| *** yangyapeng has joined #openstack-ironic | 00:00 | |
| *** Nisha_Agarwal has quit IRC | 00:00 | |
| *** yangyapeng has quit IRC | 00:04 | |
| *** dyasny has quit IRC | 00:07 | |
| *** Nisha_ has quit IRC | 00:11 | |
| *** rivers has quit IRC | 00:29 | |
| *** rivers has joined #openstack-ironic | 00:29 | |
| *** dyasny has joined #openstack-ironic | 00:50 | |
| *** dyasny has quit IRC | 00:50 | |
| *** mat128 has joined #openstack-ironic | 00:52 | |
| *** ijw_ has joined #openstack-ironic | 00:55 | |
| *** ijw has quit IRC | 00:58 | |
| *** yangyapeng has joined #openstack-ironic | 01:01 | |
| *** rbudden has joined #openstack-ironic | 01:02 | |
| *** ijw has joined #openstack-ironic | 01:04 | |
| *** ijw_ has quit IRC | 01:07 | |
| *** hshiina has joined #openstack-ironic | 01:13 | |
| *** dims has quit IRC | 01:21 | |
| *** litao__ has joined #openstack-ironic | 01:24 | |
| *** ijw has quit IRC | 01:29 | |
| *** adisky__ has joined #openstack-ironic | 01:30 | |
| *** dims has joined #openstack-ironic | 01:33 | |
| *** gcb has joined #openstack-ironic | 01:35 | |
| *** yangyapeng has quit IRC | 01:58 | |
| *** yangyapeng has joined #openstack-ironic | 01:59 | |
| *** yangyape_ has joined #openstack-ironic | 02:02 | |
| *** yangyapeng has quit IRC | 02:06 | |
| *** yangyapeng has joined #openstack-ironic | 02:12 | |
| *** yangyape_ has quit IRC | 02:15 | |
| *** rbudden has quit IRC | 02:35 | |
| *** Nisha_Agarwal has joined #openstack-ironic | 02:42 | |
| *** ianychoi_ has joined #openstack-ironic | 02:43 | |
| *** ianychoi has quit IRC | 02:46 | |
| *** mat128 has quit IRC | 02:52 | |
| *** mat128 has joined #openstack-ironic | 02:53 | |
| openstackgerrit | zenghui.shi proposed openstack/ironic-specs master: New hardware interface for BIOS configuration https://review.openstack.org/496481 | 03:15 |
|---|---|---|
| *** udesale has joined #openstack-ironic | 03:19 | |
| *** zhurong_ has joined #openstack-ironic | 03:22 | |
| *** zhurong_ has quit IRC | 03:25 | |
| *** hshiina has quit IRC | 03:34 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic-specs master: Add resource class structure to ironic https://review.openstack.org/500429 | 03:38 |
| *** Nisha_Agarwal has quit IRC | 03:42 | |
| *** stendulker has joined #openstack-ironic | 03:49 | |
| *** Dinesh_Bhor has joined #openstack-ironic | 03:53 | |
| *** hshiina has joined #openstack-ironic | 04:01 | |
| *** Nisha_Agarwal has joined #openstack-ironic | 04:01 | |
| *** wong__ has quit IRC | 04:16 | |
| *** links has joined #openstack-ironic | 04:19 | |
| *** persia has quit IRC | 04:20 | |
| *** mat128 has quit IRC | 04:21 | |
| *** Nisha_Agarwal has quit IRC | 04:36 | |
| *** MattMan_1 has quit IRC | 04:52 | |
| *** MattMan_1 has joined #openstack-ironic | 04:52 | |
| *** rbartal has joined #openstack-ironic | 05:06 | |
| *** Nisha_Agarwal has joined #openstack-ironic | 05:11 | |
| *** gcb has quit IRC | 05:15 | |
| *** gcb has joined #openstack-ironic | 05:29 | |
| *** wong_ has joined #openstack-ironic | 05:33 | |
| *** stendulker_ has joined #openstack-ironic | 05:35 | |
| *** stendulker has quit IRC | 05:35 | |
| *** hjensas has quit IRC | 05:45 | |
| *** hoonetorg has quit IRC | 06:05 | |
| *** ccamacho has joined #openstack-ironic | 06:08 | |
| *** mjura has joined #openstack-ironic | 06:10 | |
| *** mjura has quit IRC | 06:11 | |
| *** mjura has joined #openstack-ironic | 06:11 | |
| *** marios has joined #openstack-ironic | 06:17 | |
| *** hoonetorg has joined #openstack-ironic | 06:18 | |
| *** zshi_ has joined #openstack-ironic | 06:21 | |
