is now known as matty_dubs | 14:32 | |
| openstackgerrit | Max Lobur proposed a change to openstack/ironic: Fix race condition when changing node states https://review.openstack.org/69135 | 14:39 |
| *** martyntaylor has joined #openstack-ironic | 14:42 | |
| *** athomas has quit IRC | 14:44 | |
| *** jdob_ has joined #openstack-ironic | 14:46 | |
| *** jdob has quit IRC | 14:46 | |
| GheRivero | morning Ironic | 14:53 |
| matty_dubs | Howdy GheRivero | 14:53 |
| max_lobur | morning GheRivero, matty_dubs | 14:54 |
| *** athomas has joined #openstack-ironic | 14:55 | |
| yuriyz | morning/evening all | 15:00 |
| romcheg | Morning folks | 15:04 |
| *** zul has quit IRC | 15:07 | |
| *** jcooley_ has joined #openstack-ironic | 15:08 | |
| *** zul has joined #openstack-ironic | 15:10 | |
| *** jcooley_ has quit IRC | 15:14 | |
| lucasagomes | morning all :) | 15:23 |
| *** aignatov is now known as aignatov_ | 15:25 | |
| *** aignatov_ is now known as aignatov | 15:32 | |
| *** athomas has quit IRC | 15:35 | |
| NobodyCam | Good Morning Ironic | 15:41 |
| max_lobur | Good Morning NobodyCam :) | 15:44 |
| NobodyCam | morning max_lobur | 15:44 |
| NobodyCam | TGIFQ | 15:44 |
| NobodyCam | s/Q/!/ | 15:44 |
| max_lobur | NobodyCam, + | 15:46 |
| NobodyCam | lucasagomes: as soon as I figure out what I messed up in root wrap I should have a patch refactoring libvirt | 15:54 |
| lucasagomes | NobodyCam, ack! morning | 15:54 |
| lucasagomes | :) | 15:54 |
| max_lobur | lucasagomes, wanted to ask, as I guess you're the most experienced API man here :) | 15:57 |
| NobodyCam | :) | 15:57 |
| max_lobur | tempest.common.rest_client: DEBUG: Request Body: {"description": "test1", "nonexistent": "abcd"} | 15:57 |
| max_lobur | tempest.common.rest_client: INFO: Response Status: 201 | 15:57 |
| max_lobur | tempest.common.rest_client: DEBUG: Response Headers: {'date': 'Fri, 07 Feb 2014 15:54:47 GMT', 'content-length': '610', 'content-type': 'application/json; charset=UTF-8', 'server': 'WSGIServer/0.1 Python/2.7.3'} | 15:57 |
| max_lobur | tempest.common.rest_client: DEBUG: Response Body: {"description": "test1", "links": [{"href": "http://192.168.122.224:6385/v1/chassis/b84fd02e-d1be-44cb-bf18-fad321c70c81", "rel": "self"}, {"href": "http://192.168.122.224:6385/v1/chassis/b84fd02e-d1be-44cb-bf18-fad321c70c81", "rel": "bookmark"}], "extra": {}, "created_at": "2014-02-07T15:54:47.296054+00:00", "updated_at": null, "nodes": [{"href": "http://192.168.122.224:6385/v1/chassis/b84fd02e-d | 15:57 |
| max_lobur | 1be-44cb-bf18-fad321c70c81/nodes", "rel": "self"}, {"href": "http://192.168.122.224:6385/chassis/b84fd02e-d1be-44cb-bf18-fad321c70c81/nodes", "rel": "bookmark"}], "uuid": "b84fd02e-d1be-44cb-bf18-fad321c70c81"} | 15:57 |
| max_lobur | do you think we should fill a bug for this and fix? | 15:57 |
| max_lobur | nonexistent property ignored when creating nodes | 15:57 |
| max_lobur | s/nodes/chassis | 15:58 |
| lucasagomes | ohh, max_lobur +1 | 15:58 |
| max_lobur | nodes too I guess | 15:58 |
| lucasagomes | I think it's something, wsme will ignore all the other parameters sent in the request message | 15:58 |
| lucasagomes | if it's not part of the api object | 15:58 |
| max_lobur | I think it should be 400 right? | 15:58 |
| lucasagomes | max_lobur, yea, afaiui yes | 15:58 |
