| *** BadCub is now known as BadCub_Hiding | 00:01 | |
| JayF | NobodyCam: They've not even been reviewed | 00:01 |
|---|---|---|
| NobodyCam | JayF: I +1'd them | 00:01 |
| JayF | if devananda can shake a stick at people to review it, with his magic ptl power, that'd be great | 00:01 |
| NobodyCam | lol | 00:01 |
| JayF | NobodyCam: you're not a core reviewer == that's moral support | 00:01 |
| JayF | LOL | 00:01 |
| NobodyCam | g'night ya'll | 00:01 |
| devananda | NobodyCam: g'night! | 00:02 |
| devananda | (says the man wide awake at 1am) | 00:02 |
| BadCub_Hiding | JayF: let me know which ones and I will get ya some reviews tomorrow *WG* | 00:03 |
| JoshNang | BadCub_Hiding: https://review.openstack.org/#/c/168175/ | 00:04 |
| JoshNang | disables cleaning the disk in devstack by default (takes forever) and bumps the delete timeout up to give cleaning more than 60 secs to finish | 00:05 |
| * BadCub_Hiding puts on pad | 00:06 | |
| JayF | Does anyone know how this internal rst linking stuff works? | 00:06 |
| JayF | like if I'm writing cleaning docs, and want to link to the state machine doc, should I just use a full hyperlink or what? | 00:07 |
| devananda | no. because the docs must be servable from any site. eg, our doc test builds. o rmy laptop | 00:07 |
| JayF | That was the impression I had, just wasn't 100% sure from rst docs the right way to do an indirect link | 00:08 |
| devananda | JayF: `words <rel/path/to/file.html>`_ | 00:08 |
| devananda | try that | 00:08 |
| *** kkoski has joined #openstack-ironic | 00:09 | |
| BadCub_Hiding | okay, now I am off to dinner | 00:09 |
| JayF | http://docutils.sourceforge.net/docs/user/rst/quickref.html#hyperlink-targets makes me think I should just do | 00:09 |
| JayF | states_ | 00:09 |
| JayF | which seems ... wrong | 00:09 |
| JayF | and also ugly | 00:09 |
| NobodyCam | JayF: also see the edits to the readme in ipa you just landed | 00:09 |
| NobodyCam | lol | 00:09 |
| devananda | i think that works within the document, or uses some sphinx auto magic | 00:09 |
| * NobodyCam *ducks out the door* | 00:10 | |
| JayF | NobodyCam: I don't see any? | 00:10 |
| JayF | NobodyCam: there's not any patchsets pushed to it | 00:10 |
| NobodyCam | https://github.com/openstack/ironic-python-agent/commit/c2a1f5e3d4f095179e691561ad2e7055d7773cc9 line 18 | 00:11 |
| NobodyCam | :-p | 00:11 |
| JayF | Oh, see, that doesn't answer my ? though | 00:12 |
| JayF | this is an internal link | 00:12 |
| JayF | I'm going to use the relative link idea devananda had | 00:12 |
| JayF | devananda: btw, JoshNang and I saw the iLo driver docs and felt a large quantity of shame | 00:12 |
| JayF | devananda: I think adding agent driver docs is more on our radar than ever before | 00:13 |
| devananda | JayF: :) | 00:14 |
| devananda | JayF: I do not feel even a tinge of guilt that you feel shame for that | 00:15 |
| devananda | JayF: also, the agent and its driver rock. you should be proud | 00:15 |
| *** ijw_ has joined #openstack-ironic | 00:16 | |
| devananda | the agent's a major component of the project. I would like the doc quality to reflect that | 00:17 |
| *** ijw__ has joined #openstack-ironic | 00:18 | |
| *** ijw has quit IRC | 00:19 | |
| *** ijw_ has quit IRC | 00:21 | |
| devananda | JayF: however, I should be ashamed that I'm using and talking about the agent, but dont understand it well enough to go review it meaningfully | 00:23 |
| *** mtanino has quit IRC | 00:24 | |
| *** naohirot has joined #openstack-ironic | 00:25 | |
| *** davide_ has quit IRC | 00:26 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Add documentation for Cleaning https://review.openstack.org/168189 | 00:28 |
| JayF | rloo: ^ We'd love an epic-Ruby-docs-review on that, if you have the time to spare :D | 00:29 |
| rloo | huh? oh (booking my flight to summit now). did you need that reviewed this week or can it wait til monday? | 00:30 |
| JayF | It's not urgent, just all the docs for cleaning. You review documentation very well, and I appreciate your perspective :) | 00:30 |
| *** anderbubble has quit IRC | 00:31 | |
| *** igordcard_ has quit IRC | 00:34 | |
| *** kkoski has quit IRC | 00:34 | |
| *** kkoski has joined #openstack-ironic | 00:41 | |
| *** anderbubble has joined #openstack-ironic | 00:43 | |
| JoshNang | JayF: gah it still timed out. | 00:46 |
| rloo | JayF: I took a quick look; it will take me some time to review that doc and my head isn't into work right now, so probably best if I look at it later. Although I think my priority may be reviewing code for rc1 so I may not get to that til later. But i'll look even after it gets merged ;) | 00:48 |
| *** anderbubble has quit IRC | 00:59 | |
| *** stendulker has joined #openstack-ironic | 01:04 | |
| *** jmank has quit IRC | 01:04 | |
| *** jmank has joined #openstack-ironic | 01:05 | |
| *** dtantsur|afk has quit IRC | 01:05 | |
| *** dtantsur has joined #openstack-ironic | 01:10 | |
| *** derekh has quit IRC | 01:17 | |
| openstackgerrit | Josh Gachnang proposed openstack/ironic: Fix cleaning nits https://review.openstack.org/168197 | 01:18 |
| *** jmccrory has joined #openstack-ironic | 01:18 | |
| *** jmccrory has quit IRC | 01:18 | |
| *** stendulker has quit IRC | 01:21 | |
| *** anderbubble has joined #openstack-ironic | 01:21 | |
| *** ChuckC has joined #openstack-ironic | 01:26 | |
| *** kkoski has quit IRC | 01:31 | |
| *** anderbubble has quit IRC | 01:35 | |
| *** ChuckC has quit IRC | 01:36 | |
| *** penick has quit IRC | 01:53 | |
| *** dank_ has quit IRC | 01:55 | |
| *** rloo has quit IRC | 02:14 | |
| *** oomichi_ has quit IRC | 02:17 | |
| *** jerryz has joined #openstack-ironic | 02:19 | |
| *** achanda has quit IRC | 02:27 | |
| *** stendulker has joined #openstack-ironic | 02:31 | |
| *** kkoski has joined #openstack-ironic | 02:37 | |
| *** ramineni has joined #openstack-ironic | 02:40 | |
| *** chenglch has joined #openstack-ironic | 02:40 | |
| *** stendulker has quit IRC | 02:47 | |
| *** devlaps has quit IRC | 02:54 | |
| *** ijw__ has quit IRC | 03:06 | |
| *** kkoski has quit IRC | 03:09 | |
| *** kkoski has joined #openstack-ironic | 03:10 | |
| *** achanda has joined #openstack-ironic | 03:13 | |
| *** kkoski has quit IRC | 03:17 | |
| *** atuvenie has joined #openstack-ironic | 03:25 | |
| *** Marga_ has quit IRC | 03:26 | |
| *** chenglch has quit IRC | 03:37 | |
| *** chenglch has joined #openstack-ironic | 03:37 | |
| *** anderbubble has joined #openstack-ironic | 03:41 | |
| *** saripurigopi has joined #openstack-ironic | 03:53 | |
| *** atuvenie has quit IRC | 04:06 | |
| *** yuanying has joined #openstack-ironic | 04:09 | |
| *** rameshg87 has joined #openstack-ironic | 04:15 | |
| rameshg87 | good morning ironic | 04:15 |
| openstackgerrit | Merged stackforge/proliantutils: Fallback to RIBCL https://review.openstack.org/167151 | 04:37 |
| Haomeng|2 | rameshg87: ramesh, good morning:) | 04:39 |
| rameshg87 | hello Haomeng|2 | 04:39 |
| rameshg87 | morning | 04:39 |
| Haomeng|2 | rameshg87: :( | 04:39 |
| Haomeng|2 | rameshg87: :) | 04:39 |
| *** achanda has quit IRC | 05:00 | |
| *** jmanko has joined #openstack-ironic | 05:01 | |
| *** jmank has quit IRC | 05:05 | |
| *** Marga_ has joined #openstack-ironic | 05:14 | |
| *** yog__ has joined #openstack-ironic | 05:19 | |
| *** stendulker has joined #openstack-ironic | 05:31 | |
| *** jiangfei has quit IRC | 05:35 | |
| *** achanda has joined #openstack-ironic | 05:37 | |
| *** jamielennox is now known as jamielennox|away | 05:41 | |
| openstackgerrit | Merged openstack/python-ironicclient: Add support for generating a config drive https://review.openstack.org/167725 | 05:56 |
| *** Marga_ has quit IRC | 06:02 | |
| openstackgerrit | jiangfei proposed openstack/python-ironicclient: Encode exception on cli for UnicodeDecodeError https://review.openstack.org/153155 | 06:06 |
| *** saripurigopi has quit IRC | 06:06 | |
| *** jiangfei has joined #openstack-ironic | 06:10 | |
| openstackgerrit | Merged openstack/ironic: Ensure configdrive isn't mounted for ipxe/elilo https://review.openstack.org/167449 | 06:21 |
| openstackgerrit | Merged openstack/ironic: Attach ilo_boot_iso only if node is active https://review.openstack.org/167317 | 06:22 |
| *** saripurigopi has joined #openstack-ironic | 06:26 | |
| *** achanda has quit IRC | 06:32 | |
| *** anderbubble has quit IRC | 06:37 | |
| openstackgerrit | Andreas Jaeger proposed openstack/ironic-lib: Add .gitreview https://review.openstack.org/168240 | 06:45 |
| *** AJaeger has joined #openstack-ironic | 06:45 | |
| AJaeger | Hi ironic team, could you review my patch https://review.openstack.org/168240, please? Without the patch, our global requirements jobs err out | 06:46 |
| Haomeng|2 | AJaeger: sure, thanks for your patch | 06:47 |
| AJaeger | Haomeng|2: thanks! The patch is for ironic-lib and part of the basic setup that each repo needs | 06:48 |
| Haomeng|2 | AJaeger: not sure the background why we add this file for now | 06:49 |
| AJaeger | Haomeng|2: http://docs.openstack.org/infra/manual/creators.html#configure-git-review | 06:49 |
| Haomeng|2 | AJaeger: ok | 06:49 |
| AJaeger | That one was missed when setting up ironic-lib initially ;( | 06:49 |
| AJaeger | Ah, we need to go with https://review.openstack.org/#/c/162162/ instead - pep8 fails on the lib | 06:50 |
| AJaeger | and 162162 fixed that one | 06:50 |
| Haomeng|2 | AJaeger: this is new change, right? | 06:50 |
| Haomeng|2 | AJaeger: ok | 06:50 |
| Haomeng|2 | AJaeger: got, thanks for your information | 06:50 |
| AJaeger | 162162 is an old change, .gitreview is an old requirement ;) | 06:51 |
| Haomeng|2 | AJaeger: ok | 06:51 |
| AJaeger | I abandoned mine and commented on 162162 | 06:53 |
| Haomeng|2 | AJaeger: ok, that is more clear:) | 06:54 |
| *** bradjones has quit IRC | 06:55 | |
| *** pradipta has joined #openstack-ironic | 06:55 | |
| AJaeger | Let me fix the commit message since that one is not clear | 06:55 |
| openstackgerrit | Andreas Jaeger proposed openstack/ironic-lib: Add .gitreview file and some minor changes https://review.openstack.org/162162 | 06:56 |
| *** bradjones has joined #openstack-ironic | 06:56 | |
| AJaeger | Haomeng|2: hope it's clear enough for a +2 now ;) | 06:56 |
| *** bradjones has quit IRC | 06:57 | |
| *** bradjones has joined #openstack-ironic | 06:57 | |
| Haomeng|2 | AJaeger: ok | 06:58 |
| Haomeng|2 | AJaeger: :) | 06:58 |
| Haomeng|2 | AJaeger: I reviewed your code https://review.openstack.org/#/c/162162/2, looks like the unittest does not cover the change, only last line is removed that is the change I found in the test file | 07:10 |
| *** chlong has quit IRC | 07:16 | |
| Haomeng|2 | AJaeger: others looks good to me, thanks for your patch | 07:18 |
| Haomeng|2 | AJaeger: :) | 07:19 |
| *** ndipanov has joined #openstack-ironic | 08:00 | |
| *** pradipta has quit IRC | 08:06 | |
| *** zhenguo has joined #openstack-ironic | 08:09 | |
| openstackgerrit | jiangfei proposed openstack/ironic: Sync sslutils to ironic-api for support HTTPS https://review.openstack.org/168261 | 08:14 |
| *** stendulker has quit IRC | 08:15 | |
| AJaeger | thanks, Haomeng|2 | 08:21 |
| Haomeng|2 | AJaeger: welcome :) | 08:21 |
| *** ifarkas has joined #openstack-ironic | 08:28 | |
| *** JoshNang has quit IRC | 08:32 | |
| *** enikanorov has joined #openstack-ironic | 08:36 | |
| *** jistr has joined #openstack-ironic | 08:40 | |
| *** andreykurilin_ has joined #openstack-ironic | 08:43 | |
| *** jerryz_ has joined #openstack-ironic | 08:45 | |
| *** jerryz has quit IRC | 08:48 | |
| dtantsur | Morning, folks, TGIF today :) | 08:51 |
| openstackgerrit | Haomeng,Wang proposed openstack/ironic: Fix UnicodeEncodeError issue when the language is not en_US https://review.openstack.org/168268 | 08:55 |
| AJaeger | dtantsur: Ok, I revive my patch and see whether we can do something smaller... | 08:57 |
| dtantsur | AJaeger, IIUC you only need to add missing empty line to tests file :) | 08:58 |
| AJaeger | dtantsur: if that's all, I'm for it ;) | 08:58 |
| dtantsur | and g'morning/afternoon/whatever :) | 08:58 |
| AJaeger | dtantsur: morning for me ;) And morning to you as well ;) | 08:58 |
| dtantsur | :) | 08:59 |
| * AJaeger runs pep8 locally now... | 08:59 | |
| openstackgerrit | Andreas Jaeger proposed openstack/ironic-lib: Add .gitreview https://review.openstack.org/168240 | 09:02 |
| AJaeger | dtantsur, Haomeng|2: The above should be fine now, I can't make it smaller ;) | 09:02 |
| openstackgerrit | jiangfei proposed openstack/ironic: Add support for HTTPS to the Ironic API https://review.openstack.org/163735 | 09:02 |
| * dtantsur is waiting for check job to finish | 09:03 | |
| AJaeger | dtantsur: Sure! | 09:03 |
| *** atuvenie has joined #openstack-ironic | 09:10 | |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: RIS:Add get_pending_boot_mode https://review.openstack.org/168270 | 09:13 |
| *** pradipta has joined #openstack-ironic | 09:13 | |
| *** pradipta has quit IRC | 09:14 | |
| *** pradipta has joined #openstack-ironic | 09:14 | |
| *** lucasagomes has joined #openstack-ironic | 09:14 | |
| *** andreykurilin_ has quit IRC | 09:19 | |
| *** andreykurilin_ has joined #openstack-ironic | 09:19 | |
| AJaeger | dtantsur: check job is finished | 09:20 |
| dtantsur | Haomeng|2, wanna have a look before I press the button? | 09:21 |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: Fix syntax error in continue_cleaning https://review.openstack.org/168273 | 09:23 |
| dtantsur | Haomeng|2, AJaeger, ok I proceed with approving to unblock #infra | 09:25 |
| rameshg87 | lucasagomes: hi | 09:26 |
| openstackgerrit | Merged openstack/ironic-lib: Add .gitreview https://review.openstack.org/168240 | 09:27 |
| AJaeger | thanks, dtantsur | 09:27 |
| lucasagomes | rameshg87, hello there | 09:28 |
| rameshg87 | lucasagomes: do you remember i talked about not able to ssh to the instance that was created with config drive ? | 09:29 |
| rameshg87 | lucasagomes: after much of struggle i figured out | 09:29 |
| lucasagomes | rameshg87, hmm oh what was that? | 09:29 |
| rameshg87 | lucasagomes: we needed to enable ConfigDrive /etc/cloud/cloud.cfg.d/91-dib-cloud-init-datasources.cfg | 09:29 |
| rameshg87 | lucasagomes: in the built image | 09:29 |
| rameshg87 | lucasagomes: by default dib doesn't add it | 09:29 |
| lucasagomes | oh, did you open a bug against dib? | 09:30 |
| rameshg87 | lucasagomes: it has only EC2 and None as the data sources | 09:30 |
| rameshg87 | lucasagomes: dib provides an element to do that configuration | 09:30 |
| lucasagomes | I see | 09:30 |
| rameshg87 | lucasagomes: https://github.com/openstack/diskimage-builder/tree/master/elements/cloud-init-datasources | 09:30 |