| *** mjura has quit IRC | 06:23 | |
| *** zhangfei has joined #openstack-ironic | 06:24 | |
| *** mjura has joined #openstack-ironic | 06:26 | |
| *** chas has joined #openstack-ironic | 06:29 | |
| *** ijw has joined #openstack-ironic | 06:30 | |
| *** zhangfei has quit IRC | 06:34 | |
| *** ijw has quit IRC | 06:35 | |
| *** karimb has joined #openstack-ironic | 06:38 | |
| *** yangyapeng has quit IRC | 06:39 | |
| *** racedo has joined #openstack-ironic | 06:42 | |
| openstackgerrit | Merged openstack/python-ironicclient master: Fix to use "." to source script files https://review.openstack.org/500079 | 06:47 |
| *** yangyapeng has joined #openstack-ironic | 06:47 | |
| *** zhangfei has joined #openstack-ironic | 06:47 | |
| *** rcernin has quit IRC | 06:52 | |
| *** rcernin has joined #openstack-ironic | 06:52 | |
| *** hjensas has joined #openstack-ironic | 06:58 | |
| *** e0ne has joined #openstack-ironic | 07:02 | |
| *** Nisha_ has joined #openstack-ironic | 07:04 | |
| *** mvpnitesh has joined #openstack-ironic | 07:09 | |
| *** karthiks_afk is now known as karthiks | 07:10 | |
| *** tesseract has joined #openstack-ironic | 07:16 | |
| *** jtomasek has joined #openstack-ironic | 07:23 | |
| *** aarefiev_afk is now known as aarefiev | 07:24 | |
| *** etingof has quit IRC | 07:28 | |
| *** charcol has quit IRC | 07:28 | |
| *** jaganathan has joined #openstack-ironic | 07:28 | |
| *** dtantsur|afk is now known as dtantsur | 07:34 | |
| dtantsur | morning ironic | 07:34 |
| *** pcaruana has joined #openstack-ironic | 07:36 | |
| *** pcaruana has quit IRC | 07:36 | |
| *** pcaruana has joined #openstack-ironic | 07:37 | |
| *** stendulker_ has quit IRC | 07:37 | |
| *** etingof has joined #openstack-ironic | 07:39 | |
| *** lucas-afk is now known as lucasagomes | 07:41 | |
| *** rcernin has quit IRC | 07:41 | |
| *** oomichi has quit IRC | 07:51 | |
| *** oomichi has joined #openstack-ironic | 07:52 | |
| *** ralonsoh has joined #openstack-ironic | 07:53 | |
| *** Nisha_ has quit IRC | 07:56 | |
| Nisha_Agarwal | morning ironic | 07:56 |
| Nisha_Agarwal | hey dtantsur | 07:56 |
| Nisha_Agarwal | dtantsur, for py35 work we(myself and Ankit) were trying inspector manually before we modify the tempest tests in it so that it can work without swift. But we see this issue without doing any changes to inspector. http://paste.openstack.org/show/620297/ Initial ramdisk booting etc happens successfully but after IPA returns data that overwrites bmc_address, and inspection fails | 08:01 |
| Nisha_Agarwal | aarefiev, ^^^ | 08:01 |
| Nisha_Agarwal | dtantsur, this happens consistently for us. | 08:02 |
| Nisha_Agarwal | dtantsur, aarefiev is this expected behaviour? | 08:02 |
| *** stendulker_ has joined #openstack-ironic | 08:03 | |
| aarefiev | Nisha_Agarwal: hey, do you use master? | 08:07 |
| Nisha_Agarwal | aarefiev, we are using master for this | 08:07 |
| aarefiev | Nisha_Agarwal, hmmm, could you check if this change is present https://github.com/openstack/ironic-inspector/commit/2a4c9d800f24072dcbd149e194fbe0fa2464bdc4 | 08:08 |
| * Nisha_Agarwal checks | 08:11 | |
| Nisha_Agarwal | aarefiev, yes it has | 08:12 |
| aarefiev | ahh and it works on py2, right? | 08:13 |
| Nisha_Agarwal | no it is currently failing on py2 | 08:13 |
| Nisha_Agarwal | we didnt do any changes for py3 yet | 08:13 |
| Nisha_Agarwal | in past when i have tried inspector it has always worked | 08:14 |
| Nisha_Agarwal | do i need to do any conf change in ironic-inspector.conf | 08:15 |
| aarefiev | Nisha_Agarwal: do you have upstream patch where I can see failure ? | 08:15 |