| max_lobur | omg) | 15:59 |
| max_lobur | as far as I ... | 15:59 |
| lucasagomes | understand it :D | 15:59 |
| max_lobur | ah :D | 15:59 |
| max_lobur | will add it to my dictionary :D | 15:59 |
| lucasagomes | hah | 16:00 |
| *** zul has quit IRC | 16:00 | |
| * max_lobur filling a bug | 16:00 | |
| lucasagomes | max_lobur, thanks for that | 16:01 |
| max_lobur | :) | 16:01 |
| *** jcooley_ has joined #openstack-ironic | 16:01 | |
| max_lobur | I'm expanding tempest coverage, discovering a lot of interesting things :) | 16:01 |
| *** athomas has joined #openstack-ironic | 16:02 | |
| lucasagomes | nice! yea I can imagine, I bet once we get ironic in tripleO/out of incubation | 16:03 |
| lucasagomes | we are going to find so many things | 16:03 |
| max_lobur | true | 16:04 |
| *** jcooley_ has quit IRC | 16:08 | |
| *** yuriyz has quit IRC | 16:13 | |
| *** athomas has quit IRC | 16:17 | |
| *** romcheg1 has joined #openstack-ironic | 16:30 | |
| *** romcheg1 is now known as romcheg_ltp | 16:30 | |
| *** viktors has left #openstack-ironic | 16:40 | |
| SpamapS | Hey snarky Ironicites.. does Ironic support PXE booting off interface A, but configuring ports in Neutron for interface B? (as in, separate provisioning/data networks?) | 16:52 |
| NobodyCam | SpamapS: we should be able to support that | 16:53 |
| NobodyCam | and good morning :) | 16:53 |
| SpamapS | NobodyCam: ahoy there | 16:53 |
| NobodyCam | SpamapS: see lines 534 - 561 of https://review.openstack.org/#/c/51328/18/nova/virt/ironic/driver.py | 16:54 |
| lucasagomes | and https://review.openstack.org/#/c/70468/7/ironic/drivers/modules/pxe.py | 16:56 |
| lucasagomes | on review yet ^ | 16:56 |
| lucasagomes | btw, devananda I gave that a try but it didn't work for me :( wasn't able to pxe boot my machine | 16:56 |
| lucasagomes | it might be something in my enviroment as well, do I need to add some special configuration for neutron or something? | 16:57 |
| SpamapS | There's a ton of shorthand and jargon in the code there | 16:57 |
| SpamapS | vifs, pifs, etc. | 16:57 |
| NobodyCam | vif - virtual interface / pif = phyical interface | 16:58 |
| SpamapS | yes I know what the abbreviation is :) | 16:58 |
| NobodyCam | :-p | 16:58 |
| SpamapS | no idea what it means to my use case :) | 16:58 |
| NobodyCam | we're snarky | 16:58 |
| SpamapS | Ironicalites | 16:58 |
| SpamapS | doomed to wander the land of beta for 40 years | 16:58 |
| NobodyCam | lol NO!!!! | 16:59 |
| SpamapS | haha | 16:59 |
| * SpamapS watches NobodyCam's heart explode | 16:59 | |
| lucasagomes | :P | 16:59 |
| *** jcooley_ has joined #openstack-ironic | 16:59 | |
| NobodyCam | heeheh | 16:59 |
| SpamapS | So what I see is that PXE configs are written for all macs | 16:59 |
| NobodyCam | yes | 16:59 |
| SpamapS | https://git.openstack.org/cgit/openstack/ironic/tree/ironic/drivers/modules/pxe.py#n459 | 17:00 |
| SpamapS | Ok, so if my box boots and I want to distinguish one mac as "provision here" and the other as "traffic here" .. how's that done? | 17:00 |
| SpamapS | If that is "not done" .. that's fine, just something I've had on the todo list for a while. | 17:01 |
| NobodyCam | hummm | 17:01 |
| NobodyCam | set the gateway on 2nd interface and not the first? | 17:02 |