| lucasagomes | hah damn, well good you found out | 09:30 |
| rameshg87 | lucasagomes: i think we should document that | 09:30 |
| rameshg87 | lucasagomes: in ironic config drive documentation | 09:30 |
| lucasagomes | I don't know much how is the documentation in dib | 09:30 |
| lucasagomes | rameshg87, +1 | 09:31 |
| lucasagomes | yeah we can put it at the user guide | 09:31 |
| dtantsur | yeah, and in "image building" section too | 09:31 |
| lucasagomes | dtantsur, +1 | 09:31 |
| rameshg87 | yeah | 09:31 |
| lucasagomes | rameshg87, good catch :) | 09:31 |
| rameshg87 | i guess our config drive documentation is submitted | 09:31 |
| rameshg87 | i will add it to our documentation | 09:32 |
| rameshg87 | and lucasagomes one more issue i figured out | 09:32 |
| rameshg87 | fc20 comes with cloud-init 0.7.2 | 09:32 |
| rameshg87 | that doesn't support config drive on partitions | 09:32 |
| lucasagomes | rameshg87, I will update the documentation for configdrive now that it's merged on the client | 09:32 |
| lucasagomes | to reflect those changes | 09:33 |
| lucasagomes | oh seriously> | 09:33 |
| lucasagomes | ?* | 09:33 |
| rameshg87 | lucasagomes: that's why my fedora image wasn't working | 09:33 |
| rameshg87 | lucasagomes: 1 sec, sharing the code | 09:33 |
| lucasagomes | odd because we use RHEL and CentOS for some tests and AFAICT that's based on f20 | 09:33 |
| lucasagomes | and it does work | 09:33 |
| lucasagomes | right I can take a look | 09:33 |
| *** derekh has joined #openstack-ironic | 09:33 | |
| lucasagomes | 1 sec lemme grab my coffee | 09:33 |
| dtantsur | Version : 0.7.6 | 09:34 |
| dtantsur | that's in my RHEL env ^^^ | 09:34 |
| dtantsur | rameshg87, it this ^^^ enough? | 09:35 |
| rameshg87 | dtantsur: ah okay | 09:35 |
| rameshg87 | dtantsur: that works | 09:35 |
| rameshg87 | 0.7.6 is latest and is fixed | 09:35 |
| rameshg87 | copy-pasting code from 0.7.2 | 09:35 |
| rameshg87 | 1 sec | 09:35 |
| dtantsur | not sure if we can fix F20 now... I guess we should be moving to F21 | 09:35 |
| rameshg87 | cloudinit is not on github | 09:36 |
| rameshg87 | but this is an unofficial one of 0.7.6 | 09:36 |
| rameshg87 | lucasagomes: dtantsur: https://github.com/number5/cloud-init/blob/master/cloudinit/sources/DataSourceConfigDrive.py#L265-L273 | 09:36 |
| rameshg87 | see the highlighted line in def find_candidate_devs | 09:37 |
| rameshg87 | lucasagomes: dtantsur: we are supporting partitions as configdrive in 0.7.6 | 09:37 |
| dtantsur | I see... | 09:37 |
| rameshg87 | lucasagomes: dtantsur: but see paste of same method from cloudinit 0.7.2 - http://paste.openstack.org/show/197013/ | 09:37 |
| rameshg87 | see line 34,35 | 09:38 |
| rameshg87 | ignore partitions :( | 09:38 |
| lucasagomes | rameshg87, oh, yeah perhaps we should add a note on the documentation about it | 09:38 |
| dtantsur | F21 has 0.7.5 | 09:38 |
| dtantsur | this isn't enough, is it? | 09:38 |
| lucasagomes | > 0.7.2 | 09:38 |
| rameshg87 | lucasagomes: not sure >0.7.2 is enough | 09:38 |
| rameshg87 | lucasagomes: don't know when this change was made - 0.7.3, 0.7.4 or 0.7.5 | 09:39 |
| lucasagomes | gotcha | 09:39 |
| lucasagomes | :( | 09:39 |
| lucasagomes | I will try to find out, for precaution we can alert about >=0.7.6 at the documentation | 09:39 |
| rameshg87 | yeah | 09:40 |
| dtantsur | https://github.com/number5/cloud-init/blob/master/ChangeLog | 09:40 |
| rameshg87 | FC20 has only 0.7.2 | 09:40 |
| rameshg87 | dtantsur: ah good | 09:40 |
| rameshg87 | 0.7.5 | 09:40 |
| rameshg87 | https://github.com/number5/cloud-init/blob/master/ChangeLog#L93-L94 | 09:40 |
| rameshg87 | so it should be >=0.7.5 | 09:41 |
| dtantsur | yeah, also see it | 09:41 |
| dtantsur | lucasagomes, ^^^ | 09:41 |
| lucasagomes | https://launchpad.net/cloud-init/+milestone/0.7.6 | 09:41 |
| lucasagomes | ah awesome | 09:41 |
| lucasagomes | so 0.7.5 | 09:41 |
| rameshg87 | yeah | 09:41 |
| * lucasagomes feels a bit better heh | 09:41 | |
| rameshg87 | :) | 09:41 |
| rameshg87 | so two things | 09:41 |
| dtantsur | so F21 is fine | 09:41 |
| *** AJaeger has quit IRC | 09:41 | |
| rameshg87 | cloudinit>=0.7.5 and enable ConfigDrive | 09:42 |
| lucasagomes | since f20 still supported IIRC we may open a bug about it | 09:42 |
| rameshg87 | yeah i think we should | 09:42 |
| rameshg87 | lucasagomes: i checked with yum update | 09:42 |
| dtantsur | not sure it will be updated to a feature release at this point... | 09:42 |
| dtantsur | but we may try | 09:42 |
| rameshg87 | lucasagomes: it's not showing cloud-init >= 0.7.2 | 09:42 |
| lucasagomes | :( yeah | 09:42 |
| lucasagomes | well I will see internally how it would work | 09:42 |
| rameshg87 | okay | 09:43 |
| rameshg87 | lucasagomes: and dib by default build with fc20 | 09:43 |
| lucasagomes | but f21 support it is a relief for sure | 09:43 |
| rameshg87 | *builds | 09:43 |
| lucasagomes | oh perhaps that should be updated too | 09:43 |
| rameshg87 | disk-image-create fedora | 09:43 |
| dtantsur | our folks have been working on switching to F21 | 09:43 |
| lucasagomes | too many moving parts | 09:43 |
| rameshg87 | gives me fc20 based cloud image | 09:43 |
| lucasagomes | I see | 09:43 |
| dtantsur | lucasagomes, Ian Wienand is the name of the guy | 09:43 |
| lucasagomes | oh ian, ok | 09:44 |
| lucasagomes | nice, he's a sharp guy | 09:44 |
| dtantsur | https://review.openstack.org/#/c/140889/ | 09:44 |
| dtantsur | and so on | 09:44 |
| lucasagomes | cool | 09:44 |
| dtantsur | lucasagomes, we have an internal email status updates from him on this topic | 09:44 |
| dtantsur | from time to time | 09:44 |
| lucasagomes | I see, so many internal emails I can't read 1/10 of it | 09:45 |
| dtantsur | I don't even try, just read when subject is intriguing :D | 09:45 |
| lucasagomes | yeah | 09:45 |
| *** foexle has joined #openstack-ironic | 09:46 | |
| *** andreykurilin_ has quit IRC | 09:49 | |
| *** pas-ha has joined #openstack-ironic | 09:50 | |
| openstackgerrit | Harshada Mangesh Kakad proposed openstack/ironic: Update pythonseamicroclient package version https://review.openstack.org/168277 | 09:50 |
| *** naohirot has quit IRC | 09:53 | |
| *** chenglch has quit IRC | 09:53 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: pxe deploy fails for whole disk images in UEFI https://review.openstack.org/167882 | 09:54 |
| *** pelix has joined #openstack-ironic | 10:00 | |
| *** Haomeng has joined #openstack-ironic | 10:01 | |
| lucasagomes | rameshg87, added a comment on #168030 | 10:03 |
| lucasagomes | re the 2 acquires, they are different | 10:03 |
| *** Haomeng|2 has quit IRC | 10:04 | |
| rameshg87 | lucasagomes: oh okay | 10:06 |
| rameshg87 | didn't notice that .. | 10:07 |
| rameshg87 | lucasagomes: anyway to make that atomic ? | 10:07 |
| lucasagomes | yeah I probably will add a comment about it | 10:07 |
| rameshg87 | lucasagomes: anyway to make that operation atomic - either we do everything or nothing ? | 10:07 |
| rameshg87 | lucasagomes: there is a chance that we might endup doing half of it no ? | 10:07 |
| lucasagomes | rameshg87, yeah I will raise an exc if we can't stop the console | 10:08 |
| lucasagomes | so if we don't stop it we don't update | 10:08 |
| rameshg87 | lucasagomes: i meant if we were able to stop it but before we got NodeLocked while trying to acquire second lock | 10:08 |
| rameshg87 | lucasagomes: means we stopped the console, but didn't change driver | 10:08 |
| rameshg87 | should be still fine, is it ? | 10:08 |
| lucasagomes | oh gotcha | 10:09 |
| lucasagomes | hmmmmmm | 10:09 |
| lucasagomes | yeah there's a race between locks | 10:09 |
| rameshg87 | might be fine because we are not doing any partial update on node - it's just stopping console in first operation | 10:10 |
| rameshg87 | but better if it was atomic | 10:10 |
| lucasagomes | rameshg87, I think I can get the lock with the old driver | 10:11 |
| lucasagomes | then in order to check if the new driver exist in a conductor in the cluster I don't need to get another lock with it | 10:11 |
| rameshg87 | and then switch the driver of task with the lock ? | 10:11 |
| lucasagomes | I can simple check it | 10:11 |
| rameshg87 | okay | 10:11 |
| rameshg87 | brb | 10:11 |
| *** rameshg87 is now known as rameshg87-brb | 10:12 | |
| *** rameshg87-brb is now known as rameshg87 | 10:16 | |
| *** pradipta has quit IRC | 10:17 | |
| rameshg87 | lucasagomes: yeah, i think we were just saving the node object within second task block, so should be fine if we used old driver also, right ? | 10:17 |
| rameshg87 | idk | 10:17 |
| lucasagomes | yeah I think it's fine, we just need to check if the new driver exist in a ocnductor | 10:17 |
| lucasagomes | conductor* | 10:17 |
| lucasagomes | I will try to figure out a good way | 10:17 |
| rameshg87 | okay | 10:17 |
| rameshg87 | lucasagomes: but if it didn't exist, the rpc wouldn't have come no ? | 10:18 |
| rameshg87 | lucasagomes: the hash ring wouldn't have showed that conductor for the driver | 10:18 |
| lucasagomes | rameshg87, but the update RPC message was sent to the old driver | 10:19 |
| rameshg87 | lucasagomes: oh | 10:19 |
| rameshg87 | sorry :( | 10:19 |
| lucasagomes | so the RPC message comes to the conductor that is current owning the node, and then it tries to acquire a lock with a conductor that can control the new driver | 10:19 |
| rameshg87 | lucasagomes: so node-update changing driver can fail occassionaly ? | 10:19 |
| lucasagomes | rameshg87, yeah, well it's a sync operation | 10:20 |
| rameshg87 | lucasagomes: no i meant | 10:20 |
| lucasagomes | so you would know if there's no conductor that support the new driver up and running | 10:20 |
| rameshg87 | lucasagomes: unless i have both (old driver, new driver) enabled on all conductor it *can* fail sometimes | 10:20 |
| lucasagomes | rameshg87, oh... yeah changing driver can fail occasionaly yes | 10:20 |
| lucasagomes | yeah | 10:20 |
| rameshg87 | hehe | 10:21 |
| rameshg87 | strange :D | 10:21 |
| lucasagomes | shouldn't it? | 10:21 |
| rameshg87 | yeah, it should :) | 10:22 |
| rameshg87 | but we could do hash-ring rebalance at the time of node-update driver | 10:22 |
| rameshg87 | so that the new node is mapped to a conductor which has both the drivers :) | 10:22 |
| rameshg87 | too expensive may be for one node :) | 10:22 |
| lucasagomes | right, yeah | 10:23 |
| rameshg87 | lucasagomes: but it's a sort of problem i would think | 10:24 |
| lucasagomes | another problem I see is that the expected exceptions is not waiting for DriverNotFound | 10:24 |
| lucasagomes | which can be raised by the acquire | 10:24 |
| rameshg87 | hmm yeah | 10:24 |
| * lucasagomes will fix that too | 10:24 | |
| rameshg87 | lucasagomes: i can have drivers a,b enabled on conductor 1 and 2; c,d enabled on conductors 3,4 (because they require particular setups) | 10:25 |
| rameshg87 | lucasagomes: i make node x switch from driver a to driver c | 10:25 |
| rameshg87 | lucasagomes: it would never work | 10:25 |
| lucasagomes | rameshg87, why not? I mean ur just updating the db right? | 10:26 |
| lucasagomes | but still conductor 1 or 2 can check if the driver 'c' exist in the cloud | 10:26 |
| lucasagomes | if it does it updates the db, release the lock | 10:26 |
| lucasagomes | subsequent requests on that node will now go to conductors 3,4 | 10:27 |
| rameshg87 | lucasagomes: just confused about this | 10:28 |
| rameshg87 | lucasagomes: https://github.com/openstack/ironic/blob/master/ironic/conductor/task_manager.py#L202-L203 | 10:28 |
| lucasagomes | right, that's just an instance of the driver | 10:29 |
| rameshg87 | lucasagomes: but what if we never loaded that driver on conductor 1 ? | 10:29 |
| lucasagomes | oh wait... hmm | 10:29 |
| rameshg87 | lucasagomes: will driver_factory have it ? | 10:29 |
| lucasagomes | indeed | 10:29 |
| lucasagomes | gotcha | 10:29 |
| lucasagomes | so perhaps acquiring the lock with the new_driver_name is wrong there | 10:30 |
| rameshg87 | lucasagomes: i think it's unncessary | 10:30 |
| rameshg87 | lucasagomes: we just wanted to make sure noone else does db update while we are doing | 10:30 |
| rameshg87 | lucasagomes: i think it doesn't matter it is old driver or new driver | 10:30 |
| lucasagomes | rameshg87, yeah, I think the right way of doing that check | 10:31 |
| lucasagomes | is to use "get_topic_for_driver" | 10:31 |
| lucasagomes | from the rcpapi | 10:31 |
| rameshg87 | and check if topic is not None | 10:31 |
| rameshg87 | great :) | 10:31 |
| lucasagomes | that will check if the driver exist in the hash ring | 10:31 |
| lucasagomes | yeah | 10:31 |
| lucasagomes | and check | 10:31 |
| rameshg87 | +1 | 10:31 |
| lucasagomes | but not try to acquire it on the old conductor | 10:31 |
| lucasagomes | ack | 10:31 |
| lucasagomes | damn haha | 10:32 |
| lucasagomes | loads of bugs there | 10:32 |
| rameshg87 | :) | 10:32 |
| rameshg87 | lucasagomes: but we will need to remap the node too, right ? | 10:32 |
| rameshg87 | lucasagomes: to another conductor which has the driver | 10:33 |
| rameshg87 | i mean new driver | 10:33 |
| lucasagomes | I thought it would happen automagically, I will check | 10:34 |
| lucasagomes | cause new requests will call get_topic_for() | 10:35 |
| lucasagomes | and that resets the hash ring | 10:35 |
| lucasagomes | which will then reload the stuff | 10:35 |
| rameshg87 | got it | 10:36 |
| rameshg87 | yeah should work automatically after that | 10:36 |
| lucasagomes | probably I should separate the bugs, fix the problem trying to load a new driver which might not be enabled on that particularly conductor, and then fix the stop console thing | 10:39 |
| * lucasagomes works | 10:39 | |
| ramineni | dtantsur: hi | 10:41 |
| rameshg87 | lucasagomes: yeah i agree | 10:43 |
| *** atuvenie has quit IRC | 10:47 | |
| openstackgerrit | Haomeng,Wang proposed openstack/ironic: Fix UnicodeEncodeError issue when the language is not en_US https://review.openstack.org/168268 | 10:51 |
| *** priteau has joined #openstack-ironic | 10:53 | |
| *** yuanying has quit IRC | 10:54 | |
| *** ramineni has quit IRC | 10:57 | |
| *** yuanying has joined #openstack-ironic | 10:57 | |
| tiagogomes_ | Hi, my baremetal server is not finding the Neutron's DHCP server. I can see `DHCPDISCOVER(tap4be9b5ea-00) 9c:b6:54:bb:67:05 no address available` on the dhcp-agent log | 10:59 |
| openstackgerrit | Pierre Riteau proposed openstack/python-ironicclient: Add support for filtering nodes by chassis UUID https://review.openstack.org/167964 | 11:00 |
| tiagogomes_ | That MAC address was added to Ironic: http://paste.openstack.org/show/197020/ | 11:00 |