| Nisha_Agarwal | No, this is master ironic-inspectr which we were trying to use with pxe_ipmitool driver for inband inspection on proliant hardwares. This is just a testing prior to make any changes for py35 | 08:16 |
| Nisha_Agarwal | so fr us inspection using inspector is always failing at this point. | 08:17 |
| aarefiev | Nisha_Agarwal: ok, do you have ironic port with one of [u'9c:b6:54:01:90:2e', u'9c:b6:54:01:90:2f'] | 08:19 |
| aarefiev | macs | 08:19 |
| Nisha_Agarwal | yes | 08:19 |
| dtantsur | aarefiev: I wonder if it's related to https://github.com/openstack/ironic-python-agent/commit/7c28332e9a59da316a9d6e6652f4269b27b3ca66 | 08:21 |
| dtantsur | Nisha_Agarwal: ^^ | 08:21 |
| *** mvpnitesh has quit IRC | 08:21 | |
| *** mvpnitesh has joined #openstack-ironic | 08:22 | |
| Nisha_Agarwal | dtantsur, hmmm may be . but we see the bmc address overwritten with 0.0.0.0 after IPA returns the data | 08:22 |
| dtantsur | also I still suspect that https://github.com/openstack/ironic-inspector/commit/2a4c9d800f24072dcbd149e194fbe0fa2464bdc4 is not present.. | 08:22 |
| Nisha_Agarwal | dtantsur, ironic-inspector is latest. it has changes till Aug24 | 08:23 |
| dtantsur | Nisha_Agarwal: anyway, can you grab IPA logs please? (the file with them is mentioned in the logs) | 08:23 |
| dtantsur | UGH https://github.com/openstack/ironic-inspector/blob/master/ironic_inspector/process.py#L86 | 08:24 |
| dtantsur | this needs fixing ^^^ | 08:24 |
| dtantsur | Nisha_Agarwal: mind filing a bug against both inspector and IPA? at least the inspector part has to be fixed (but I think IPA not returning BMC address for you is also wrong) | 08:25 |
| Nisha_Agarwal | dtantsur, sure. will do / | 08:26 |
| * Nisha_Agarwal gets IPA logs | 08:26 | |
| aarefiev | but there is mac == mac from ironic port in introspection data | 08:26 |
| *** derekh has joined #openstack-ironic | 08:26 | |
| *** mvpnitesh has quit IRC | 08:29 | |
| *** mgoddard has joined #openstack-ironic | 08:30 | |
| *** mvpnitesh has joined #openstack-ironic | 08:31 | |
| dtantsur | aarefiev: this is strange too.. | 08:33 |
| dtantsur | there are up to 3 different issues here :) | 08:33 |
| dtantsur | Nisha_Agarwal: I'll post the 1st patch when I have a bug number. could you please paste the complete inspector logs from the start of the introspection and till the finish? | 08:34 |
| dtantsur | * attach to the bug, I meant | 08:35 |
| Nisha_Agarwal | dtantsur, yes. re IPA logs, we will need to redo inspection, i somehow deleted all the logs :( | 08:35 |
| Nisha_Agarwal | we wil try it on different hardware and see if issue is seen | 08:36 |
| * Nisha_Agarwal will get back on this after sometime with re-inspection | 08:36 | |
| dtantsur | okay. currently I just need a bug # to use Partial-Bug on :) | 08:38 |
| *** mvpnitesh has quit IRC | 08:38 | |
| *** ankkumar has joined #openstack-ironic | 08:41 | |
| Nisha_Agarwal | dtantsur, sure. ankit is redoing the test. I will raise the bug with details then. But it was a strange thing for us too. | 08:41 |
| openstackgerrit | Merged openstack/ironic master: Updated from global requirements https://review.openstack.org/499998 | 08:44 |
| *** mvpnitesh has joined #openstack-ironic | 08:48 | |
| stendulker_ | dtantsur: Hi | 08:52 |
| dtantsur | hi stendulker_ | 08:52 |
| stendulker_ | dtantsur: Can you please have a look at this 'ilo' hardware type doc. https://review.openstack.org/#/c/496480/ | 08:52 |
| patchbot | patch 496480 - ironic - Update documentation for ``ilo`` hardware type | 08:52 |
| dtantsur | sure | 08:53 |