| *** jcooley_ has quit IRC | 17:03 | |
| SpamapS | NobodyCam: gateway? | 17:04 |
| SpamapS | NobodyCam: what I want to prevent is pxe traffic travelling on the data network. | 17:04 |
| SpamapS | I guess | 17:05 |
| SpamapS | really | 17:05 |
| NobodyCam | ahh | 17:05 |
| SpamapS | the server has to never ask for PXE on the data network | 17:05 |
| SpamapS | because that is where the trouble starts ;) | 17:05 |
| *** shausy has quit IRC | 17:05 | |
| NobodyCam | humm, | 17:05 |
| SpamapS | ok n/m I think Ironic is in the clear for the ase that I'm proxying for somebody else ;) | 17:05 |
| SpamapS | case | 17:05 |
| NobodyCam | :) | 17:05 |
| *** shausy has joined #openstack-ironic | 17:05 | |
| SpamapS | NobodyCam: I'm trying ot prevent "evil box observes or mitm's pxe" .. but you can't do that if your boxes are spewing DHCP discovers on the data network. | 17:06 |
| NobodyCam | SpamapS: that I think we will need to think about | 17:06 |
| SpamapS | It really boils down to "openflow" so you can just shut down the provisioning ports when they're not in use. | 17:06 |
| NobodyCam | ya | 17:07 |
| *** aignatov is now known as aignatov_ | 17:07 | |
| NobodyCam | how are you setting the functions of the interfaces | 17:07 |
| SpamapS | also ipxe | 17:07 |
| SpamapS | NobodyCam: preconfig -> tomorrow's problem | 17:08 |
| NobodyCam | ie pif 1 is pxe only pif2 is data only | 17:08 |
| NobodyCam | :-p | 17:08 |
| SpamapS | sits along side raid and firmware | 17:08 |
| NobodyCam | SpamapS: is there a bug / blueprint to address this on the nova BM side? | 17:09 |
| NobodyCam | lucasagomes: LOL .... Because I stole the tests from the libvirt test file and modifyed for our use! <- in ref to 71429 | 17:12 |
| lucasagomes | NobodyCam, ahhh hahahah | 17:13 |
| lucasagomes | NobodyCam, ack, well I thought we would be using mock | 17:13 |
| lucasagomes | at least we should :P | 17:13 |
| NobodyCam | lol ya | 17:14 |
| SpamapS | NobodyCam: nah.. I'm just riffing ;) | 17:14 |
| SpamapS | NobodyCam: getting my head on straight for the impending "can we do this now?" question flood :) | 17:14 |
| *** matty_dubs is now known as matty_dubs|lunch | 17:15 | |
| SpamapS | btw somebody in the know should respond to http://lists.openstack.org/pipermail/openstack-dev/2014-February/026617.html | 17:15 |
| NobodyCam | :) was about to | 17:16 |
| *** martyntaylor has left #openstack-ironic | 17:17 | |
| *** dshulyak has quit IRC | 17:20 | |
| *** ndipanoff has quit IRC | 17:22 | |
| *** lazy_prince has quit IRC | 17:23 | |
| *** martyntaylor has joined #openstack-ironic | 17:24 | |
| max_lobur | lucasagomes, what do you think about this | 17:29 |
| max_lobur | http://paste.openstack.org/show/63075/ | 17:29 |
| max_lobur | this more like a jsonpatch lib bug | 17:30 |
| max_lobur | existing property silently replaced by 'add' op | 17:30 |
| max_lobur | or maybe I missed that in RFC... | 17:30 |
| lucasagomes | max_lobur, yea, it's a jsonpatch thing, I don't remember if that was a problem tho | 17:31 |
| lucasagomes | gotta take a look at the RFC as u said | 17:32 |
| lucasagomes | add will add if the value is not present | 17:32 |
| lucasagomes | but replace with the new value if it's already there | 17:32 |
| max_lobur | this was a member of extra | 17:33 |