| tiagogomes_ | anyone knows what it could be missing | 11:00 |
| *** saripurigopi has quit IRC | 11:01 | |
| tiagogomes_ | I am not using Nova | 11:01 |
| Haomeng | tiagogomes_: if not run with nova, the neutron dhcp port will not be updated to support pxe options I think | 11:02 |
| *** dtantsur is now known as dtantsur|brb | 11:02 | |
| tiagogomes_ | I thought I could use Ironic without nova | 11:02 |
| tiagogomes_ | anyway, the problem is not pxe now, is not being able to get an IP address from the DHCP server | 11:03 |
| Haomeng | tiagogomes_: yes, ironic can run on standalone mode, but the dhcp/images should be prepared by manually | 11:03 |
| tiagogomes_ | doesn't Neutron take care of that | 11:04 |
| Haomeng | tiagogomes_: yes, that is the nova pass the ironic port to neutron, so without nova, neutron does not konw ironic mac, so neutron will not response with the unknow mac to assign ip | 11:04 |
| rameshg87 | tiagogomes_: are you using neutron + ironic without nova ? | 11:04 |
| tiagogomes_ | I am using Neutron, Ironic, Glance and Keystone | 11:04 |
| Haomeng | tiagogomes_: you can add neutron port and update dhcp port by neutron command by manually I think | 11:05 |
| tiagogomes_ | So what is `ironic port-add` about | 11:05 |
| tiagogomes_ | `port-create` sorry | 11:06 |
| Haomeng | tiagogomes_: ironic port-add is try to add baremetal nic mac to ironic, it will be passed to neutron by nova if run with nova | 11:07 |
| devananda | morning, all | 11:07 |
| Haomeng | devananda: good morning deva:) | 11:08 |
| devananda | lucasagomes: seen https://bugs.launchpad.net/ironic/+bug/1405131 ? | 11:08 |
| openstack | Launchpad bug 1405131 in OpenStack Compute (nova) "Ports cannot be mapped to networks" [Low,In progress] - Assigned to Mark Goddard (mgoddard) | 11:08 |
| Haomeng | AJaeger: sorry for late response, I am back now:) | 11:08 |
| *** dprince has joined #openstack-ironic | 11:08 | |
| lucasagomes | devananda, morning | 11:08 |
| lucasagomes | devananda, lemme take a look | 11:08 |
| devananda | mgoddard: ohhai! I havent seen updates on ^ in a while - are you still working on it? seems like something we should actually fix now | 11:09 |
| Haomeng | tiagogomes_: we have the document for standalone mode - http://docs.openstack.org/developer/ironic/deploy/install-guide.html | 11:09 |
| lucasagomes | devananda, yup, remember we talked about in the mid-cycle about having a attribute on ports | 11:09 |
| Haomeng | tiagogomes_: hope this can help you | 11:09 |
| lucasagomes | that tells whether it can be used for PXE or not? | 11:09 |
| tiagogomes_ | Haomeng I was following that | 11:09 |
| lucasagomes | we should add it, and modify the nova driver to filter the ports by that attribute | 11:09 |
| Haomeng | tiagogomes_: yes, but it is the dhcp_provider=none case | 11:10 |
| tiagogomes_ | I created a port in neutron with port-create, and I added it to Ironic with `ironic port-update $PORT add extra/vif_port_id=$NEUTRON_PORT_UUID` | 11:11 |
| Haomeng | tiagogomes_: ok, can you check your neutron port - neutron port-show | 11:11 |
| vdrok | tiagogomes_, did you add fixed_ip and mac address to neutron port? | 11:12 |
| Haomeng | tiagogomes_: to find "device_owner | network:dhcp" in the result | 11:12 |
| *** yog__ has quit IRC | 11:12 | |
| Haomeng | tiagogomes_: make sure we have neutron dhcp port | 11:13 |
| vdrok | Haomeng, tiagogomes_ IIRC when I tried it I just created neutron port with fixed_ip and mac address in private network | 11:13 |
| Haomeng | tiagogomes_: and check dnsmasq process | 11:13 |
| vdrok | and it worked | 11:13 |
| Haomeng | vdrok: I think we need to update neutron dhcp port to add pxe options, such as tftp address, and pxe file | 11:14 |
| vdrok | Haomeng, if we have dhcp_provider=neutron it will do that | 11:14 |
| devananda | lucasagomes: ack. /me targets to rc1 | 11:14 |
| tiagogomes_ | Ah, I didn't set up any macaddresss or device_owner | 11:14 |
| tiagogomes_ | so the MAC address needs to correspond to the one on the baremetal NIC | 11:15 |
| Haomeng | tiagogomes_, vdrok: , so not only dhcp works but also the pxe options are ready for pxe boot | 11:15 |
| devananda | lucasagomes: mgoddard: that means we'll actually need to work on it, and it looks like the fix is on the nova side ... what's required on Iornic's side? | 11:15 |
| lucasagomes | devananda, the problem is the nova :( will they accept the change? | 11:15 |
| devananda | lucasagomes: possibly. the one mgoddard proposed was just a WIP | 11:16 |
| lucasagomes | I see | 11:16 |
| Haomeng | tiagogomes_: the pxe options are - option:bootfile-name,pxelinux.0 / option:server-ip-address,9.119.58.235 / tftp-server,9.119.58.235 | 11:16 |
| devananda | if it were cleaned up, there's a good chance they'll take it | 11:16 |
| vdrok | Haomeng, neutron will be able to update pxe options itself | 11:16 |
| tiagogomes_ | I've dhcp_provider=neutron, so I only need the remaining steps | 11:16 |
| Haomeng | Haomeng: they are requried for pxe boot | 11:16 |
| lucasagomes | I think in Ironic we should have a "bootable" attribute on the port | 11:16 |
| lucasagomes | that can say whether the port can be used for network boot or not | 11:16 |
| lucasagomes | and some way to filter ports that are bootable on our api | 11:17 |
| lucasagomes | the later is not 100% necessary for a 1st version | 11:17 |
| Haomeng | vdrok: ok, I remember the neutron dhcp port update which is call from ironic nova driver , let me check code | 11:17 |
| Haomeng | tiagogomes_: so we can just check the dnsmasq process | 11:17 |
| tiagogomes_ | I am using Juno/Stable | 11:17 |
| Haomeng | tiagogomes_: to see --dhcp-optsfile | 11:17 |
| Haomeng | tiagogomes_: juno? | 11:17 |
| devananda | lucasagomes: that sounds like API changes // feature work | 11:17 |
| tiagogomes_ | Haomeng,yes juno | 11:17 |
| lucasagomes | yeah, we can ignore it for now | 11:18 |
| lucasagomes | and do it in L | 11:18 |
| Haomeng | tiagogomes_: it is better that you run with latest kilo code | 11:18 |
| vdrok | tiagogomes_, not sure about juno, you can just check dnsmasq configs | 11:18 |
| lucasagomes | since the attribute is indexable on the db, it makes sense to make querying it on the API #futurework | 11:18 |
| vdrok | tiagogomes_, if there is entry with your mac and pxe options than it will work fine | 11:18 |
| tiagogomes_ | vdrok how can I do that | 11:18 |
| Haomeng | tiagogomes_, vdrok: yes, check --dhcp-optsfile option file content | 11:18 |
| Haomeng | tiagogomes_: share my case - http://paste.openstack.org/show/197023/ | 11:19 |
| * rameshg87 goes home | 11:19 | |
| rameshg87 | see you all in a couple of hours | 11:19 |
| Haomeng | tiagogomes_: and check --dhcp-hostsfile file | 11:19 |
| *** rameshg87 has quit IRC | 11:19 | |
| tiagogomes_ | Haomeng mine http://paste.openstack.org/show/197027/ | 11:20 |
| Haomeng | tiagogomes_: it is fine | 11:20 |
| vdrok | tiagogomes_, tag is neutron port uuid | 11:20 |
| Haomeng | tiagogomes_: how about --dhcp-hostsfile file content | 11:20 |
| vdrok | tiagogomes_, so if that matches port you created then it should be fine | 11:21 |
| tiagogomes_ | fa:16:3e:a9:1e:cb,host-192-168-100-15.openstacklocal,192.168.100.15,set:b025320d-3825-47e2-a0e3-cc1a8c216840 | 11:21 |
| tiagogomes_ | fa:16:3e:a6:0a:87,host-192-168-100-17.openstacklocal,192.168.100.17 | 11:21 |
| Haomeng | tiagogomes_: another question, what is your l2 net type, flat or vlan or gre? | 11:21 |
| Haomeng | tiagogomes_: ironic supports flat now | 11:21 |
| tiagogomes_ | flat I thik | 11:21 |
| Haomeng | tiagogomes_: ok, neutron dnsmasq looks fine, so can you try to tcpdump the dhcp package from your ironic side? | 11:22 |
| Haomeng | tiagogomes_: make sure the baremetal dhcp requests are send to the neutron dhcp machine, and check the interface, if there is namespace or not | 11:22 |
| tiagogomes_ | Haomeng not sure what you mean by my ironic side, my I am receiving DHCP discover packets | 11:22 |
| Haomeng | tiagogomes_: flat is fine and easy to debug | 11:22 |
| Haomeng | tiagogomes_: ok, which interface you are listening to capture the dhcp siccover packets | 11:23 |
| Haomeng | tiagogomes_: check dnsmasq process --bind-interfaces arguments to see which interface is bind to get dhcp request | 11:24 |
| Haomeng | tiagogomes_: --interface argument | 11:24 |
| tiagogomes_ | the interface is virtual interface in a VM dedicated only to contact with the baremetal | 11:24 |
| Haomeng | tiagogomes_: ok, did you run with ovs driver for neutron | 11:24 |
| tiagogomes_ | `--bind-interfaces --interface=tap4be9b5ea-00 --except-interface=lo` | 11:25 |
| Haomeng | tiagogomes_: can you share "# ovs-vsctl show" output | 11:25 |
| Haomeng | tiagogomes_: to check if the tap4be9b5ea-00 interface is bind to physical port | 11:25 |
| tiagogomes_ | http://paste.openstack.org/show/197031/ | 11:25 |
| Haomeng | tiagogomes_: ok, tap4be9b5ea-00 is bind to br-int and br-int is bind to eth2 physical nic, right? | 11:26 |
| Haomeng | tiagogomes_: so you said you can receive dhcp discovery packets, which interface you bind to capture the tcp/udp packages | 11:27 |
| tiagogomes_ | Haomeng ens6 | 11:28 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: boot_mode is overwritten in node properties https://review.openstack.org/167952 | 11:28 |
| Haomeng | tiagogomes_: I think we can get baremetal dhcp request from physical eth2 nic, but can not get it for tap4be9b5ea-00 interface | 11:28 |
| Haomeng | tiagogomes_: ok, try to tcpdump to check tap4be9b5ea-00 packages, and redhcp from baremetal to see if we can get dhcp packages, I think no | 11:29 |
| Haomeng | tiagogomes_: some times, neutron have issue with ovs netflow table, so have to check netflow table to see how the packages is routed by netflow | 11:29 |
| Haomeng | tiagogomes_: try with "tcpdump -i tap4be9b5ea-00 '(port 67 or port 68)'" | 11:31 |
| Haomeng | tiagogomes_: if tap4be9b5ea-00 is in down status, just up it - "ip link set tap4be9b5ea-00 up" | 11:32 |
| *** tiagogomes_ has quit IRC | 11:33 | |
| Haomeng | tiagogomes_: and check " ovs-ofctl dump-flows br-int" flow table to see if priority is correct value | 11:33 |
| *** tiagogomes_ has joined #openstack-ironic | 11:33 | |
| tiagogomes_ | <tiagogomes_> Haomeng, I'm confused | 11:34 |
| tiagogomes_ | <tiagogomes_> would it help if I described all the network steps that I've done | 11:34 |
| tiagogomes_ | <tiagogomes_> I don't know much about neutron, so maybe I am doing something silly | 11:34 |
| tiagogomes_ | <tiagogomes_> I also didn't set the mac address when creating the port with neutron, could that be the proble | 11:34 |
| Haomeng | tiagogomes_: let me check my env | 11:34 |
| Haomeng | tiagogomes_: neutron port-list should return 2 rows at list, one is dhcp port, another one is for baremetal port | 11:35 |
| *** vdrok has quit IRC | 11:36 | |
| Haomeng | tiagogomes_: if you just have dhcp port, then the baremetal port should be created by neutron, then neutron can support it | 11:36 |
| *** vdrok has joined #openstack-ironic | 11:37 | |
| Haomeng | tiagogomes_: and check dnsmasq --dhcp-hostsfile file content, to see if there is your baremetal mac as well | 11:37 |
| tiagogomes_ | The baremetal port MAC address should correspond to the baremetal server MAC address? Because it doesn't | 11:38 |
| tiagogomes_ | the baremetal mac is not on --dhcp-hostsfile | 11:39 |
| Haomeng | tiagogomes_: yes that is reasone why neutron dhcp does not response your baremetal dhcp request | 11:39 |
| Haomeng | tiagogomes_: neutorn dhcp work on white-list mode, just response the know mac which registed into neutron as neutorn port | 11:39 |
| Haomeng | tiagogomes_: so run "neutron port-list" to check if your baremetal mac is in the list | 11:40 |
| Haomeng | tiagogomes_: if no, have to create by manually with neutron port-create command to register the baremetal mac | 11:40 |
| *** dprince has quit IRC | 11:42 | |
| tiagogomes_ | Haomeng, done, should I associate the port with Ironic now | 11:44 |
| *** dprince has joined #openstack-ironic | 11:44 | |
| openstackgerrit | Yuiko Takada proposed stackforge/ironic-discoverd: Add unit tests for firewall.py https://review.openstack.org/168304 | 11:45 |
| Haomeng | tiagogomes_: I understand you have run ironic port-create to register the baremetal mac already | 11:45 |
| Haomeng | tiagogomes_: so check dnsmasq --dhcp-hostsfile again | 11:46 |
| Haomeng | tiagogomes_: make sure the mac is ready in the list | 11:46 |
| tiagogomes_ | Haomeng, yes, but normamilly I run `ironic port-update $PORT add extra/vif_port_id=$NEUTRON_PORT_UUID` | 11:46 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: When updating node's driver do not try to load the new driver https://review.openstack.org/168307 | 11:46 |
| tiagogomes_ | Haomeng, if I didn't add that extra/vif_port I would get a no VIFs found error | 11:47 |
| tiagogomes_ | Haomeng, the MAC address is on --dhcp-hostsfile | 11:47 |
| Haomeng | tiagogomes_: if the mac is on dhcp-hostfile, it will work with neutron | 11:47 |
| Haomeng | tiagogomes_: so re-dhcp again and capture packages with command - "tcpdump -i tap4be9b5ea-00 '(port 67 or port 68)'" | 11:48 |
| Haomeng | tiagogomes_: yes, we should add extra/vif-port first then run ironic to kickoff the provision | 11:49 |
| *** jjohnson2 has joined #openstack-ironic | 11:49 | |
| tiagogomes_ | I can't tcpdump on that interface: | 11:50 |
| tiagogomes_ | ~ # tcpdump -i tap4be9b5ea-00 | 11:50 |
| tiagogomes_ | tcpdump: tap4be9b5ea-00: SIOCETHTOOL(ETHTOOL_GET_TS_INFO) ioctl failed: No such device | 11:50 |
| tiagogomes_ | hold on, DHCPOFFER on dhcp-agent log | 11:50 |
| Haomeng | tiagogomes_: try "ip link set tap4be9b5ea-00 up" | 11:51 |
| tiagogomes_ | ~ # ip link set tap4be9b5ea-00 up | 11:51 |
| tiagogomes_ | Cannot find device "tap4be9b5ea-00" | 11:51 |
| Haomeng | tiagogomes_: run ip netns | 11:52 |
| tiagogomes_ | ~ # ip netns | 11:52 |
| tiagogomes_ | qdhcp-e2e418d7-f430-45f0-8c34-a742105c4f9a | 11:52 |
| Haomeng | tiagogomes_: did you enable netspace, it is in dhcp netspace | 11:52 |
| Haomeng | tiagogomes_: ip netns exec qdhcp-563b5dee-01a0-4bb3-b575-fdfdc4cb513e tcpdump '(port 67 or port 68)' -i tap80a5fe3b-77 | 11:52 |
| tiagogomes_ | ~ # ip netns exec qdhcp-563b5dee-01a0-4bb3-b575-fdfdc4cb513e true | 11:53 |
| tiagogomes_ | Cannot open network namespace "qdhcp-563b5dee-01a0-4bb3-b575-fdfdc4cb513e": No such file or directory | 11:53 |
| tiagogomes_ | ah wrong namespace | 11:53 |
| Haomeng | tiagogomes_: sorry for confusing, that is my netns:) | 11:53 |
| Haomeng | tiagogomes_: ip netns exec qdhcp-e2e418d7-f430-45f0-8c34-a742105c4f9a tcpdump '(port 67 or port 68)' -i tap4be9b5ea-00 | 11:53 |