| stendulker_ | dtantsur: Will we be able to backport doc to pike release? | 08:53 |
| dtantsur | stendulker_: yes, as soon as the master version merges | 08:53 |
| stendulker_ | dtantsur:yes, upon merge. | 08:53 |
| stendulker_ | dtantsur: thank you. | 08:53 |
| openstackgerrit | zenghui.shi proposed openstack/ironic-specs master: New hardware interface for BIOS configuration https://review.openstack.org/496481 | 08:55 |
| *** mrch has joined #openstack-ironic | 09:02 | |
| openstackgerrit | yolanda.robla proposed openstack/bifrost stable/pike: Fix the dhcp checks on centos https://review.openstack.org/500485 | 09:05 |
| openstackgerrit | yolanda.robla proposed openstack/bifrost stable/ocata: Fix the dhcp checks on centos https://review.openstack.org/500486 | 09:06 |
| Tengu | dtantsur: do you have the issue link at hand? | 09:07 |
| dtantsur | I don't | 09:07 |
| dtantsur | note that the validate command should be only used to check power/management at the early stage | 09:08 |
| dtantsur | like in the example: openstack baremetal node validate <IRONIC NODE> | grep -E '(power|management)\W*False' | 09:08 |
| Tengu | dtantsur: ah, ok. thus the grep pipe. | 09:08 |
| Tengu | as it didn't show anything, I launched the command without the pipe, and saw the issues I mentionned on the other channel. | 09:08 |
| Tengu | dtantsur: here's the link: https://bugs.launchpad.net/tripleo/+bug/1714887 | 09:08 |
| openstack | Launchpad bug 1714887 in tripleo "[pike] Openstack overcloud deploy failed: not enough nodes" [Undecided,New] | 09:08 |
| openstackgerrit | yolanda.robla proposed openstack/bifrost stable/newton: Fix the dhcp checks on centos https://review.openstack.org/500487 | 09:10 |
| dtantsur | Tengu: how did you end up with these flavors? they don't look correct for pike.. | 09:10 |
| dtantsur | is it just the way it installed? if so, when did you install the undercloud? | 09:10 |
| Tengu | dtantsur: I created them in an ansible playbook installing the undercloud | 09:11 |
| Tengu | dtantsur: just a small note: that was working fine until I activate the disk_label thing in order to be able to take profit of 2.2To on the compute nodes | 09:11 |
| dtantsur | okay, it's not a blocking issue, but please check Fhttps://docs.openstack.org/ironic/latest/install/configure-nova-flavors.html#scheduling-based-on-resource-classes and adjust your flavors (tripleo uses resource class 'baremetal') | 09:12 |
| Tengu | I suspect that label creates something weird somewhere. First I thought is was due to the fact the server would boot in UEFI mode instead of BIOS (default when no disk_label is set), but it doesn't seem to be the case: the servers are in compatibility mode | 09:12 |
| dtantsur | mmm https://docs.openstack.org/ironic/latest/install/configure-nova-flavors.html#scheduling-based-on-resource-classes | 09:12 |
| dtantsur | Tengu: what exactly did you do with disk_label? | 09:13 |
| Tengu | dtantsur: set it to "gpt", according to https://docs.openstack.org/ironic/pike/install/advanced.html#choosing-the-disk-label | 09:13 |
| dtantsur | so, everything was working, then you set this, and the next deployment attempt failed, right? | 09:14 |
| Tengu | dtantsur: yup. altough I wasn't there when my colleagues tried to deploy the overcloud. But according to the git and file status we have, they actually didn't do anything else. | 09:15 |
| dtantsur | Tengu: I just wonder if you had several deployment attempts failing for another reason, and they you got hit by https://github.com/openstack/instack-undercloud/commit/b366467f6d596017913167cf3e6c158805c8ae52 | 09:16 |