| max_lobur | I also just tested on root properties | 17:33 |
| max_lobur | tempest.common.rest_client: INFO: Request: PATCH http://192.168.122.224:6385/v1/chassis/e7f706b2-f5b1-455a-a4b2-9e4f4b637dcd | 17:33 |
| max_lobur | tempest.common.rest_client: DEBUG: Request Headers: {'Content-Type': 'application/json', 'Accept': 'application/json', 'X-Auth-Token': '<Token omitted>'} | 17:33 |
| max_lobur | tempest.common.rest_client: DEBUG: Request Body: [{"path": "/description", "value": "whatever", "op": "add"}] | 17:33 |
| max_lobur | tempest.common.rest_client: INFO: Response Status: 400 | 17:33 |
| max_lobur | tempest.common.rest_client: DEBUG: Response Headers: {'date': 'Fri, 07 Feb 2014 17:33:00 GMT', 'content-length': '173', 'content-type': 'application/json', 'server': 'WSGIServer/0.1 Python/2.7.3'} | 17:33 |
| max_lobur | tempest.common.rest_client: DEBUG: Response Body: {"error_message": "{\"debuginfo\": null, \"faultcode\": \"Client\", \"faultstring\": \"Adding a new attribute (/description) to the root of the resource is not allowed\"}"} | 17:33 |
| lucasagomes | max_lobur, ah, ^ that is our test I think | 17:34 |
| max_lobur | this looks like a normal behavior | 17:34 |
| max_lobur | hmm | 17:34 |
| max_lobur | on Ironic side right? | 17:34 |
| lucasagomes | maybe we should allow add to act like a replace on the attributes that is part of the resource | 17:34 |
| lucasagomes | max_lobur, yes | 17:34 |
| lucasagomes | take a look at | 17:34 |
| lucasagomes | the PatchType custom type for wsme | 17:35 |
| max_lobur | RE: maybe we should allow add to act like a replace on the attributes that is part of the resource | 17:35 |
| lucasagomes | max_lobur, if th RFC says that's allowed I think it's a bug on us | 17:35 |
| max_lobur | need to search this in RFC first :) | 17:35 |
| lucasagomes | max_lobur, yea | 17:35 |
| max_lobur | lucasagomes, + | 17:35 |
| max_lobur | I'll make a note | 17:35 |
| max_lobur | will proceed with this on Monday | 17:35 |
| lucasagomes | max_lobur, https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/types.py#L157 | 17:36 |
| max_lobur | heh | 17:36 |
| max_lobur | indeed | 17:36 |
| max_lobur | k, thanks :) | 17:37 |
| lucasagomes | max_lobur, :) thank u for finding all this! | 17:37 |
| *** anniec has joined #openstack-ironic | 17:39 | |
| NobodyCam | bbt .. brb | 17:39 |
| max_lobur | Folks, pls take a look when you have time https://review.openstack.org/#/c/67854/2 Expanded tempest tests for chassis | 17:45 |
| max_lobur | there will be some failing tests | 17:46 |
| *** killer_prince has joined #openstack-ironic | 17:46 | |
| max_lobur | due to bugs | 17:46 |
| max_lobur | or cause I'm not sure how it should be (like with "add" op) | 17:46 |
| max_lobur | but you already can suggest more scenarios if you see that we missing something :) | 17:46 |
| max_lobur | Have a good rest! bye! | 17:49 |
| *** jcooley_ has joined #openstack-ironic | 17:50 | |
| *** max_lobur is now known as max_lobur_afk | 17:52 | |
| *** jcooley_ has quit IRC | 17:56 | |
| lucasagomes | NobodyCam, there's a new patch-set up and I hope to finish the tests and some refactor in the code at monday tops | 18:08 |
| lucasagomes | and with that, I'm going to enjoy the irish weekend with a cold beer :) | 18:08 |