| Haomeng | tiagogomes_: I forgot to replace to yours:) | 11:53 |
| tiagogomes_ | no, mea culpa | 11:53 |
| Haomeng | tiagogomes_: that is my case:) | 11:53 |
| Haomeng | tiagogomes_: just run any commad with this "ip netns exec qdhcp-e2e418d7-f430-45f0-8c34-a742105c4f9a" as prefix, that will check in that net namespace | 11:56 |
| tiagogomes_ | Haomeng, yes I read about that some days ago | 11:56 |
| Haomeng | tiagogomes_: cool:) | 11:57 |
| Haomeng | tiagogomes_: you can just reboot the baremeatl and see the dhcp output, no need to run ironic command now, becaue we are checking neutron dhcp services now | 11:57 |
| Haomeng | tiagogomes_: to see if neutron dhcp is ready to serv the pxe dhcp options | 11:57 |
| Haomeng | tiagogomes_: another easy way, try to start a vm which mac is set to same with baremetal and the vm is conected to same flat bridge | 11:58 |
| tiagogomes_ | sounds like a plan, doing it mow | 11:58 |
| tiagogomes_ | now | 11:58 |
| Haomeng | tiagogomes_: ok, good luck:) | 11:58 |
| *** subscope has joined #openstack-ironic | 12:00 | |
| tiagogomes_ | I had to stop ironic-conductor, as he was shutting down the node | 12:03 |
| Haomeng | tiagogomes_: I think we can just go head to check neutron dhcp first, make sure neutorn dhcp works, than go ahead to run ironic command to kickoff the provision | 12:05 |
| Haomeng | tiagogomes_: did you have remote console for baremetal? | 12:05 |
| *** atuvenie has joined #openstack-ironic | 12:06 | |
| tiagogomes_ | Haomeng kind of, sometimes it doesn't work | 12:06 |
| Haomeng | tiagogomes_: ok | 12:06 |
| tiagogomes_ | Haomeng, yes that's what I am doing, that's why I stopped Ironic | 12:06 |
| Haomeng | tiagogomes_: ok, good luck for you:) | 12:07 |
| tiagogomes_ | Haomeng, ok, I received some packets on tap4be9b5ea-00 | 12:07 |
| Haomeng | tiagogomes_: cool | 12:07 |
| Haomeng | tiagogomes_: check if it is dhcp packages | 12:08 |
| tiagogomes_ | Haomeng, yep, the full set, discover,request,offer,ack | 12:08 |
| Haomeng | tiagogomes_: great, should works now:) | 12:08 |
| Haomeng | tiagogomes_: so check the baremetal console output | 12:09 |
| Haomeng | tiagogomes_: if it is request the NBP pxe file | 12:09 |
| *** subscope has quit IRC | 12:11 | |
| tiagogomes_ | Haomeng, PXE-E53: No boot filename received, but this is expected no? | 12:12 |
| Haomeng | tiagogomes_: have to check dnsmasq process | 12:12 |
| tiagogomes_ | although I can see some PXE settings on the baremetal port | 12:12 |
| *** dtantsur|brb is now known as dtantsur | 12:13 | |
| Haomeng | tiagogomes_: e | 12:13 |
| Haomeng | tiagogomes_: --dhcp-optsfile | 12:13 |
| Haomeng | tiagogomes_: should have "option:bootfile-name,pxelinux.0" | 12:13 |
| tiagogomes_ | Haomeng, it does | 12:14 |
| Haomeng | tiagogomes_: check dnsmasq log, maybe that can be find in /var/log/message | 12:14 |
| Haomeng | tiagogomes_: what your systems is, x86 server? | 12:15 |
| tiagogomes_ | Haomeng, yes x86 | 12:16 |
| Haomeng | tiagogomes_: some pxe rom is not working as well | 12:16 |
| Haomeng | tiagogomes_: but it is few case for x86 | 12:16 |
| tiagogomes_ | I don't have anything on /var/log/message, it is systemd, but I don't have an unit for dnsmasq | 12:17 |
| Haomeng | tiagogomes_: got a reference - http://www.symantec.com/business/support/index?page=content&id=TECH10532 | 12:17 |
| Haomeng | tiagogomes_: another way is, we just run dnsmasq proces by manually, then we can have the dhcp pxe detail logs | 12:18 |
| Haomeng | tiagogomes_: some time it is caused by tftp folder mapping | 12:19 |
| tiagogomes_ | right, but know I know that Neutron and baremetal server are communicating properly, which was my issue for some days | 12:19 |
| Haomeng | tiagogomes_: can you check your /tftpboot folder if pxelinux.0 is there | 12:19 |
| Haomeng | tiagogomes_: yes | 12:19 |
| Haomeng | tiagogomes_: I have dnsmasq config for manually run, you can have a reference to debug pxe booting | 12:20 |
| tiagogomes_ | Haomeng that dir it is not on a conventional place, I set the right dir in the ironic configuration file, but I am not using Ironic | 12:20 |
| tiagogomes_ | s/conventional/default | 12:20 |
| Haomeng | tiagogomes_: http://paste.openstack.org/show/197072/ | 12:21 |
| Haomeng | tiagogomes_: we can run tftp client command to get pxelinux.0 by manully to verify if tftp server is working or ont | 12:22 |
| Haomeng | tiagogomes_: netstat -an|grep ":69" | 12:23 |
| tiagogomes_ | Haomeng, good idea. I don't have that configuration file | 12:23 |
| tiagogomes_ | ~ # netstat -an | grep ":69" | 12:23 |
| tiagogomes_ | udp 0 0 0.0.0.0:69 0.0.0.0:* | 12:23 |
| Haomeng | tiagogomes_: try to tftp download pxelinux.0 by manually | 12:24 |
| Haomeng | tiagogomes_: or run dnsmasq with my configuration to check dnsmasq full logs, we can see all the pxe booting actions | 12:25 |
| *** subscope has joined #openstack-ironic | 12:25 | |
| Haomeng | tiagogomes_: my log - http://paste.openstack.org/show/197077/ | 12:25 |
| Haomeng | tiagogomes_: that is the way to debug pxe booting | 12:26 |
| tiagogomes_ | Haomeng I'll do, thanks | 12:28 |
| Haomeng | tiagogomes_: ok, good luck:) | 12:28 |
| tiagogomes_ | Haomeng, thanks for all the help ! | 12:29 |
| Haomeng | tiagogomes_: that is my pleasure, welcome:) | 12:30 |
| Haomeng | tiagogomes_: good luck and enjoy weekend:) | 12:30 |
| Haomeng | tiagogomes_: try to tftp command by manually to check if we can get pxelnux.0 by tftp or not | 12:31 |
| Haomeng | tiagogomes_: if not working, that is tftp issue | 12:31 |
| *** atuvenie has quit IRC | 12:32 | |
| TheJulia | good morning ironic | 12:32 |
| tiagogomes_ | Haomeng, it worked | 12:34 |
| tiagogomes_ | that is, getting pxelinux.0 through tftp | 12:35 |
| Haomeng | tiagogomes_: ok | 12:35 |
| Haomeng | Haomeng: run dnsmasq process with my conf file and make sure the configuration is changed for your env/case, and try to check dnsmasq log to check more in details why pxe not working | 12:36 |
| Haomeng | tiagogomes_: because for baremetal pxe side, there is not enough information to debug/investigation the root cause | 12:36 |
| Haomeng | tiagogomes_: suggest to boot a vm to test the pxe/dhcp/tftp env first | 12:37 |
| Haomeng | tiagogomes_: dnsmasq --conf-file=dnsmasq.conf | 12:40 |
| *** yuanying has quit IRC | 12:42 | |
| *** rameshg87 has joined #openstack-ironic | 12:43 | |
| *** kkoski has joined #openstack-ironic | 12:44 | |
| *** jcoufal has joined #openstack-ironic | 12:46 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Stop console when updating node's driver https://review.openstack.org/168030 | 12:47 |
| yuriyz | hello Ironic | 12:52 |
| *** atuvenie has joined #openstack-ironic | 12:52 | |
| yuriyz | lucasagomes please look at https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L1068-L1079 | 12:52 |
| lucasagomes | yuriyz, hi there | 12:53 |
| * lucasagomes clicks | 12:53 | |
| *** athomas has joined #openstack-ironic | 12:53 | |
| lucasagomes | yuriyz, oh, but that's testing the topic for the old driver | 12:53 |
| lucasagomes | yuriyz, I'm worried about the new one, when the user is updating the driver from A to B | 12:53 |
| lucasagomes | we should check if there's a conductor that can manage B | 12:54 |
| rameshg87 | yuriyz: i think lucasagomes will need the conductor to load the old driver instead | 12:54 |
| rameshg87 | yuriyz: to stop the console | 12:54 |
| lucasagomes | yeah, I have to get a lock on the old conductor, so it loads the current driver (prior to update) | 12:54 |
| lucasagomes | stop the console | 12:55 |
| rameshg87 | so should we change https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L1068-L1079 to get a conductor which old driver ? | 12:55 |
| rameshg87 | *which has old driver | 12:55 |
| lucasagomes | rameshg87, yeah | 12:55 |
| lucasagomes | x.x | 12:55 |
| * lucasagomes mark it all as WIP | 12:56 | |
| lucasagomes | rameshg87, I wonder if perhaps we should just check if the conductor has the console enabled and fail update | 12:58 |
| lucasagomes | at the API layer | 12:58 |
| lucasagomes | rameshg87, like, can't change node's <blah> driver because the console is enabled, stop it first | 12:58 |
| lucasagomes | something like that | 12:59 |
| rameshg87 | oh, asking the user to stop the console ? | 12:59 |
| lucasagomes | well not sure, I think not right? maybe we should be smarter | 12:59 |
| lucasagomes | and just try to stop it | 12:59 |
| rameshg87 | hmm..may be..even i am in 2 minds.. | 13:00 |
| lucasagomes | yeah | 13:00 |
| rameshg87 | it should easy for them to stop the console themselves | 13:00 |
| rameshg87 | and how ofter does someone change the driver ? :) | 13:00 |
| lucasagomes | not often I suppose | 13:00 |
| rameshg87 | *often | 13:00 |
| rameshg87 | yeah, may be even for it seems fine if we just ask them to stop the console themselves | 13:01 |
| lucasagomes | fair enuff | 13:01 |
| lucasagomes | I will grab some food and update the patches | 13:01 |
| rameshg87 | yuriyz: ^^^ | 13:01 |
| rameshg87 | okay | 13:01 |
| lucasagomes | if you and yuriyz agrees with that | 13:02 |
| *** lucasagomes is now known as lucas-hungry | 13:02 | |
| lucas-hungry | bbl | 13:02 |
| rameshg87 | anyone has bug# for usual postgres failures in gate ? | 13:02 |
| yuriyz | lucasagomes ok | 13:04 |
| *** dlpartain has joined #openstack-ironic | 13:12 | |
| *** dlpartain has left #openstack-ironic | 13:13 | |
| devananda | going to add a couple bugs that I think were fixed already to the list | 13:14 |
| devananda | rameshg87: didn't you fix this already? https://bugs.launchpad.net/ironic/+bug/1412561 | 13:17 |
| openstack | Launchpad bug 1412561 in Ironic "IPA/DIB is not compatible to create the uefi capable images" [Medium,Confirmed] | 13:17 |
| devananda | also, I'm bumping https://bugs.launchpad.net/ironic/+bug/1422723 to critical | 13:18 |
| openstack | Launchpad bug 1422723 in Ironic "Ironic sets instance as active even if bootloader installation fails when boot_option is 'local'" [Critical,Confirmed] | 13:18 |
| devananda | because that's a serious failure in Ironic | 13:18 |
| rameshg87 | devananda: checking ... | 13:19 |
| devananda | also https://bugs.launchpad.net/ironic/+bug/1432761 is a fairly serious issue IMO. the REST API can't be synchronous with respect to hardware like this, and raising an error is silly. | 13:20 |
| openstack | Launchpad bug 1432761 in Ironic "Cannot perform get_cleaning_steps on powered off node." [Critical,Confirmed] | 13:20 |
| dtantsur | fair | 13:21 |
| rameshg87 | devananda: https://bugs.launchpad.net/ironic/+bug/1412561 is only partly true | 13:21 |
| openstack | Launchpad bug 1412561 in Ironic "IPA/DIB is not compatible to create the uefi capable images" [Medium,Confirmed] | 13:21 |
| rameshg87 | devananda: since we added a new element ironic-agent in diskimage-builder | 13:22 |
| *** ifarkas has quit IRC | 13:22 | |
| rameshg87 | devananda: we can create uefi-capable agent iso images using ironic-agent in dib | 13:22 |
| rameshg87 | devananda: for building ipa with coreos - we had added a script in juno - https://github.com/openstack/ironic-python-agent/blob/master/imagebuild/coreos/iso-image-create | 13:23 |
| rameshg87 | devananda: that needs to be updated to generate uefi capable images | 13:23 |
| *** ifarkas has joined #openstack-ironic | 13:24 | |
| rameshg87 | devananda: i will take care of that .. | 13:24 |
| rameshg87 | devananda: fixing https://bugs.launchpad.net/ironic/+bug/1422723 is a bit hard for kilo-rc1 | 13:24 |
| openstack | Launchpad bug 1422723 in Ironic "Ironic sets instance as active even if bootloader installation fails when boot_option is 'local'" [Critical,Confirmed] | 13:24 |
| rameshg87 | devananda: the bug is for localboot with dib ramdisk | 13:25 |
| rameshg87 | devananda: currently there is no way dib ramdisk can communicate back to ironic the status of the bootloader installation | 13:25 |
| rameshg87 | devananda: in my thinking, adding a new vendor passthru to communicate status of boot loader isntallation is the only way | 13:26 |
| *** atuvenie has quit IRC | 13:27 | |
| *** kkoski has quit IRC | 13:39 | |
| *** kkoski has joined #openstack-ironic | 13:43 | |
| *** stendulker has joined #openstack-ironic | 13:54 | |
| *** r-daneel has joined #openstack-ironic | 13:58 | |
| NobodyCam | good morning Ironic AND T.G.I.F :) | 14:00 |
| TheJulia | good morning and indeed! | 14:02 |
| NobodyCam | :) | 14:02 |
| NobodyCam | morning TheJulia :) | 14:02 |
| NobodyCam | oh coffee sounds like its ready..brb | 14:02 |
| dtantsur | NobodyCam, TheJulia, morning! | 14:02 |
| * TheJulia needs coffee horribly | 14:02 | |
| *** dencaval has joined #openstack-ironic | 14:03 | |
| *** Marga_ has joined #openstack-ironic | 14:06 | |
| *** lucas-hungry is now known as lucasagomes | 14:07 | |
| *** Marga_ has quit IRC | 14:08 | |
| *** Marga_ has joined #openstack-ironic | 14:09 | |
| jroll | 00:23:30 devananda | JayF: however, I should be ashamed that I'm using and talking about the agent, but dont understand it well enough to go review it meaningfully <- I think you overestimate the complexity of the agent, you should be able to grok it in a couple of hours | 14:10 |
| jroll | morning NobodyCam TheJulia dtantsur devananda lucasagomes rameshg87 and everyone above the fold :) | 14:10 |
| * jroll highlights half the channel | 14:10 | |
| dtantsur | jroll, o/ | 14:10 |
| lucasagomes | jroll, NobodyCam TheJulia morning | 14:10 |
| jroll | rameshg87: lucasagomes: funny story about cloud-init using a configdrive on a partition, we patched that over a year ago, before we even started working on ironic :P | 14:11 |
| rameshg87 | jroll: morning | 14:11 |
| rameshg87 | jroll: oh | 14:11 |
| lucasagomes | jroll, oh damn | 14:11 |
| lucasagomes | jroll, you gotta submit the patch upstream man! heh | 14:12 |
| lucasagomes | but yeah, I'm glad that f21 has the version that works with partitions | 14:12 |
| lucasagomes | and rhel/centos7 too | 14:12 |
| jroll | lucasagomes: it is upstream, as evidenced by you finding it in newer versions :P | 14:13 |
| lucasagomes | jroll, :) fair enuff | 14:14 |
| rameshg87 | lucasagomes: we will start getting f21 with diskimage-create soon ? | 14:14 |
| NobodyCam | morning dtantsur, jroll and lucasagomes | 14:14 |
| NobodyCam | and rameshg87 | 14:14 |
| NobodyCam | :) | 14:14 |
| jroll | lucasagomes: something something old packages in fedora :P | 14:14 |
| rameshg87 | morning NobodyCam | 14:14 |
| lucasagomes | rameshg87, I hope so, as dtantsur pointed out there's one guy (ian) already working on it | 14:14 |
| rameshg87 | okay | 14:14 |
| NobodyCam | :) | 14:14 |