| Tengu | dtantsur: re: resource_class: if I understand well, I should do nova flavor-key c2c_baremetal set resources:CUSTOM_C2C_BAREMETAL=1 right ? | 09:16 |
| dtantsur | CUSTOM_BAREMETAL (because tripleo uses "baremetal" as its resource class, it does not have to match the flavor name) | 09:16 |
| Nisha_Agarwal | dtantsur, i raised the spec https://review.openstack.org/500429 for resource class topic. Have a look and pour in ur opinion on the spec. Thank you. :) | 09:17 |
| patchbot | patch 500429 - ironic-specs - Add resource class structure to ironic | 09:17 |
| dtantsur | Tengu: anyway, I'd like to see nova-compute logs, is it possible to attach them to the bug? | 09:17 |
| dtantsur | Nisha_Agarwal: will put it on my list | 09:17 |
| * Nisha_Agarwal will be away for half an hour. will join back | 09:18 | |
| Nisha_Agarwal | dtantsur, thanks | 09:18 |
| Nisha_Agarwal | dtantsur, re inspector issue ankit(ankkumar) will raise the bug with details. | 09:19 |
| Tengu | dtantsur: hmm, I can attach the log, yes. 22. | 09:20 |
| Tengu | *2s | 09:20 |
| dtantsur | Tengu: I'll have to run for a couple of hours, maybe other folks (sambetts?) can help you too | 09:20 |
| *** dtantsur is now known as dtantsur|bbl | 09:21 | |
| Tengu | dtantsur|bbl: fine for me, I'll try to dig a bit deeper in all of that. happy running :) | 09:21 |
| *** Nisha_Agarwal has quit IRC | 09:22 | |
| *** amarao has joined #openstack-ironic | 09:28 | |
| *** milan has joined #openstack-ironic | 09:38 | |
| milan | morning Ironic! :) | 09:39 |
| *** baoli has joined #openstack-ironic | 09:48 | |
| milan | aarefiev, morning, would you mind https://review.openstack.org/#/c/487909/11 ;) I've added the semaphore release (and luckily enough CI gave it the green light ;) | 09:50 |
| patchbot | patch 487909 - ironic-inspector - Allow periodics to terminate inspector | 09:50 |
| *** Nisha_Agarwal has joined #openstack-ironic | 09:51 | |
| *** lucasagomes is now known as lucas-brb | 09:52 | |
| *** gcb has quit IRC | 09:52 | |
| aarefiev | milan: morning :) I'm looking | 09:52 |
| *** Nisha_ has joined #openstack-ironic | 09:56 | |
| aarefiev | done | 09:56 |
| *** baoli has quit IRC | 09:57 | |
| *** rbartal has quit IRC | 09:57 | |
| *** Nisha_Agarwal has quit IRC | 09:58 | |
| milan | aarefiev, thanks a lot! :D | 10:00 |
| * milan needs to bother sambetts|afk and dtantsur|bbl with the same ;) | 10:00 | |
| vdrok | good morning all! | 10:02 |
| milan | morning vdrok! :) | 10:03 |
| vdrok | good morning milan :) | 10:04 |
| *** Nisha_ has quit IRC | 10:14 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic master: Update documentation for ``ilo`` hardware type https://review.openstack.org/496480 | 10:15 |
| *** zhangfei has quit IRC | 10:17 | |
| *** milan has quit IRC | 10:20 | |
| aarefiev | milan, could you check https://review.openstack.org/#/c/495751/2/doc/source/user/http-api.rst | 10:39 |
| patchbot | patch 495751 - ironic-inspector - Docs: fix typo in HTTP API | 10:39 |
| aarefiev | please | 10:39 |
| *** adisky__ has quit IRC | 10:39 | |
| *** yangyapeng has quit IRC | 10:44 | |
| *** yangyapeng has joined #openstack-ironic | 10:45 | |
| *** udesale has quit IRC | 10:46 | |
| *** lucas-brb is now known as lucasagomes | 10:48 | |
| *** yangyapeng has quit IRC | 10:50 | |
| openstackgerrit | Merged openstack/ironic-inspector master: Updated from global requirements https://review.openstack.org/499999 | 10:55 |
| *** rbartal has joined #openstack-ironic | 11:01 | |
| *** litao__ has quit IRC | 11:04 | |
| *** links has quit IRC | 11:07 | |