| NobodyCam | :) I'm working on libvirt refactor and then I push up a new vol driver patch too | 18:08 |
| lucasagomes | NobodyCam, have a good night, enjoy the weekend | 18:08 |
| lucasagomes | NobodyCam, awesome! | 18:08 |
| NobodyCam | have one for me :) | 18:08 |
| lucasagomes | NobodyCam, there's one already | 18:09 |
| lucasagomes | vol driver patch | 18:09 |
| lucasagomes | NobodyCam, https://review.openstack.org/#/c/71026/ | 18:09 |
| NobodyCam | ya.. I push up over that | 18:09 |
| lucasagomes | ah | 18:09 |
| lucasagomes | sure | 18:09 |
| lucasagomes | :) | 18:09 |
| NobodyCam | :) | 18:09 |
| lucasagomes | NobodyCam, I will have one for u for sure | 18:09 |
| lucasagomes | see ya | 18:09 |
| lucasagomes | night everybody | 18:09 |
| NobodyCam | :) | 18:09 |
| *** shausy has quit IRC | 18:10 | |
| *** lucasagomes has quit IRC | 18:11 | |
| *** martyntaylor has left #openstack-ironic | 18:13 | |
| *** matty_dubs|lunch is now known as matty_dubs | 18:13 | |
| *** jcooley_ has joined #openstack-ironic | 18:16 | |
| *** igor_ has quit IRC | 18:17 | |
| *** igor_ has joined #openstack-ironic | 18:18 | |
| *** harlowja_away is now known as harlowja | 18:18 | |
| *** jcooley_ has quit IRC | 18:20 | |
| *** igor_ has quit IRC | 18:24 | |
| *** jcooley_ has joined #openstack-ironic | 18:40 | |
| *** jcooley_ has quit IRC | 18:47 | |
| *** jcooley_ has joined #openstack-ironic | 18:48 | |
| *** igor_ has joined #openstack-ironic | 18:50 | |
| devananda | SpamapS: right now ironic does not differentiate between boot-interfaces and data-interfaces | 18:51 |
| devananda | SpamapS: shouldn't be hard to add, though | 18:51 |
| NobodyCam | Good Mornig devananda :) | 18:52 |
| *** jcooley_ has quit IRC | 18:52 | |
| SpamapS | devananda: I think it is probably moot.. but it came up when talking to some HP folk. | 18:53 |
| *** jcooley_ has joined #openstack-ironic | 18:53 | |
| devananda | SpamapS: *nod* | 18:53 |
| *** shausy has joined #openstack-ironic | 18:55 | |
| *** igor_ has quit IRC | 18:55 | |
| *** anniec has quit IRC | 18:56 | |
| * NobodyCam makes a bagel | 18:56 | |
| *** martyntaylor has joined #openstack-ironic | 18:56 | |
| *** martyntaylor has left #openstack-ironic | 18:56 | |
| *** wusWWW has joined #openstack-ironic | 19:00 | |
| *** jbjohnso has quit IRC | 19:01 | |
| *** epim has joined #openstack-ironic | 19:02 | |
| *** anniec has joined #openstack-ironic | 19:03 | |
| *** jcooley_ has quit IRC | 19:06 | |
| *** jcooley_ has joined #openstack-ironic | 19:13 | |
| *** jbjohnso has joined #openstack-ironic | 19:16 | |
| * devananda has his second cup of coffee | 19:17 | |
| NobodyCam | :) | 19:17 |
| * NobodyCam is on 5 or 6 he thinks | 19:17 | |
| *** shausy has quit IRC | 19:18 | |
| *** anniec has quit IRC | 19:35 | |
| devananda | https://review.openstack.org/#/c/71211/ could use a quick check | 19:36 |
| *** jbjohnso has quit IRC | 19:39 | |
| russell_h | to create a node in ironic, do I need to specify a chassis_uuid? | 19:39 |
| devananda | russell_h: no | 19:40 |
| NobodyCam | ^^ lgtm | 19:41 |
| *** aignatov_ is now known as aignatov | 19:45 | |
| *** igor_ has joined #openstack-ironic | 19:49 | |
| *** mdurnosvistov_ has joined #openstack-ironic | 19:50 | |
| NobodyCam | do you have to rebuild test evn after making changes to rootwrap? | 19:53 |