| lucasagomes | jroll, indeed | 14:14 |
| lucasagomes | rameshg87, but I don't know exactly when | 14:14 |
| rameshg87 | lucasagomes: i tried to update the cloudinit using easy_install after going into fedora image | 14:15 |
| rameshg87 | lucasagomes: and it got screwed :) | 14:15 |
| jroll | heh | 14:15 |
| lucasagomes | x.x | 14:15 |
| rameshg87 | lucasagomes: after that when i tried to boot the image, cloudinit was trying to create ubuntu user on fedora and was failing :D | 14:15 |
| rameshg87 | supposed to fail btw :) | 14:16 |
| *** mtanino has joined #openstack-ironic | 14:16 | |
| lucasagomes | yeah looks very likely to fail | 14:16 |
| lucasagomes | rameshg87, btw, re installing grub2 on images | 14:16 |
| lucasagomes | did you get it working on ubuntu | 14:16 |
| lucasagomes | I put a patch upstream on DIB yesterday doing that, seems to fail on ubuntu with DIB | 14:16 |
| lucasagomes | I will take a look later see if I can find the cause | 14:16 |
| * rameshg87 checks the patch | 14:17 | |
| lucasagomes | (it's like one 1 line patch, lemme get the link) | 14:17 |
| lucasagomes | https://review.openstack.org/#/c/167929/ | 14:17 |
| lucasagomes | rameshg87, ^ | 14:17 |
| rameshg87 | lucasagomes: for my testing i install grub on the partition image after mounting it with qemu-nbd | 14:17 |
| NobodyCam | lucasagomes: when you have a free minute can take a look at mrda's comments on https://review.openstack.org/#/c/165666 | 14:17 |
| lucasagomes | rameshg87, I see, yeha that was what I was doing too | 14:18 |
| lucasagomes | but it would be nice to be able to generate the image with it installed already | 14:18 |
| lucasagomes | NobodyCam, sure | 14:18 |
| rameshg87 | lucasagomes: oh for that i mounted /dev, /proc and /sys after chrooting | 14:18 |
| jroll | btw, I'm off today, so y'all have a happy weekend :) | 14:18 |
| lucasagomes | jroll, enjoy the weekend! | 14:18 |
| jroll | NobodyCam: ^ super tgif today | 14:18 |
| lucasagomes | and today too :) | 14:18 |
| NobodyCam | oh jroll rub it in .. so early :-p | 14:18 |
| jroll | hehehe | 14:18 |
| rameshg87 | lucasagomes: i always do that when i chroot :) | 14:18 |
| NobodyCam | lol | 14:18 |
| rameshg87 | jroll: happy weekend .. | 14:18 |
| NobodyCam | have a great day of PTO jroll :) | 14:19 |
| jroll | thanks :) | 14:19 |
| lucasagomes | NobodyCam, if we merge that code we are intentionally adding a bug in ironic | 14:19 |
| rameshg87 | lucasagomes: stendulker had a patch which installs grub2 on the image for copying efi files (and then uninstall it) | 14:19 |
| lucasagomes | NobodyCam, I'm not sure if it's a good idea :/ | 14:20 |
| lucasagomes | rameshg87, right, stendulker r u going to propose it upstream? | 14:21 |
| stendulker | lucasgomes: the patch is already merged into DIB for iso element | 14:21 |
| lucasagomes | NobodyCam, maybe it's better to keep the 63 characters that doesn't comply to FQDN, but at least check if it comply to a valid hostname? | 14:21 |
| rameshg87 | lucasagomes: i think we should install grub2, grub2-efi as well | 14:22 |
| rameshg87 | lucasagomes: let me try your patch | 14:22 |
| stendulker | lucasgomes: Its https://review.openstack.org/#/c/153987/ | 14:22 |
| lucasagomes | rameshg87, yeah grub-efi sounds good | 14:22 |
| NobodyCam | lucasagomes: :/ also see irc chat http://eavesdrop.openstack.org/irclogs/%23openstack-ironic/%23openstack-ironic.2015-03-26.log | 14:22 |
| NobodyCam | starting around 22:01:17 | 14:22 |
| stendulker | lucasgomes: Check this review if that is something you are looking for.. | 14:23 |
| lucasagomes | NobodyCam, oh that's just the test that doesn't work!? | 14:23 |
| lucasagomes | hmm | 14:24 |
| lucasagomes | that's odd | 14:24 |
| stendulker | lucasgomes, rameshg87: But be careful with grub-efi. Fedora and Ubuntu have different implementation, packaging and installation for grub2 | 14:24 |
| stendulker | i think grub-efi is valid for only for ubuntu | 14:25 |
| rameshg87 | stendulker: okay | 14:25 |
| *** BadCub_Hiding is now known as BadCub | 14:25 | |
| pshige | morning | 14:25 |
| rameshg87 | stendulker: do you remember equivalent in ubuntu ? | 14:25 |
| rameshg87 | i mean fedora | 14:25 |
| BadCub | MOrning Folks | 14:25 |
| rameshg87 | BadCub: o/ | 14:25 |
| rameshg87 | morning | 14:25 |
| BadCub | Happy Friday!!! | 14:25 |
| NobodyCam | morning BadCub, and pshige | 14:26 |
| stendulker | rameshg87: its grub2-efi for fedora | 14:26 |
| pshige | BadCub: morning | 14:26 |
| NobodyCam | lucasagomes: yep from what I take from mrda it's only tests | 14:27 |
| rameshg87 | stendulker: okay | 14:27 |
| NobodyCam | morning stendulker :) | 14:27 |
| stendulker | rameshg87: grub-efi-amd64 for ubuntu | 14:27 |
| stendulker | NobodyCam: Morning | 14:27 |
| NobodyCam | :) | 14:27 |
| pshige | rameshg87: morning | 14:27 |
| lucasagomes | NobodyCam, I seem I wonder if it has to deal with the db backend :/ | 14:27 |
| lucasagomes | we use sqlite for tests right? and run mysql on devstack | 14:28 |
| stendulker | rameshg87: grub-efi is 32-bit and we should not use in uefi | 14:28 |
| stendulker | NobodyCam :) | 14:28 |
| lucasagomes | NobodyCam, hmm I don't wanna be the one taking the decision to merge it without tests really | 14:28 |
| lucasagomes | sounds risky for rc-1 | 14:28 |
| pshige | NodyCam: morning | 14:28 |
| *** athomas has quit IRC | 14:29 | |
| rameshg87 | pshige: morning :) | 14:29 |
| *** athomas has joined #openstack-ironic | 14:29 | |
| rameshg87 | stendulker: yeah, may be we should decide the package name based on architecture and distro ? | 14:29 |
| rameshg87 | stendulker: i think we should have a map file | 14:30 |
| rameshg87 | wdyt ? | 14:30 |
| lucasagomes | devananda, r u around? maybe you can take a look at mrda comments on https://review.openstack.org/#/c/165666 ? | 14:30 |
| NobodyCam | lucasagomes: ya, I told mrda I would have you look at it, I think I tend to agree... it's to late, unless we can get the test going is where my head is kinda at | 14:30 |
| lucasagomes | you know databases better than I. Is it too risky? I'm most wondering if the reasons why tests doesn't work is because of the database backend | 14:30 |
| lucasagomes | NobodyCam, yeah | 14:31 |
| lucasagomes | NobodyCam, I'm afraid we merge something and then boom we break like postgres or other databases | 14:31 |
| lucasagomes | :/ | 14:31 |
| NobodyCam | :) | 14:32 |
| NobodyCam | oh and a /shamelessplug for a doc patch.. https://review.openstack.org/#/c/168095 | 14:32 |
| *** stendulker has quit IRC | 14:32 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Proper errors handling for client https://review.openstack.org/168354 | 14:33 |
| *** stendulker has joined #openstack-ironic | 14:34 | |
| BadCub | Need some core reviews on: https://review.openstack.org/#/c/168175/ today too | 14:34 |
| *** achanda has joined #openstack-ironic | 14:35 | |
| BadCub | anyone know devstack folks to help out on ^ | 14:36 |
| NobodyCam | JayF: as jroll is on PTO and Josh is not yet on I'm pinging you. why a ten minute timeout? | 14:37 |
| jroll | NobodyCam: who what | 14:37 |
| NobodyCam | lol | 14:37 |
| NobodyCam | oh your still here | 14:38 |
| jroll | what's the question? | 14:38 |
| NobodyCam | jroll: on the patch BadCub linked ^^^ | 14:38 |
| jroll | yeah, hacking on a side project | 14:38 |
| jroll | :| | 14:38 |
| NobodyCam | josh bumped the timeout to ten minutes | 14:38 |
| jroll | I have no idea | 14:38 |
| jroll | seems like we shouldn't need to change it | 14:39 |
| NobodyCam | thats going make the gate slower :( | 14:39 |
| jroll | and if we do probably not so much | 14:39 |
| NobodyCam | 180 I understood | 14:39 |
| NobodyCam | but 600 seems a bit high for gate testing | 14:39 |
| NobodyCam | esp because ereasing hdd is also turned off with that patch | 14:40 |
| jroll | right, I don't understand why that change is needed | 14:40 |
| * NobodyCam waits for Josh | 14:40 | |
| NobodyCam | lol | 14:40 |
| *** zz_jgrimm is now known as jgrimm | 14:41 | |
| *** achanda has quit IRC | 14:42 | |
| * BadCub needs more coffee | 14:44 | |
| * TheJulia needs coffee in general | 14:44 | |
| *** stendulker has quit IRC | 14:44 | |
| *** mgoddard1 has joined #openstack-ironic | 14:45 | |
| NobodyCam | TheJulia: are you out of coffee? | 14:45 |
| BadCub | hey jroll NobodyCam you guys (and hopefully a few others) up for a bug squash session on MOnday? | 14:45 |
| NobodyCam | sure | 14:45 |
| *** mgoddard has quit IRC | 14:45 | |
| jroll | BadCub: I think I could, later in the week would be better for me | 14:45 |
| TheJulia | NobodyCam: can't have any until after a medical appointment | 14:46 |
| BadCub | jroll: later in week works too :-) How about we group up on Monday and pick a day for it | 14:46 |
| NobodyCam | TheJulia: :( oh I'm sorry... I hate those tests | 14:46 |
| jroll | BadCub: sounds good | 14:47 |
| BadCub | awesomeness! :-) | 14:47 |
| *** saripurigopi has joined #openstack-ironic | 14:47 | |
| NobodyCam | brb | 14:48 |
| TheJulia | Excpet to see my hyped up on caffee around 7 PM Eastern :) | 14:48 |
| TheJulia | caffine | 14:49 |
| *** rameshg87 is now known as rameshg87-dinner | 14:54 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: [WIP] Add DevStack plugin for ironic-discoverd https://review.openstack.org/164782 | 14:55 |
| *** Marga_ has quit IRC | 14:57 | |
| * NobodyCam is back | 14:58 | |
| NobodyCam | TheJulia: how long are you excluded from caffine? | 14:59 |
| TheJulia | NobodyCam: until after my appointment today | 15:02 |
| NobodyCam | :( | 15:02 |
| * BadCub is back | 15:07 | |
| BadCub | TheJulia: my gods, what time is your appointment? | 15:08 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Prevent updating the node's driver if console is enabled https://review.openstack.org/168030 | 15:09 |
| TheJulia | BadCub: 4.... I'm already starting to get a headache :( | 15:10 |
| BadCub | :( | 15:10 |
| *** devlaps has joined #openstack-ironic | 15:10 | |
| NobodyCam | TheJulia: will we be seeing a new rev of 168120 today? | 15:11 |
| NobodyCam | just checking... not pushing... | 15:11 |
| *** atuvenie has joined #openstack-ironic | 15:12 | |
| TheJulia | NobodyCam: Not likely given the caffine headache, likely over the weekend | 15:13 |
| *** rameshg87-dinner is now known as rameshg87 | 15:14 | |
| NobodyCam | :) that works :) thou weekend is personal time. so monday is also a good time :-p | 15:14 |
| TheJulia | Maybe later tonight after I get to have coffee | 15:14 |
| NobodyCam | wb rameshg87 | 15:14 |
| rameshg87 | lucasagomes: it goes fine for me for ubuntu | 15:14 |
| rameshg87 | thanks NobodyCam :) | 15:14 |
| NobodyCam | :) | 15:14 |
| rameshg87 | lucasagomes: will try doing a localboot once for bios and uefi | 15:15 |
| lucasagomes | rameshg87, hmm odd | 15:16 |
| lucasagomes | rameshg87, so it builds the image with grub2 installed? | 15:16 |
| * lucasagomes lemme try for ubuntu | 15:16 | |
| rameshg87 | lucasagomes: yeah it does | 15:16 |
| rameshg87 | lucasagomes: trying with grub-efi | 15:17 |
| rameshg87 | lucasagomes: if it goes through, i will try a deploy with that image with localboot | 15:18 |
| rameshg87 | lucasagomes: https://review.openstack.org/#/c/168030/3/ironic/api/controllers/v1/node.py | 15:18 |
| rameshg87 | lucasagomes: isn't 409 better ? | 15:18 |
| rameshg87 | lucasagomes: 400 - The request could not be understood by the server due to malformed syntax. | 15:18 |
| rameshg87 | lucasagomes: 409 - The request could not be completed due to a conflict with the current state of the resource. | 15:18 |
| lucasagomes | rameshg87, oh yeah 409 seems better indeed | 15:19 |
| lucasagomes | rameshg87, will update 1 sec | 15:19 |
| rameshg87 | okay | 15:19 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Prevent updating the node's driver if console is enabled https://review.openstack.org/168030 | 15:20 |
| lucasagomes | here we go | 15:20 |
| rameshg87 | :) | 15:21 |
| *** kkoski has quit IRC | 15:23 | |
| lucasagomes | rameshg87, odd, generating with ubuntu for me stops without any error message | 15:23 |
| lucasagomes | rameshg87, http://paste.openstack.org/show/197176/ | 15:23 |
| rameshg87 | lucasagomes: has it something to got with the machine you are building ? | 15:23 |
| rameshg87 | lucasagomes: i am on Ubuntu 14.04 LTS | 15:24 |
| lucasagomes | testing on my ubuntu vm, so the host is Ubuntu 14.04.1 LTS | 15:24 |
| * lucasagomes will try on a f21 host | 15:24 | |
| rameshg87 | :) | 15:24 |
| rameshg87 | lucasagomes: i am facing dependency issues when i add grub-efi :) | 15:24 |
| lucasagomes | I thought that installing a package would be like, trivial! | 15:24 |
| rameshg87 | hehe | 15:25 |
| rameshg87 | seems like it is not | 15:25 |
| rameshg87 | lucasagomes: when i add grub-efi, i am getting stranger errors like "grub2 : Depends: grub-pc (= 2.02~beta2-9ubuntu1) but it is not going to be installed" | 15:25 |
| rameshg87 | " but it is not going to be installed", wondering why ? | 15:25 |
| lucasagomes | grub-efi is grub 2 ? or grub legacy (1) ? | 15:26 |
| lucasagomes | tho it says 2.02~beta seems like grub2 | 15:26 |
| rameshg87 | lucasagomes: it says it's grub2 | 15:26 |
| rameshg87 | grub-efi - GRand Unified Bootloader, version 2 (dummy package) | 15:27 |
| lucasagomes | right | 15:28 |
| lucasagomes | hmm /me don't know | 15:29 |
| lucasagomes | I'm testing grub2 on f21 now | 15:29 |
| JayF | NobodyCam: I suspect joshnang bumped the timeout so aggressively because we wanted to see howlong it would take | 15:29 |
| lucasagomes | takes a while cause I dpon't have the images cached there | 15:29 |
| lucasagomes | base images* | 15:29 |
| *** jjohnson2 has quit IRC | 15:31 | |
| NobodyCam | JayF: should it be adjusted down on that patch to land? | 15:32 |
| *** anderbubble has joined #openstack-ironic | 15:32 | |
| rameshg87 | lucasagomes: it's going to be more than a simple package-installs.yaml i think | 15:32 |
| rameshg87 | lucasagomes: i can't install grub-pc and grub-efi together | 15:32 |
| rameshg87 | it conflicts | 15:32 |
| rameshg87 | :( | 15:32 |
| JayF | NobodyCam: I don't know until I see the results of the tempest run on the Ironic side | 15:32 |
| JayF | NobodyCam: but realistically, the devstack change shouldn't land until we have tempest passing on the ironic chang | 15:33 |
| JayF | *change | 15:33 |
| rameshg87 | makes sense because you cannot have both on the same node | 15:33 |
| lucasagomes | rameshg87, got the same result on f21 http://paste.openstack.org/show/197177/ | 15:33 |
| lucasagomes | rameshg87, x.x | 15:33 |
| NobodyCam | JayF: ++ | 15:33 |
| rameshg87 | lucasagomes: this succeeded no ? | 15:33 |