| openstackgerrit | Ilya Etingof proposed openstack/ironic-python-agent master: Report /dev/disk/by-path on inspection https://review.openstack.org/498489 | 11:14 |
| *** links has joined #openstack-ironic | 11:20 | |
| *** jlabarre has joined #openstack-ironic | 11:23 | |
| *** aparnav has joined #openstack-ironic | 11:25 | |
| *** fellypefca has joined #openstack-ironic | 11:38 | |
| *** stendulker_ has quit IRC | 11:40 | |
| *** links has quit IRC | 11:42 | |
| *** lucasagomes is now known as lucas-hungry | 11:43 | |
| *** yufei has joined #openstack-ironic | 11:44 | |
| *** yufei has left #openstack-ironic | 11:44 | |
| *** hshiina has quit IRC | 11:46 | |
| fellypefca | Good morning everyone. | 11:48 |
| *** rbartal has quit IRC | 11:48 | |
| *** jlabarre has quit IRC | 11:49 | |
| *** jlabarre has joined #openstack-ironic | 11:49 | |
| *** jkilpatr has joined #openstack-ironic | 11:51 | |
| *** dtantsur|bbl is now known as dtantsur | 11:52 | |
| *** baoli has joined #openstack-ironic | 11:54 | |
| *** links has joined #openstack-ironic | 11:56 | |
| *** jlabarre has quit IRC | 11:56 | |
| *** baoli has quit IRC | 11:59 | |
| *** jlabarre has joined #openstack-ironic | 12:00 | |
| dtantsur | TheJulia: morning! do we need a backport of https://review.openstack.org/#/c/499926/ ? | 12:02 |
| patchbot | patch 499926 - ironic - Boot from volume fails with 'iscsi' deploy interface (MERGED) | 12:02 |
| Tengu | hello there! interesting issue: openstack hypervisor show <UID> shows the node with "status disabled" and "state up" - any way to know why that's the case? ironic apparently does see them as "available", without any Maintenance… | 12:04 |
| openstackgerrit | Ilya Etingof proposed openstack/ironic-lib master: Added `by_path` root device hint https://review.openstack.org/500524 | 12:05 |
| openstackgerrit | Ilya Etingof proposed openstack/ironic-python-agent master: Report /dev/disk/by-path on inspection https://review.openstack.org/498489 | 12:07 |
| dtantsur | Tengu: it seems indeed to be a case of https://bugs.launchpad.net/tripleo/+bug/1712632 | 12:07 |
| openstack | Launchpad bug 1712632 in tripleo "Check that enough nodes are in "available" state with maintenance mode off error" [High,Fix released] - Assigned to Alex Schultz (alex-schultz) | 12:07 |
| dtantsur | try getting that patch or applying to nova configuration manually | 12:08 |
| dtantsur | see #3 at https://docs.openstack.org/ironic/latest/install/configure-compute.html | 12:08 |
| Tengu | darn. | 12:08 |
| Tengu | dtantsur: in fact, if I do an "openstack undercloud upgrade", it should get that patch, shouldn't it? | 12:09 |
| *** zshi_ has quit IRC | 12:10 | |
| hjensas | bfournie: dtantsur: know if anyone made an effort to package openstack/networking-baremetal in RDO yet? (If not, I will give it a go... :) ) | 12:12 |
| dtantsur | Tengu: I think so | 12:13 |
| dtantsur | hjensas: it was on bfournie's radar, but I don't think it was started | 12:13 |
| dtantsur | hjensas: feel free to add me as a co-maintainer or ask question, I have some experience with packaging | 12:14 |
| Tengu | dtantsur: apparently not. I juste edited the file and am running a new `openstack undercloud upgrade' in order to ensure it's applied. | 12:14 |
| dtantsur | the patch has just recently merged, maybe it has not hit the repos even | 12:14 |
| * dtantsur -> coffee | 12:14 | |
| Tengu | dtantsur: apparently yep. so now I wait for the upgrade to be done, and cross my fingers. | 12:14 |
| hjensas | dtantsur: ok, thanks I did some rpm packaging in the past but no python apps so guidance is appreciated. :) I will will sync with bfournie first in case he did start something already. | 12:17 |