| *** anniec has joined #openstack-ironic | 19:55 | |
| russell_h | devananda: so is a chassis more like optional metadata? And the provisioning is all based on the parameters passed to the create node call? | 19:56 |
| devananda | russell_h: for now, yes | 19:58 |
| devananda | russell_h: in the initial approach, chassis was mandatory, but we've since realized that more types of grouping need to be supported. rather than ripping out the code which is there, we've made it optional and will expand on it in Juno | 19:59 |
| russell_h | gotcha | 20:00 |
| russell_h | what do you mean by more types of grouping? | 20:00 |
| russell_h | the original source of my interest is I was trying to figure out why the API acts like there might be more than one node per chassis | 20:01 |
| devananda | yep | 20:03 |
| devananda | chassis 1:N nodes | 20:03 |
| devananda | russell_h: chassis could be a rack. or it could be an 8U enclosure. or what ever. | 20:03 |
| NobodyCam | devananda: I thought there was a patch (thou I cann't seem to find it atm) that removed the [required] from here : https://github.com/openstack/python-ironicclient/blob/master/ironicclient/v1/node_shell.py#L67 | 20:04 |
| russell_h | devananda: gotcha, thanks | 20:04 |
| devananda | russell_h: there is interest in modelling multiple failure domains and mapping nodes accordingly to provide better hints to schedulers | 20:04 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Use threading.semaphore instead of lockutils https://review.openstack.org/71974 | 20:04 |
| russell_h | devananda: right, "get me a node in (the same | a different) cabinet" and so on | 20:06 |
| devananda | right | 20:08 |
| *** anniec has quit IRC | 20:18 | |
| *** comstud is now known as bearhands | 20:18 | |
| *** aignatov is now known as aignatov_ | 20:20 | |
| *** aignatov_ is now known as aignatov | 20:26 | |
| devananda | hmm, i would like to stop all the spurious debug messages from lockutils | 20:28 |
| devananda | NobodyCam: oh, it's not required on the server side. the CLI help needs to be updated, though, it seems | 20:28 |
| devananda | NobodyCam: "ironic node-create -d fake" works with no chassis-id | 20:29 |
| NobodyCam | ya just the doc string | 20:30 |
| devananda | ya | 20:30 |
| devananda | i'm waffling | 20:31 |
| NobodyCam | on requiring chassis? | 20:31 |
| devananda | the patch i just posted, 71974, doesn't change the functionality in any meaningful way afaict. but it stops all the LOG.debug spam | 20:31 |
| NobodyCam | lol | 20:31 |
| NobodyCam | what log are you getting spam in? | 20:32 |
| devananda | debug | 20:32 |
| devananda | NobodyCam: http://paste.openstack.org/show/njde6BKP7gosiWAmMsaW/ | 20:33 |
| NobodyCam | can filter it out in logging-* conf | 20:33 |
| devananda | no - devug is good | 20:33 |
| devananda | but this mkes it unreadable | 20:33 |
| devananda | every time task_manager.acquire is called, we get taht whole series of debug messages | 20:33 |
| *** openstackgerrit has quit IRC | 20:34 | |
| devananda | it's useful in other situations, eg. around instantiating a drver instance | 20:34 |
| *** openstackgerrit has joined #openstack-ironic | 20:34 | |
| *** igor___ has joined #openstack-ironic | 20:34 | |