| BadCub | heya folks. Can I get a non-HP review/+a on https://review.openstack.org/#/c/165559 ? We have two +2s on it already | 15:33 |
| lucasagomes | rameshg87, nop | 15:34 |
| lucasagomes | rameshg87, just exit, no image was created | 15:34 |
| rameshg87 | lucasagomes: oh | 15:34 |
| NobodyCam | lol so happy to not be at the seattle office today :-p as its Comicon! | 15:35 |
| *** JoshNang has joined #openstack-ironic | 15:36 | |
| lucasagomes | NobodyCam, oh sounds good actually | 15:36 |
| NobodyCam | lol | 15:36 |
| NobodyCam | they just sent out a email saying no parking at the office :-p | 15:36 |
| JayF | BadCub: I put a +1 on that, so has mrda, it's proabbly safe for NobodyCam to land :) https://review.openstack.org/#/c/165559 | 15:36 |
| lucasagomes | NobodyCam, I see | 15:37 |
| lucasagomes | well at least you could have lunch with spiderman | 15:37 |
| NobodyCam | lol... JayF awesome ! | 15:37 |
| NobodyCam | lol | 15:37 |
| BadCub | JayF: thanks!1 :-) | 15:37 |
| JayF | We'll use the crap out of that env var too :) | 15:37 |
| NobodyCam | so click the +a will make our client's default 1.6 | 15:37 |
| NobodyCam | lucasagomes: did you want to have a look at ^^^ | 15:38 |
| JayF | Yeah, that's fine with me from my perspective; we already pass around files full of environment to setup creds for ironicclient, it's nbd to add an API version to that which we adjust as needed | 15:38 |
| lucasagomes | NobodyCam, what was the consensus on the client? | 15:39 |
| NobodyCam | lucasagomes: what part of the client? | 15:40 |
| lucasagomes | NobodyCam, passing a version >1.1 | 15:40 |
| lucasagomes | NobodyCam, I remember we had a patch in Nova | 15:40 |
| lucasagomes | was it merged? /me reads the comments | 15:40 |
| lucasagomes | oh ok it got merged | 15:42 |
| lucasagomes | https://review.openstack.org/#/c/149760/ | 15:43 |
| lucasagomes | +a | 15:44 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: [WIP] Add DevStack plugin for ironic-discoverd https://review.openstack.org/164782 | 15:44 |
| BadCub | brb | 15:44 |
| *** kkoski has joined #openstack-ironic | 15:49 | |
| *** dencaval has quit IRC | 15:52 | |
| *** jrist has quit IRC | 15:55 | |
| rameshg87 | lucasagomes: it goes through again for me :) | 15:57 |
| rameshg87 | with efi | 15:57 |
| lucasagomes | rameshg87, ouch, how come | 15:57 |
| lucasagomes | :/ | 15:57 |
| lucasagomes | this vm is like almost fresh install | 15:57 |
| *** kkoski has quit IRC | 15:58 | |
| rameshg87 | lucasagomes: http://paste.openstack.org/show/197187/ | 15:58 |
| *** kkoski has joined #openstack-ironic | 15:59 | |
| rameshg87 | lucasagomes: okay, time to go for me .. | 16:01 |
| rameshg87 | lucasagomes: i will try to check regarding efi vs bios thing | 16:01 |
| *** kkoski has quit IRC | 16:02 | |
| rameshg87 | lucasagomes: please drop me some indication in review if you happen to solve your ubuntu failure | 16:02 |
| rameshg87 | lucasagomes: i will try once in fedora as well | 16:02 |
| rameshg87 | good night all | 16:02 |
| rameshg87 | see you next week .. happy weekend | 16:03 |
| lucasagomes | rameshg87, ok thanks for the tests, have a good night! | 16:03 |
| rameshg87 | bye | 16:03 |
| lucasagomes | bye bye | 16:03 |
| lucasagomes | enjoy the weekend | 16:03 |
| *** rameshg87 has quit IRC | 16:03 | |
| * BadCub goes to hunt breakfast things bbiab | 16:03 | |
| *** foexle has quit IRC | 16:05 | |
| NobodyCam | have a good weekend rameshg87 | 16:06 |
| *** jrist has joined #openstack-ironic | 16:08 | |
| *** jrist has quit IRC | 16:08 | |
| *** jrist has joined #openstack-ironic | 16:08 | |
| *** kkoski has joined #openstack-ironic | 16:08 | |
| *** Marga_ has joined #openstack-ironic | 16:09 | |
| *** atuvenie has quit IRC | 16:09 | |
| *** ChuckC has joined #openstack-ironic | 16:11 | |
| *** kkoski has quit IRC | 16:19 | |
| *** kkoski has joined #openstack-ironic | 16:20 | |
| tiagogomes_ | ok, I have now my baremetal server having DHCP from Neutron (192.168.100.18), but I seems that he can't contact the TFTP server (10.24.1.113), do I need to add a route somewhere | 16:24 |
| NobodyCam | tiagogomes_: humm is this a flat network? | 16:27 |
| tiagogomes_ | NobodyCam, yes | 16:27 |
| *** ijw has joined #openstack-ironic | 16:27 | |
| *** ijw has quit IRC | 16:28 | |
| tiagogomes_ | If I set the IP of the tftp server to be on 192.168.100.10 e.g. I get an ARP timeout error | 16:28 |
| NobodyCam | the addresses seem strange to me, what type of setup are you using, DIB, Devstack or??? | 16:28 |
| *** ijw has joined #openstack-ironic | 16:28 | |
| tiagogomes_ | I don't know what DIV is | 16:28 |
| tiagogomes_ | I am integrating Ironic on baserock | 16:29 |
| NobodyCam | Dib == Disk Image Builder == TripleO | 16:29 |
| tiagogomes_ | I created a network for 192.168.100.0./24, where I run the DHCP server | 16:29 |
| tiagogomes_ | I've have a physical interface with IP 10.24.1.113, which is the IP for the TFTP server passed to the baremetal server | 16:30 |
| *** kkoski has quit IRC | 16:31 | |
| *** kkoski has joined #openstack-ironic | 16:32 | |
| NobodyCam | when you got the arp error was the tftp server bound to a 192.168.100 ip? | 16:32 |
| JayF | Does anyone know how to do a docs build locally? | 16:33 |
| tiagogomes_ | NobodyCam yes | 16:33 |
| NobodyCam | JayF: for ironic I do tox -egendocs | 16:33 |
| JayF | sweet, I thought it was something like that but didn't see any ref to it in the tox.ini | 16:34 |
| NobodyCam | command is in tox.ini I think | 16:34 |
| NobodyCam | https://github.com/openstack/ironic/blob/master/tox.ini#L45 | 16:35 |
| * BadCub updates pad with new bug(s) | 16:35 | |
| NobodyCam | tiagogomes_: any firewalls? anything in the log about tftp requests? | 16:36 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Use node UUID in logs instead of node ID https://review.openstack.org/167623 | 16:36 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add DevStack plugin for ironic-discoverd https://review.openstack.org/164782 | 16:43 |
| dtantsur | folks, I can barely believe myself, but seems like devstack support for discoverd started working somehow | 16:43 |
| dtantsur | and you can try it by following https://etherpad.openstack.org/p/DiscoverdDevStack | 16:43 |
| tiagogomes_ | NobodyCam no firewalls, the tftp server didn't received any request | 16:44 |
| tiagogomes_ | on tcpdump: 16:43:19.853798 ARP, Ethernet (len 6), IPv4 (len 4), Request who-has onenode tell 192.168.100.18, length 46 | 16:45 |
| *** subscope has quit IRC | 16:47 | |
| NobodyCam | humm | 16:48 |
| NobodyCam | dtantsur: awesome!!!!! | 16:48 |
| dtantsur | ifarkas, re rebooting in power off: there's no such written rule, but IIRC majority of drivers do handle this case properly (aka just power on) | 16:50 |
| *** alexpilotti has joined #openstack-ironic | 16:50 | |
| dtantsur | NobodyCam, devananda, your opinion ^^^ | 16:50 |
| ifarkas | dtantsur, yeah, for the sake of consistency it should this way | 16:50 |
| ifarkas | dtantsur, maybe it should also be added somewhere in the docs | 16:51 |
| dtantsur | ifarkas, makes sense to me. I will propose the patch, we'll see what people think | 16:52 |
| ifarkas | dtantsur, ack | 16:52 |
| NobodyCam | dtantsur: I'd have to test ipmitool will try and power off a node : https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/ipmitool.py#L671 | 16:56 |
| dtantsur | NobodyCam, well, it works for me for some time already :) | 16:57 |
| dtantsur | we had problems with DRAC driver | 16:57 |
| JayF | WTF? Something I put in the cleaning.rst apparently broke the docs build... | 16:57 |
| JayF | The traceback is a mile long and includes little/no useful output | 16:57 |
| NobodyCam | dtantsur: ssh driver? | 16:57 |
| NobodyCam | JayF: :( | 16:58 |
| lucasagomes | dtantsur, w00t | 16:58 |
| dtantsur | NobodyCam, works for both ssh and ipmitool in my tests | 16:58 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Explictly state that reboot is expected to work with powered off nodes https://review.openstack.org/168420 | 16:59 |
| dtantsur | NobodyCam, ifarkas, ^^^ | 17:00 |
| *** davide has joined #openstack-ironic | 17:00 | |
| NobodyCam | cool! | 17:00 |
| dtantsur | small docstring change to make it the official behavior | 17:00 |
| NobodyCam | ya ssh has some funky sutff like: https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/ssh.py#L439-L440 | 17:00 |
| dtantsur | (we don't have docs on how to write drivers, do we?) | 17:00 |
| NobodyCam | dtantsur: I think we do | 17:00 |
| NobodyCam | noe sec | 17:00 |
| ifarkas | dtantsur, thanks! | 17:00 |
| *** vdrok is now known as vdrok_afk | 17:02 | |
| NobodyCam | dtantsur: I was wrong I was thinking about this: http://docs.openstack.org/developer/ironic/dev/vendor-passthru.html#writing-vendor-methods | 17:02 |
| *** Marga_ has quit IRC | 17:03 | |
| NobodyCam | brb | 17:12 |
| *** anderbubble has quit IRC | 17:12 | |
| *** jerryz_ has quit IRC | 17:12 | |
| *** jistr has quit IRC | 17:13 | |
| devananda | zigo_: ping | 17:13 |
| *** jmccrory has joined #openstack-ironic | 17:14 | |
| devananda | zigo_: nvm :) | 17:14 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Add documentation for Cleaning https://review.openstack.org/168189 | 17:14 |
| devananda | dtantsur: yes, reboot of powered-off node should just work | 17:14 |
| JayF | aweeks: ^ that's what I wanted you to review, please | 17:14 |
| aweeks | JayF: looking | 17:15 |
| *** kkoski has quit IRC | 17:15 | |
| devananda | dtantsur: NobodyCam: afaik, no. we have some docs (possibly old) on the driver API and what the design goals of that API were, but not explicitly a "how to write a driver" doc | 17:15 |
| devananda | dtantsur: or if there is, it's recent and I haven't seen it yet | 17:15 |
| dtantsur | at least I haven't found it, so I proposed a change to docstring | 17:16 |
| devananda | NobodyCam: fwiw, I think diskimage-builder is useful (and is actually being used) outside of tripleo now | 17:16 |
| *** achanda has joined #openstack-ironic | 17:16 | |
| devananda | dtantsur: cool, thanks | 17:16 |
| *** achanda has quit IRC | 17:16 | |
| *** achanda has joined #openstack-ironic | 17:17 | |
| aweeks | JayF: are you looking for just content, or also style comments? | 17:18 |
| JayF | aweeks: I'm looking to make the best doc possible | 17:19 |
| dtantsur | and now calling it a day/week! have a nice weekend | 17:19 |
| *** dtantsur is now known as dtantsur|afk | 17:19 | |
| JayF | aweeks: any suggestions that involve "rewrite most of it" will be ignored :P | 17:19 |
| aweeks | lol | 17:19 |
| * aweeks has a tendency to heavily rewrite anything he edits, so be warned | 17:19 | |
| JayF | Note this is a review, not an edit :P | 17:19 |
| aweeks | touché | 17:20 |
| JayF | JoshNang: you probably want to look at https://review.openstack.org/#/c/168189/ to look over my changes | 17:20 |
| *** ijw has quit IRC | 17:22 | |
| *** ndipanov has quit IRC | 17:25 | |
| NobodyCam | morning devananda | 17:26 |
| NobodyCam | :-p | 17:26 |
| NobodyCam | devananda: oh cool! | 17:26 |
| devananda | NobodyCam: good afternoon | 17:26 |
| NobodyCam | :) | 17:26 |
| BadCub | heya devananda | 17:26 |
| *** anderbubble has joined #openstack-ironic | 17:27 | |
| NobodyCam | have a good weekend dtantsur|afk | 17:28 |
| *** ijw has joined #openstack-ironic | 17:30 | |
| *** bauzas is now known as bauwser | 17:30 | |
| *** anderbubble has quit IRC | 17:33 | |
| openstackgerrit | Merged openstack/python-ironicclient: Send version header by default https://review.openstack.org/165559 | 17:34 |
| NobodyCam | devananda: ^^^^^ :) | 17:34 |
| devananda | wooot! | 17:34 |
| devananda | did someone test that with stable/juno nova? | 17:35 |
| lucasagomes | devananda, nop, but adam_g has a patch there | 17:35 |
| lucasagomes | to reset the version to 1.0 | 17:35 |
| lucasagomes | I think that was the plan no? | 17:35 |
| devananda | no | 17:35 |
| devananda | https://review.openstack.org/#/c/167038/ | 17:35 |
| NobodyCam | adam_g: have this for stable juno https://review.openstack.org/#/c/167038/ | 17:35 |
| *** Marga_ has joined #openstack-ironic | 17:35 | |
| devananda | i mean, yes, that's the plan | 17:35 |
| devananda | but it hasn't merged yet | 17:35 |
| lucasagomes | yeah, becasue that was waiting #165559 to merge first | 17:36 |
| devananda | so I still can't tag a release of the client without breaking stable/juno :( | 17:36 |
| lucasagomes | yeah :/ | 17:36 |
| NobodyCam | devananda: adam_g wip'd that waiting to see the outcome of https://review.openstack.org/#/c/164369 | 17:36 |
| lucasagomes | there's some stuff that needs to get merged first | 17:36 |
| devananda | ugh. no. that's NOT the right fix. | 17:36 |
| lucasagomes | https://review.openstack.org/#/c/164369/ | 17:36 |
| devananda | it will work. but we'll still break any *existing* installations of Nova | 17:37 |
| devananda | if they somehow update the version of python-ironicclient | 17:37 |
| devananda | I'm not thrilled by that, but we need to move forard | 17:37 |
| devananda | forward | 17:37 |
| lucasagomes | :/ | 17:37 |
| lucasagomes | yeah it doesn't sounds good | 17:37 |
| * devananda starts compiling changelog for the client | 17:37 | |
| lucasagomes | we should merge the dependent stuff soonish | 17:38 |
| NobodyCam | lucasagomes: got a free minute to take a look at 164369? | 17:38 |
| lucasagomes | NobodyCam, yup | 17:39 |
| lucasagomes | devananda, hmm will that be equivalent to juno? I'm thinking about the JSON fields of the API | 17:40 |
| lucasagomes | that merged in kilo | 17:40 |
| lucasagomes | that wasn't supported by juno | 17:40 |
| lucasagomes | I thought we would kept the 1.1 to indicate that | 17:40 |
| NobodyCam | devananda: if you could also take a quck look just to make sure my changes are still in line with what you had in mind. | 17:40 |
| NobodyCam | oh also while your still up! JayF found a issue with target_provision_state yesterday | 17:41 |
| NobodyCam | JayF: I've lost the link.. do you still happen to have the link for target_provision_state thing you found yesterday | 17:42 |
| JayF | http://logs.openstack.org/96/165496/2/check/check-tempest-dsvm-ironic-agent_ssh/4d11e22/console.html#_2015-03-24_19_33_03_046 | 17:42 |
| JayF | NobodyCam: I was already fishing through my logs :D | 17:43 |
| NobodyCam | TY :) | 17:43 |
| NobodyCam | lol | 17:43 |
| BadCub | brb | 17:44 |
| aweeks | JayF: commented | 17:47 |
| JayF | ty | 17:47 |
| devananda | lucasagomes: which dependent stuff? | 17:47 |
| *** spandhe has joined #openstack-ironic | 17:47 | |
| lucasagomes | devananda, https://review.openstack.org/#/c/137762/ | 17:47 |
| lucasagomes | devananda, if we meant Kilo 1.0 to be exactly like Juno | 17:47 |