| openstackgerrit | Ilya Etingof proposed openstack/ironic-python-agent master: Report /dev/disk/by-path on inspection https://review.openstack.org/498489 | 12:17 |
| *** aparnav has quit IRC | 12:23 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-ironicclient master: Updated from global requirements https://review.openstack.org/500023 | 12:26 |
| *** dr_gogeta86 has quit IRC | 12:26 | |
| *** lucas-hungry is now known as lucasagomes | 12:27 | |
| *** udesale has joined #openstack-ironic | 12:28 | |
| openstackgerrit | Fellype Cavalcante proposed openstack/ironic master: Introduce hpOneView and ilorest to OneView https://review.openstack.org/468129 | 12:31 |
| *** jaypipes has joined #openstack-ironic | 12:33 | |
| ankkumar | hii dtantsur | 12:34 |
| ankkumar | dtantsur, i raised bug in ironic-inspector.. https://bugs.launchpad.net/ironic-inspector/+bug/1714944 | 12:35 |
| openstack | Launchpad bug 1714944 in Ironic Inspector "BMC address is getting replaced by 0.0.0.0 in the ramdisk data received" [Undecided,New] | 12:35 |
| *** yangyapeng has joined #openstack-ironic | 12:37 | |
| dtantsur | thanks ankkumar | 12:38 |
| dtantsur | lemme fix the inspector part first, it may unblock you | 12:38 |
| bfournie | hi hjensas, dtantsur: no, I haven't started any of the packaging for openstack/networking-baremetal | 12:39 |
| dtantsur | hjensas: start with reading https://www.rdoproject.org/documentation/add-packages/ | 12:39 |
| dtantsur | hjensas: you may want to cargo-cult another networking-* packaging | 12:40 |
| *** mvpnitesh has quit IRC | 12:40 | |
| *** dr_gogeta86 has joined #openstack-ironic | 12:41 | |
| *** dr_gogeta86 has quit IRC | 12:41 | |
| *** dr_gogeta86 has joined #openstack-ironic | 12:41 | |
| ankkumar | dtantsur, thanks | 12:41 |
| hjensas | dtantsur: thanks, will do. | 12:42 |
| *** yangyapeng has quit IRC | 12:42 | |
| *** yangyapeng has joined #openstack-ironic | 12:42 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector master: Do not rely on the older ipmi_address field on lookup https://review.openstack.org/500534 | 12:44 |
| dtantsur | aarefiev: ^^^ | 12:44 |
| *** ankkumar has quit IRC | 12:45 | |
| dtantsur | "Due to Harvey Hurricane at Houston the HPE third party CI setup is down" wow, scary | 12:47 |
| dtantsur | I hope everyone is OK | 12:47 |
| aarefiev | dtantsur: I still don't understand why look up fails if `mac` attribute is match | 12:50 |
| dtantsur | aarefiev: becase 0.0.0.0 does not | 12:51 |
| dtantsur | IIRC we need all present fields to match | 12:52 |
| aarefiev | dtantsur: hmm, I think we count nodes with matching attribues | 12:53 |
| dtantsur | yes, but nodes with IPMI address a.b.c.d do not match data with IPMI address 0.0.0.0 (before my patch) | 12:53 |
| * aarefiev checks | 12:53 | |
| Tengu | dtantsur: humpf… after the upgrade of the VM, I still get a "disabled" status… I re-imported all the nodes in order to ensure I'm in a clean state :( | 12:56 |
| aarefiev | dtantsur: yes, IPMI address doesn't match, but next `macs` attribute should match anyway https://github.com/openstack/ironic-inspector/blob/master/ironic_inspector/node_cache.py#L797 | 12:57 |
| aarefiev | hmm, maybe I'm missing something | 12:58 |
| *** gcb has joined #openstack-ironic | 13:00 | |
| *** jaganathan has quit IRC | 13:01 | |
| dtantsur | Tengu: I think now you need to enable the compute, I'm not sure how, some nova command | 13:01 |
| dtantsur | aarefiev: think of it as an AND operation | 13:01 |