| devananda | oslo lockutils isn't meant to be called 100's of times per second | 20:34 |
| devananda | or even 10's | 20:34 |
| devananda | oooh, duh. | 20:35 |
| NobodyCam | add to https://github.com/openstack/ironic/blob/master/etc/ironic/ironic.conf.sample#L159 | 20:36 |
| *** igor_ has quit IRC | 20:36 | |
| *** anniec has joined #openstack-ironic | 20:36 | |
| NobodyCam | brb | 20:39 |
| devananda | i need to move my car, but i think ResourceManager is being naive about its locking | 20:39 |
| devananda | going to toss up an improvement in a bit | 20:39 |
| devananda | bbiab | 20:39 |
| *** anniec has quit IRC | 20:42 | |
| *** jbjohnso has joined #openstack-ironic | 20:43 | |
| *** mdurnosvistov_ has quit IRC | 20:53 | |
| *** aignatov is now known as aignatov_ | 20:59 | |
| *** anniec has joined #openstack-ironic | 21:03 | |
| NobodyCam | quick walkies... | 21:11 |
| openstackgerrit | Michael Davies proposed a change to openstack/ironic: Ensure parameter order of assertEqual correct https://review.openstack.org/71486 | 21:27 |
| openstackgerrit | Michael Davies proposed a change to openstack/ironic: Ensure parameter order of assertEqual correct https://review.openstack.org/71486 | 21:28 |
| *** anniec has quit IRC | 21:32 | |
| NobodyCam | devananda: https://review.openstack.org/#/c/72007/1 | 21:33 |
| devananda | i'm not happy with ersource_manager.py | 21:39 |
| *** anniec has joined #openstack-ironic | 21:41 | |
| devananda | NobodyCam: lgtm | 21:41 |
| NobodyCam | :-p | 21:42 |
| NobodyCam | getting there | 21:42 |
| devananda | so now that we've got a lot more things doing stuff, i'm thinking about how taskmanager and resorucemanager are actually being used | 21:42 |
| devananda | vs what i initialy thought | 21:42 |
| devananda | and really unhappy with the locking that's going on | 21:42 |
| devananda | resourcemanager is doing waay too much unnecessary locking | 21:43 |
| NobodyCam | gah H306 imports not in alphabetical order | 21:43 |
| NobodyCam | :-p | 21:43 |
| devananda | lol | 21:43 |
| NobodyCam | gota check my local flake8 settings | 21:44 |
| *** john-n-seattle has quit IRC | 21:44 | |
| *** john-n-seattle1 has joined #openstack-ironic | 21:48 | |
| *** jcooley_ has quit IRC | 21:48 | |
| NobodyCam | devananda: re the locking ... thought was prob better safe then sorry / easier to remove (in most cases) then add :-p | 21:48 |
| NobodyCam | :-/ | 21:48 |
| devananda | max_lobur_afk: i may make smoe changes to task_manager that conflict with your changes. i will help with refactoring if needed | 21:48 |
| devananda | NobodyCam: yea | 21:48 |
| devananda | NobodyCam: so this is very pessimistic, heh. there's a process-wide lock around entering any task, and a DB query is done inside that lock. | 21:49 |
| *** jcooley_ has joined #openstack-ironic | 21:49 | |
| devananda | NobodyCam: essentially, everything could stall if one DB query gets dropped | 21:49 |
| NobodyCam | oh, I would expect several tasks to fail if the db get dropped out from under a running app | 21:50 |
| devananda | err, yes | 21:50 |
| *** romcheg1 has joined #openstack-ironic | 21:51 | |
| *** romcheg_ltp has quit IRC | 21:51 | |
| NobodyCam | but surlly we dont need to lock for things like get power state | 21:51 |
| devananda | NobodyCam: well, we do | 21:56 |