| lucasagomes | that won't be true | 17:47 |
| lucasagomes | because in Kilo we added support for other types of value on some attributes in our api | 17:48 |
| lucasagomes | that is not supported by Juno | 17:48 |
| *** pelix has quit IRC | 17:49 | |
| lucasagomes | so one still can put a dict or a list in the Node.extra attribute for e.g with Kilo 1.0 | 17:49 |
| lucasagomes | but not with Juno | 17:49 |
| devananda | lucasagomes: meh. it's not going to raise new or different errors for things that worked in Juno | 17:49 |
| *** htrmeira has joined #openstack-ironic | 17:49 | |
| devananda | lucasagomes: someone who thinks they're talking to Juno will get the results they expect | 17:49 |
| lucasagomes | devananda, it will raise 200 if you try with Kilo 1.0 and 400 Juno | 17:49 |
| devananda | right? | 17:49 |
| lucasagomes | no for things that worked with Juno it will work fine | 17:50 |
| lucasagomes | because juno just worked with strings and ints for those fields | 17:50 |
| lucasagomes | and those types still supported | 17:50 |
| devananda | yah. so it's not breaking | 17:50 |
| lucasagomes | we just added more | 17:50 |
| devananda | right | 17:50 |
| lucasagomes | yeah it's not breaking | 17:50 |
| devananda | SO | 17:50 |
| lucasagomes | but I wanted to point out that it's not 100% the same | 17:50 |
| *** rsalevsky has left #openstack-ironic | 17:50 | |
| devananda | we collectively need to decide whether the right thing is a) only bump microversions for breaking changes, b) bump for any API change, c) something in between these | 17:50 |
| JayF | aweeks: great review, ty | 17:51 |
| devananda | lucasagomes: thanks. i htink it's fine | 17:51 |
| aweeks | JayF: my pleasure | 17:51 |
| lucasagomes | devananda, yeah, to be honest. At first I thought it would be only for breaking changes | 17:51 |
| lucasagomes | because things like adding a new endpoint shouldn't break old versions | 17:51 |
| lucasagomes | it's just an addon | 17:52 |
| lucasagomes | but apparently we are doing it for any API change | 17:52 |
| lucasagomes | so if we get to vote I'm more inclined to go with A) | 17:52 |
| *** kbs has quit IRC | 17:52 | |
| *** kbs1 has joined #openstack-ironic | 17:53 | |
| devananda | lucasagomes: see, adding a new endpoint without bumping it kinda bothers me, if I can't discover that it's there | 17:53 |
| devananda | lucasagomes: one of the things that some REST APIs do is actually return the whole schema from eg, /v1/schema | 17:54 |
| devananda | lucasagomes: so that a client can actually discover what end points are there | 17:54 |
| devananda | I think we should do that. will make me much happier about going with (A) | 17:54 |
| NobodyCam | devananda: lucasagomes: should we add that question to the agenda for the next meeting? | 17:55 |
| lucasagomes | devananda, yeah the schema is a good idea | 17:55 |
| lucasagomes | we should add support for tht | 17:55 |
| devananda | NobodyCam: probably to the summit discussion on APIs | 17:55 |
| lucasagomes | that* | 17:55 |
| NobodyCam | devananda: ack | 17:55 |
| devananda | the next few meetings should continue to focus on bugs and summit planning | 17:55 |
| lucasagomes | NobodyCam, hmm wouldn't do any harm | 17:55 |
| lucasagomes | but it's really something that we need to discuss deep | 17:55 |
| lucasagomes | API is something that if you break it's very annoying | 17:55 |
| NobodyCam | ya so summit is best | 17:55 |
| lucasagomes | and users will hate us | 17:55 |
| *** derekh has quit IRC | 17:56 | |
| NobodyCam | devananda: added to https://etherpad.openstack.org/p/liberty-ironic-design-summit-ideas | 17:57 |
| devananda | NobodyCam: ty | 17:57 |
| NobodyCam | oh devananda did you see lucasagomes question ealier? | 17:58 |
| NobodyCam | 14:30 | lucasagomes > devananda, r u around? maybe you can take a look at mrda comments on https://review.openstack.org/#/c/165666 ? | 17:58 |
| lucasagomes | and folks I will call it a day | 17:58 |
| NobodyCam | have a great weekend lucasagomes | 17:58 |
| lucasagomes | oh yeah | 17:58 |
| jlvillal | lucasagomes: Good night | 17:58 |
| lucasagomes | devananda, specially because u understand dbs very well | 17:58 |
| * devananda looks | 17:58 | |
| NobodyCam | morning jlvillal :) | 17:59 |
| jlvillal | NobodyCam: Thanks :) | 17:59 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Add documentation for Cleaning https://review.openstack.org/168189 | 17:59 |
| JayF | aweeks: ^ | 17:59 |
| lucasagomes | jlvillal, night :) | 17:59 |
| lucasagomes | see y'all later, enjoy the weekend | 17:59 |
| * jlvillal wonders if NobodyCam has a WeeChat script that says good morning ;) | 17:59 | |
| devananda | lucasagomes: ciao! | 17:59 |
| NobodyCam | nope.. lol | 17:59 |
| jlvillal | NobodyCam: Your very fast! | 17:59 |
| *** lucasagomes is now known as lucas-beer | 17:59 | |
| devananda | NobodyCam, lucas-beer: what's the db concern on 165666? | 18:00 |
| lucas-beer | devananda, AFAIUI the tests doesn't work | 18:00 |
| NobodyCam | just basic irssi with only a changed theme otherwise stock out of the box | 18:00 |
| lucas-beer | but the migration script does | 18:00 |
| devananda | oh | 18:00 |
| lucas-beer | which makes me wonder, whether it's because of the tests itself | 18:00 |
| devananda | there's no migration test | 18:00 |
| lucas-beer | or some different db backend | 18:00 |
| lucas-beer | it's sounds risky to me | 18:00 |
| lucas-beer | devananda, yeah that's my -1 there | 18:01 |
| lucas-beer | cause there's no tests | 18:01 |
| devananda | jenkins passed the unit test changes | 18:01 |
| devananda | but yah. no db migration test | 18:01 |
| *** andreykurilin_ has joined #openstack-ironic | 18:02 | |
| lucas-beer | alright I'll step out | 18:02 |
| lucas-beer | see y'all! | 18:02 |
| NobodyCam | devananda: also see also see irc chat http://eavesdrop.openstack.org/irclogs/%23openstack-ironic/%23openstack-ironic.2015-03-26.log starting around 22:01:17 | 18:03 |
| NobodyCam | g'night lucas-beer | 18:03 |
| devananda | I see | 18:04 |
| aweeks | JayF: reviewed | 18:13 |
| JoshNang | aweeks: thanks! | 18:14 |
| aweeks | np | 18:14 |
| NobodyCam | oh hey JoshNang | 18:14 |
| NobodyCam | wanted to ask about the 600 second time out on 168175 | 18:15 |
| NobodyCam | is that needed for gate tests :( | 18:15 |
| JoshNang | NobodyCam: no no no | 18:15 |
| NobodyCam | phew :-p | 18:16 |
| JoshNang | NobodyCam: i was just trying to find if it was an error in the code and/or what the proper timeout would be | 18:16 |
| devananda | NobodyCam: patch tested and reviewed | 18:16 |
| JoshNang | trying to run tempest locally to do more testing | 18:16 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Add documentation for Cleaning https://review.openstack.org/168189 | 18:16 |
| NobodyCam | ack... :) | 18:16 |
| NobodyCam | TY devananda :) | 18:17 |
| devananda | NobodyCam: looks like it needs a tiny bit more work; i'm fine landing it soon and fixing the migration test situation following that | 18:17 |
| JayF | aweeks: fixed 2, disagreed about having the info about how to add steps to drivers, as that doc is currently targeted at deployers/operators and not developers | 18:17 |
| NobodyCam | ack :) | 18:17 |
| aweeks | JayF: yeah, that's reasonable | 18:17 |
| JayF | aweeks: IMO if drivers want to document how to add steps, they can (and I will add that when we get agent docs) | 18:17 |
| *** ijw has quit IRC | 18:17 | |
| NobodyCam | Ty devananda :) | 18:17 |
| NobodyCam | we'll make it happen while you running all over the EU :-p heheheh j/k | 18:18 |
| NobodyCam | s/you/your/ | 18:18 |
| NobodyCam | :-p | 18:18 |
| aweeks | JayF: giving it one final read over | 18:18 |
| JayF | aweeks: you should also review it in it's rendered form if you haven't | 18:18 |
| aweeks | JayF: ah, where would I find that? | 18:19 |
| JayF | aweeks: either wait a while for tests to run or tox -egendocs on that branch and file:///path/to/ironic/doc/build/ | 18:19 |
| aweeks | kk | 18:19 |
| JayF | aweeks: where, given the time and how many patchsets I've pushed "a while" can be a REALLY LONG TIME | 18:19 |
| devananda | something ya'll may be interested in (one of the outputs from this week of running all over the EU :P ) -- https://github.com/devananda/python-redfish | 18:19 |
| aweeks | heh, ok | 18:19 |
| devananda | all hail our new RESTful BMC overlords | 18:20 |
| NobodyCam | oh cool beans | 18:21 |
| * NobodyCam has to read | 18:21 | |
| * devananda drops offline to give a presentation to the local meetup group | 18:21 | |
| NobodyCam | enjoy the weekend devananda :) | 18:21 |
| NobodyCam | brb | 18:22 |
| *** hp-hj has joined #openstack-ironic | 18:23 | |
| *** hj-hp has quit IRC | 18:24 | |
| *** ifarkas has quit IRC | 18:25 | |
| *** hp-hj has quit IRC | 18:26 | |
| *** hp-hj has joined #openstack-ironic | 18:27 | |
| aweeks | JayF: +1ed | 18:28 |
| JayF | woo | 18:28 |
| *** penick has joined #openstack-ironic | 18:31 | |
| *** enikanorov has quit IRC | 18:32 | |
| *** Sukhdev has joined #openstack-ironic | 18:32 | |
| Sukhdev | Ironic folks, I have a need to access the ironic DB from the Neutron ML2 driver - any suggestion as to what is the best way to go? | 18:33 |
| JayF | That's not reasonable at all, I don't think | 18:34 |
| JayF | I would never give Ironic DB creds to my Neutron driver | 18:34 |
| JayF | What are you trying to do? There should be a way to make it work using API calls | 18:34 |
| *** penick_ has joined #openstack-ironic | 18:35 | |
| Sukhdev | JayF: I need to get the port information (the mac address and extra dict) from the Ironic port | 18:35 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Add documentation for Cleaning https://review.openstack.org/168189 | 18:35 |
| JayF | Sukhdev: that can be gotten with an API call | 18:35 |
| Sukhdev | JayF: can you please suggest the cleanest way? | 18:36 |
| JayF | jlvillal: JoshNang aweeks ^ Fixed jlvillal and aweeks's comments, would love to get your votes back | 18:36 |
| *** ijw has joined #openstack-ironic | 18:36 | |
| *** penick has quit IRC | 18:36 | |
| *** penick_ is now known as penick | 18:36 | |
| JayF | Sukhdev: on an Ironic node object, there are ports attached. Query the node, get the ports for the node, then query the ports | 18:36 |
| JayF | you'll get the port information returned | 18:36 |
| *** htrmeira has quit IRC | 18:37 | |
| jlvillal | JayF: done | 18:37 |
| JayF | ty | 18:37 |
| Sukhdev | JayF: I do not have node Id. All I have is mac address of the port - I saw in the models.py, there is method to get get_port_by_address() - wanted to use this method to get the extra dict info | 18:38 |
| JayF | Ah, so you have a MAC address, you need to map it back to a node | 18:38 |
| JayF | hm. | 18:38 |
| Sukhdev | JayF: correct | 18:38 |
| *** ijw_ has joined #openstack-ironic | 18:39 | |
| NobodyCam | can port filter by mac? | 18:39 |
| JayF | NobodyCam: it totally exists as a vendor_passthru | 18:39 |
| JayF | because this is how Agent looks up its node | 18:39 |
| NobodyCam | :) | 18:39 |
| JayF | Sukhdev: would an API call that gave you a node for a mac work? | 18:39 |
| JayF | Sukhdev: you give the macs, get the node back, then query the port object from there? | 18:39 |
| JayF | https://github.com/openstack/ironic-python-agent/blob/master/ironic_python_agent/ironic_api_client.py#L93 | 18:40 |
| Sukhdev | JayF: In neutron ML2 driver, I have mac address of the port of the BM server, want to get the connectivity information - which lives in extra dict | 18:40 |
| NobodyCam | JayF: seems like that a useful thing, are there plans U.S. it? | 18:40 |
| JayF | Yeah, so what you do is call vendor_passthru lookup like IPA does in that code ^ I just linked | 18:40 |
| JayF | then query the node object returned, it'll have the port objects attached, then just get those | 18:40 |
| JayF | NobodyCam: U.S. ? | 18:41 |
| JayF | NobodyCam: that is upstream | 18:41 |
| NobodyCam | ya | 18:41 |
| NobodyCam | just saw that | 18:41 |
| JayF | lol | 18:41 |
| NobodyCam | :-p | 18:41 |
| NobodyCam | its friday | 18:41 |
| * NobodyCam is slow | 18:41 | |
| JayF | that's actually one thing we should think about | 18:41 |
| JayF | I guess it doesn't matter though, only API call IPA makes is lookup | 18:41 |
| JayF | so microversioning doesn't matter, because lol vendor passthru | 18:42 |
| * BadCub adds cleaning doc to review-day pad & slaps it with a +1 | 18:42 | |
| JayF | that's likely something that can be upgraded out of passhtru and just added to the API | 18:42 |
| Sukhdev | JayF: in order to use iornic API - I was trying to use client_wrapper class - is that a wrong way to go? | 18:42 |
| *** ijw has quit IRC | 18:42 | |
| NobodyCam | JayF: ya in L :) | 18:42 |
| JayF | Sukhdev: I don't know what you're talking about, but when I've needed to talk to Ironic API I've used ironicclient. | 18:42 |
| JayF | python-ironicclient | 18:42 |
| Sukhdev | JayF: this is what I was trying to use - https://github.com/openstack/nova/blob/master/nova/virt/ironic/client_wrapper.py#L34 | 18:44 |
| JayF | Perfect, that uses the ironicclient under the hood | 18:44 |
| JayF | and has all the code to cache tokens and do reasonable retries | 18:45 |
| Sukhdev | JayF: would I have issue with the credentials for using python-ironicclient, when I am using it from neutron driver? | 18:46 |
| JayF | You'd have to have valid credentials to use Ironic if auth is enabled, yes | 18:46 |
| JayF | Whether or not you can reuse whatever creds Neutron uses for other things seems like an operational config question | 18:46 |
| Sukhdev | JayF: Is there a utility in ironic that will give me the admin credentials? | 18:47 |
| JayF | This is typical openstack auth; we use keystone | 18:47 |
| JayF | So you'd just have to have a keystone user with the proper role | 18:48 |
| Sukhdev | JayF: Is there admin role contex available in Ironic - something similar to Neutron | 18:50 |
| Sukhdev | ^context^ | 18:51 |
| JayF | https://github.com/openstack/ironic/blob/master/etc/ironic/policy.json | 18:51 |
| Sukhdev | JayF: as an example, this is how I access Neutron DB without going through API and avoiding the whole credential issue - https://github.com/stackforge/networking-arista/blob/master/networking_arista/common/db_lib.py#L315 | 18:52 |
| Sukhdev | JayF: the link which you gave me will not give the current admin context - I do not think so | 18:53 |
| JayF | You have to make calls to Ironic; it's a separate API service. Your code won't even be running on the same machines as Ironic. | 18:53 |
| JayF | NobodyCam: ^ do you know a basic guide for OpenStack SOA that could help Sukhdev | 18:53 |