| devananda | NobodyCam: but not right now | 21:57 |
| NobodyCam | to ensure another conductor is not talking to our node | 21:57 |
| devananda | NobodyCam: in principle, there are situations where we should respect $other_task and not tickle IPMI at all | 21:57 |
| devananda | NobodyCam: but taht's a differnt type of lock | 21:58 |
| NobodyCam | if node locked return db.value for $request ? | 21:58 |
| devananda | NobodyCam: that's the difference between a shared & exclusive lock. get_power_state should b a shared lock | 21:58 |
| devananda | as should validate, etc | 21:58 |
| devananda | hmm | 21:58 |
| NobodyCam | humm? | 21:59 |
| devananda | ok, we're goign to eventually need 3 types of lock: 1) i'm touching but not changing the node 2) i'm changing the node, no one else can change it now 3) i'm changing the node, no one else can touch it now | 22:00 |
| devananda | >1 of (1) can run | 22:00 |
| devananda | >1 of (1) can run while exactly 1 of (2) runs, but no more than 1 of (2) can ever run at a time | 22:00 |
| *** anniec has quit IRC | 22:00 | |
| devananda | if a (3) wants to start, it needs to block the creation of new (1) nd (2), and wait for any existing (1) and (2) to finish | 22:01 |
| devananda | right now we don't support (3) | 22:01 |
| devananda | which is fine -- afaik, that's only needed for firmware udpate | 22:01 |
| devananda | but, i side track. the semaphore around resouce_manager has nothing to dowith this type of locking | 22:02 |
| NobodyCam | kinda wait I have the lock... get in line behind me | 22:02 |
| NobodyCam | a queue of who gets the lock | 22:02 |
| devananda | i don't thnk we're actally getting any benefit from resource_manager right now, aside from bundling all the ports associated with a node | 22:02 |
| *** matty_dubs is now known as matty_dubs|gone | 22:03 | |
| *** jdob_ has quit IRC | 22:07 | |
| *** jrist has quit IRC | 22:18 | |
| *** epim has quit IRC | 22:28 | |
| *** jrist has joined #openstack-ironic | 22:30 | |
| *** jcooley_ has quit IRC | 22:31 | |
| *** epim has joined #openstack-ironic | 22:32 | |
| *** anniec has joined #openstack-ironic | 22:38 | |
| *** anniec has quit IRC | 22:41 | |
| *** anniec has joined #openstack-ironic | 22:43 | |
| NobodyCam | humm my dependicy didn't get added to the volume review.... oh but it already a dep on another patch | 22:44 |
| *** rwsu has quit IRC | 23:01 | |
| *** jcooley_ has joined #openstack-ironic | 23:01 | |
| *** rwsu has joined #openstack-ironic | 23:02 | |
| *** rwsu has quit IRC | 23:06 | |
| *** jcooley_ has quit IRC | 23:06 | |
| *** rwsu has joined #openstack-ironic | 23:19 | |
| devananda | i must have brokensomething... this is acting all single threaded even after I remove all the locks that I can find | 23:21 |
| devananda | yep. i issue parallel requests for DIFFERENT nodes, and even that gets serialized | 23:25 |
| *** wusWWW has quit IRC | 23:26 | |
| *** anniec has quit IRC | 23:31 | |
| NobodyCam | :-p | 23:32 |
| *** epim has quit IRC | 23:34 | |
| devananda | looks like it is serialized in the API layer | 23:37 |
| devananda | wtf | 23:37 |
| *** anniec has joined #openstack-ironic | 23:40 | |
| * NobodyCam goes for walkies bbiaf | 23:43 | |
| *** anniec has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!