| Sukhdev | JayF: I am assuming Ironc DB will be on the controller - and If I can peek into it, that would do the job | 18:54 |
| JayF | That's a terrible assumption | 18:54 |
| JayF | and as someone who runs Neutron and Ironic in production, there's no way I'd ever give direct access to my Ironic DB to Neutron or a neutron plugin | 18:55 |
| JayF | that's just not how service oriented things work | 18:55 |
| Sukhdev | JayF: Understood - | 18:55 |
| NobodyCam | hummm | 18:57 |
| NobodyCam | JayF: maybe https://github.com/openstack/api-site | 19:00 |
| NobodyCam | ??? | 19:00 |
| JayF | maybe | 19:01 |
| NobodyCam | Sukhdev: take a look at ^^^^ | 19:01 |
| NobodyCam | also https://wiki.openstack.org/wiki/Documentation/HowTo/FirstTimers | 19:01 |
| Sukhdev | NobodyCam: Will do - thanks | 19:02 |
| NobodyCam | may be a starting point | 19:02 |
| NobodyCam | :) | 19:02 |
| * NobodyCam is going to step away for a bit for a hairs cutting :-p | 19:02 | |
| Sukhdev | NobodyCam: I am familiar with Neutron and Nova APIs and their DB models and I am able to pull the admin credentials and access the DB. | 19:03 |
| * BadCub does too | 19:03 | |
| NobodyCam | Sukhdev: i would recamend not going direct to the DB | 19:04 |
| NobodyCam | back in a bit.. | 19:04 |
| Sukhdev | NobodyCam: can you suggest a painless way to get the credentials to be able to use Ironic API (I want to do it from the driver) | 19:05 |
| Sukhdev | NobodyCam: NP - thanks | 19:05 |
| *** tiagogomes_ has quit IRC | 19:06 | |
| *** alexpilotti has quit IRC | 19:07 | |
| *** kkoski has joined #openstack-ironic | 19:15 | |
| *** jmccrory has quit IRC | 19:26 | |
| *** andreykurilin__ has joined #openstack-ironic | 19:35 | |
| *** andreykurilin_ has quit IRC | 19:35 | |
| *** igordcard_ has joined #openstack-ironic | 19:35 | |
| clif_h | for the agent driver, it looks like deploy_kernel and deploy_ramdisk control which images get loaded for the agent | 19:36 |
| *** devlaps has quit IRC | 19:36 | |
| clif_h | can I manually change those on a created node and then ironic will pull the images I want? | 19:37 |
| clif_h | or will it still use value from node creation time? I'm sure I can do an experiment and find out, but if someone knows off the top of their head.. | 19:37 |
| JayF | I think you should change it, stop conductor and remove pxe configs, start conductor | 19:38 |
| JayF | there are likely ways that are ... less brutal ... to get it to pull down the other ramdisk | 19:38 |
| JayF | but I don't know them :) | 19:38 |
| clif_h | remove pxe configs? | 19:39 |
| JayF | the ones Ironic writes | 19:40 |
| JayF | likely in /tftpboot/, but I'm not 100% sure for devstack | 19:40 |
| JoshNang | /opt/stack/data/ironic/tftpboot | 19:41 |
| clif_h | JoshNang: thanks | 19:41 |
| JoshNang | clif_h: np. if you figure out an easier way to get a new agent image in than restacking, i'll be grateful | 19:42 |
| clif_h | what format are the agent images in that ipa creates? | 19:42 |
| JayF | think of them as like a kernel/initrd | 19:42 |
| JayF | basically a kernel + a CPIO archive that holds the root fs | 19:42 |
| clif_h | well, when I create an image in glance, don't I have to tell it? | 19:43 |
| clif_h | or will it detect on its own? | 19:43 |
| JayF | clif_h: I got it for you, 1s | 19:44 |
| JayF | https://github.com/openstack-dev/devstack/blob/master/lib/ironic#L723 | 19:44 |
| JayF | that's the method in devstack that builds + uploads the images to glance | 19:44 |
| JayF | you should be able to glean the command to run from there | 19:44 |
| clif_h | aha | 19:45 |
| clif_h | yep, thanks | 19:45 |
| JayF | line 771 looks like exactly what you want | 19:45 |
| clif_h | I should've looked at devstack | 19:45 |
| JayF | It's fine, we should have all this stuff documented | 19:45 |
| JayF | JoshNang: any success with agent tempest? | 19:45 |
| JoshNang | JayF: got distracted working on/testing nits | 19:46 |
| JayF | ah. | 19:46 |
| JayF | Just curious if you want to parlay this conversation about how agent things happen into starting on agent doc | 19:46 |
| JayF | that's what I'm going to do after lunch | 19:46 |
| JoshNang | i'm gonna keep cranking on this | 19:47 |
| JoshNang | deadlines and such | 19:47 |
| JayF | yeah igi | 19:47 |
| *** andreykurilin__ has quit IRC | 19:57 | |
| *** jmccrory has joined #openstack-ironic | 19:58 | |
| *** jmccrory1 has joined #openstack-ironic | 20:01 | |
| *** jmccrory has quit IRC | 20:03 | |
| *** jmccrory1 has quit IRC | 20:06 | |
| *** jmccrory has joined #openstack-ironic | 20:06 | |
| jroll | Sukhdev: curl $IRONIC_API_URL/v1/ports/detail?address=00:11:22:33:44:55 | 20:07 |
| *** mjturek1 has quit IRC | 20:08 | |
| *** achanda has quit IRC | 20:08 | |
| JayF | jroll: we added a real API endpoint for that? nice | 20:08 |
| jroll | we've had it for a while iirc | 20:09 |
| JayF | win 40 | 20:09 |
| JayF | oops | 20:09 |
| jroll | heh, this is 40 for me | 20:10 |
| JayF | you can use timing to guess which one is 40 for me | 20:11 |
| JayF | this is 7 for me | 20:11 |
| jroll | :P | 20:12 |
| *** achanda has joined #openstack-ironic | 20:12 | |
| openstackgerrit | Pierre Riteau proposed openstack/python-ironicclient: Add support for filtering nodes by chassis UUID https://review.openstack.org/167964 | 20:13 |
| *** mjturek1 has joined #openstack-ironic | 20:17 | |
| *** priteau has quit IRC | 20:20 | |
| *** lucas-beer has quit IRC | 20:24 | |
| openstackgerrit | John L. Villalovos proposed openstack/python-ironicclient: Add unittests for resrource_fields https://review.openstack.org/168476 | 20:32 |
| *** Sukhdev has quit IRC | 20:32 | |
| openstackgerrit | John L. Villalovos proposed openstack/python-ironicclient: Add unittests for resource_fields https://review.openstack.org/168476 | 20:32 |
| *** athomas has quit IRC | 20:34 | |
| *** achanda has quit IRC | 20:43 | |
| *** achanda has joined #openstack-ironic | 20:50 | |
| *** spandhe has quit IRC | 21:00 | |
| -openstackstatus- NOTICE: Gerrit maintenance commences in 1 hour at 22:00 UTC http://lists.openstack.org/pipermail/openstack-dev/2015-March/059948.html | 21:00 | |
| *** spandhe has joined #openstack-ironic | 21:02 | |
| *** pas-ha has quit IRC | 21:03 | |
| *** dprince has quit IRC | 21:06 | |
| *** achanda has quit IRC | 21:07 | |
| *** absubram has quit IRC | 21:11 | |
| *** absubram has joined #openstack-ironic | 21:12 | |
| jlvillal | BadCub: Is https://review.openstack.org/#/c/157602/ being tracked? | 21:18 |
| jlvillal | BadCub: I was just curious if that was one that has to wait for Liberty or not. | 21:19 |
| * BadCub looks | 21:19 | |
| BadCub | jlvillal: it doesn't appear to be a new feature. I tend to agree with the notes | 21:20 |
| *** Marga_ has quit IRC | 21:21 | |
| jlvillal | BadCub: Thanks. Seems like a good fix to me :) | 21:21 |
| *** trown is now known as trown|outttypeww | 21:21 | |
| BadCub | jlvillal: it does to me as well | 21:21 |
| *** achanda has joined #openstack-ironic | 21:27 | |
| *** saripurigopi has quit IRC | 21:31 | |
| *** jcoufal has quit IRC | 21:41 | |
| *** spandhe has quit IRC | 21:44 | |
| *** Marga_ has joined #openstack-ironic | 21:46 | |
| *** ChuckC has quit IRC | 21:56 | |
| *** kkoski has quit IRC | 21:57 | |
| *** sambetts has quit IRC | 22:01 | |
| -openstackstatus- NOTICE: Gerrit is offline for maintenance, ETA 22:30 UTC http://lists.openstack.org/pipermail/openstack-dev/2015-March/059948.html | 22:04 | |
| *** ChanServ changes topic to "Gerrit is offline for maintenance, ETA 22:30 UTC http://lists.openstack.org/pipermail/openstack-dev/2015-March/059948.html" | 22:04 | |
| *** sambetts has joined #openstack-ironic | 22:04 | |
| JayF | Anyone around in channel who knows things about our doc build? | 22:06 |
| JayF | Trying to port the Ironic sphinx build to IPA (to build dev docs) and not having any success | 22:06 |
| NobodyCam | i know we have a thing that does it | 22:07 |
| JayF | I have the build running in IPA, but it's erroring out in a ... unhelpful way | 22:07 |
| NobodyCam | what are you getting? | 22:08 |
| JayF | I think I may just need an index.rst | 22:09 |
| NobodyCam | your doing somehting like python setup.py build_sphinx ? | 22:09 |
| JayF | let me try this and if it doesn't work I'll push up my WIP and gist the output | 22:09 |
| JayF | gendocs: commands succeeded | 22:10 |
| JayF | \o/ | 22:10 |
| JayF | now to see if it's readable at all | 22:10 |
| JayF | give me about 5m and I'll have something useful to post and wip, I got the build working and I think I understand what it's doing... maybe. | 22:11 |
| NobodyCam | woo hoo | 22:12 |
| JayF | I was just thinking, agent docs shoudl be in the agent, re: docs on how to build hardware managers and similar | 22:12 |
| JayF | agent *driver and config* docs should go in Ironic | 22:12 |
| JayF | so first real step to agent docs == build the dev docs we already have | 22:13 |
| NobodyCam | :) yay for docs... | 22:13 |
| NobodyCam | lol if I understand BadCub will be looking into our docs to see what needed. ;p | 22:14 |
| NobodyCam | I.e. I don't think we have anything writtend down on the iboot powerdriver | 22:14 |
| *** spandhe has joined #openstack-ironic | 22:16 | |
| * BadCub thinks everyone needs docsw just because | 22:17 | |
| JayF | doh | 22:17 |
| JayF | I can't push my docs wip[ | 22:17 |
| JayF | because gerrit | 22:17 |
| BadCub | docs even lol | 22:17 |
| NobodyCam | doh | 22:18 |
| NobodyCam | lol | 22:18 |
| BadCub | docs are good for the soul! *WG* | 22:18 |
| *** ChuckC has joined #openstack-ironic | 22:20 | |
| *** kkoski has joined #openstack-ironic | 22:23 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: WIP: Initial commit of support for docs build https://review.openstack.org/168503 | 22:24 |
| NobodyCam | oh gerrit back? | 22:24 |
| JayF | they just said "gerrit is back, lets start zuul" in -infra, so I tried git-review again | 22:24 |
| BadCub | nice | 22:25 |
| NobodyCam | :) | 22:25 |
| *** ChuckC has quit IRC | 22:25 | |
| *** ChuckC has joined #openstack-ironic | 22:26 | |
| *** spandhe has quit IRC | 22:28 | |
| *** anderbubble has joined #openstack-ironic | 22:29 | |
| *** achanda_ has joined #openstack-ironic | 22:31 | |
| JayF | NobodyCam: question for you bro, proxying for clif_h | 22:32 |
| NobodyCam | lol really: It's an agent! Hooray! "-p | 22:32 |
| NobodyCam | hehehe | 22:32 |
| JayF | NobodyCam: We need a console in one of the devstack vm-baremetal-nodes | 22:32 |
| JayF | NobodyCam: note the "WIP" part of that :P | 22:32 |
| NobodyCam | lol | 22:32 |
| *** ChanServ changes topic to "Bare Metal Provisioning | Status: http://bit.ly/ironic-whiteboard | Docs: http://docs.openstack.org/developer/ironic/ | Bugs: https://bugs.launchpad.net/ironic" | 22:33 | |
| NobodyCam | I tend to just open the local kvm mgnt tool and click console | 22:33 |
| NobodyCam | but I expect your looking for more then that | 22:33 |
| JayF | Well most of us rackers run our devstacks in the cloud | 22:34 |
| JayF | so there's not a local kvm management tool :) | 22:34 |
| NobodyCam | ahh I have my taco laptop | 22:34 |
| *** achanda has quit IRC | 22:34 | |
| clif_h | NobodyCam: I try to use say, virsh console <bm_vm_name>, and it says error: internal error: character device <null> is not using a PTY | 22:34 |
| NobodyCam | so shellnabox stuff? | 22:34 |
| NobodyCam | ahh | 22:35 |
| NobodyCam | dib build image? | 22:35 |
| clif_h | dib? | 22:35 |
| NobodyCam | http://wiki.libvirt.org/page/Error_%22internal_error_cannot_find_character_device%22_when_trying_to_connect_a_domain's_console | 22:36 |
| *** achanda_ has quit IRC | 22:36 | |
| NobodyCam | disk image builder? | 22:36 |
| clif_h | it shows a serial port on port 1 | 22:37 |
| clif_h | can I not just use that | 22:37 |
| *** jcoufal has joined #openstack-ironic | 22:38 | |
| NobodyCam | the tripleo scripts do this: https://github.com/openstack/tripleo-incubator/blob/master/scripts/configure-vm#L120-L128 | 22:38 |
| JayF | There's def a way built-in to devstack to get the consoles | 22:39 |
| JayF | with no modification needed beyond that which is done in devstack-gate | 22:39 |
| JayF | adam_g: don't happen to be around, do you? ^ I feel like this is a question you know the answer to immediately | 22:39 |
| *** spandhe has joined #openstack-ironic | 22:44 | |
| NobodyCam | JayF: lol http://docs.openstack.org/user-guide/content/instance_console.html | 22:44 |
| * NobodyCam *ducks* | 22:45 | |
| JayF | honestly, it never occurred to me to think to use the api for the console | 22:46 |
| JayF | heh | 22:46 |
| JayF | clif_h: ^ maybe try that? Get the console URL from nova and tunnel to ti? | 22:46 |
| clif_h | vnc? | 22:47 |
| clif_h | wow | 22:47 |
| clif_h | through nova? | 22:48 |
| clif_h | I'm not provisioning nodes through nova | 22:48 |
| JayF | Yep, Ironic got support for it | 22:48 |
| JayF | ... | 22:48 |
| JayF | oh, that's right :( | 22:49 |
| clif_h | I'm making calls to cache the image directly to ironic | 22:49 |
| clif_h | which does the pxe boot stuff | 22:49 |
| JayF | there's an Ironic API endpoint to get the console working, but it occurs to me it might not be implemented by the ssh driver | 22:49 |
| JayF | you're right, bah, nevermind the crummy advice | 22:49 |
| clif_h | it's ok, thanks for trying | 22:49 |
| clif_h | my motivation meter for the day is almost toast though, so I'll probably hang it up until monday | 22:50 |
| *** kbs1 has quit IRC | 22:51 | |
| NobodyCam | :) | 22:52 |
| *** mark-lk has joined #openstack-ironic | 22:55 | |
| *** kbs has joined #openstack-ironic | 23:01 | |
| *** jmccrory has quit IRC | 23:06 | |
| *** kkoski has quit IRC | 23:25 | |
| openstackgerrit | Josh Gachnang proposed openstack/ironic: Add agent cleaning reboot and fix nits https://review.openstack.org/168197 | 23:30 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Add sphinx build + basic documentation https://review.openstack.org/168503 | 23:33 |
| JayF | ^ that is potentially landable | 23:33 |
| *** mark-lk has quit IRC | 23:33 | |
| JayF | until someone figures out the long list of things I did wrong with the build | 23:33 |
| JayF | heh | 23:33 |
| *** mark-lk has joined #openstack-ironic | 23:34 | |
| NobodyCam | :-p | 23:35 |
| openstackgerrit | Haomeng,Wang proposed openstack/ironic: Fix UnicodeEncodeError issue when the language is not en_US https://review.openstack.org/168268 | 23:38 |
| openstackgerrit | Josh Gachnang proposed openstack/ironic: Enable cleaning by default https://review.openstack.org/165496 | 23:40 |
| *** anderbubble has quit IRC | 23:47 | |
| *** andreykurilin_ has joined #openstack-ironic | 23:47 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!