| Tengu | dtantsur: hmm. ok | 13:01 |
| dtantsur | aarefiev: in pseude-code (data.bmc_address is None or data.bmc_address == node.bmc_address) and (data.macs is None or data.macs in node.macs) | 13:02 |
| Tengu | dtantsur: even when I deleted all nodes and re-imported them properly? | 13:02 |
| dtantsur | Tengu: this is in Nova, not in Ironic. | 13:02 |
| dtantsur | tl;dr Nova can disable a compute instance, if it failed to build several times in a row | 13:02 |
| dtantsur | aarefiev: actually you may be right and we use OR there. lemme check again | 13:03 |
| Tengu | dtantsur: darn -.- | 13:04 |
| aarefiev | dtantsur: so we are counting the number of matching attributes, and choose the node with max number | 13:06 |
| dtantsur | Tengu: this may help https://docs.openstack.org/nova/latest/admin/services.html | 13:10 |
| Tengu | duh. thanks ! | 13:10 |
| Tengu | (sorry, on the phone, my turn to run in all directions -.-' | 13:10 |
| Tengu | have to wait a bit more in order to ensure all is really up-to-date - just saw the trunk source-list points to a specific hash, and thus I juste updated the yum config now -.-.' | 13:13 |
| Tengu | maybe that will help as well. | 13:13 |
| *** baoli has joined #openstack-ironic | 13:14 | |
| *** rbudden has joined #openstack-ironic | 13:18 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic master: Troubleshooting docs: explain disabled compute services https://review.openstack.org/500542 | 13:20 |
| dtantsur | Tengu: troubleshooting guide update with your problem ^^ | 13:20 |
| dtantsur | vdrok, sambetts|afk ^^ | 13:20 |
| dtantsur | aarefiev: ok, I must admin I don't understand the problem.. | 13:21 |
| Tengu | dtantsur: cool, thanks! I'm just waiting for the deploy to finish before I can say "yep, duplicated issue". | 13:22 |
| Tengu | dtantsur: and I'll push the troubleshooting doc link in our wiki so that we can access it in case of emergency. | 13:22 |
| aarefiev | dtantsur: yeah, it's strange, I've just modified unit test to match bug | 13:27 |
| aarefiev | and it works fine | 13:27 |
| openstackgerrit | Fellype Cavalcante proposed openstack/ironic master: Introduce hpOneView and ilorest to OneView https://review.openstack.org/468129 | 13:29 |
| aarefiev | and it means that discovered mac address didn't match with ironic one | 13:29 |
| aarefiev | or there is bug in attributes adding logic | 13:30 |
| aarefiev | which is unlikely imo | 13:31 |
| TheJulia | good morning everyone | 13:32 |
| aarefiev | dtantsur: btw ipmi address matches bmc https://github.com/openstack/ironic-python-agent/blob/master/ironic_python_agent/inspector.py#L244 | 13:33 |
| aarefiev | morning TheJulia | 13:33 |
| dtantsur | aarefiev: yep, but we forgot to get rid of 0.0.0.0 there (see my patch) | 13:33 |
| dtantsur | morning TheJulia | 13:33 |
| Tengu | dtantsur: humpf. "Failed to set service status to enabled" - need more digging, I'll go check with either #tripleo or, if it exists, #openstack-nova | 13:34 |
| dtantsur | Tengu: yep, the latter channel exists. please let me know how it goes, I need it for this troubleshooting guide | 13:35 |
| TheJulia | dtantsur: re bacport, I believe xo | 13:35 |
| TheJulia | so | 13:35 |
| * TheJulia needs coffee | 13:35 | |
| dtantsur | k, lemme propose | 13:35 |
| TheJulia | dtantsur: thanks | 13:36 |
| Tengu | dtantsur: found out : the example command was misleading. | 13:36 |
| dtantsur | TheJulia: two patches with the same summary, isn't it cool? https://review.openstack.org/#/c/499446/ and https://review.openstack.org/#/c/499926/ | 13:36 |
| patchbot | patch 499446 - ironic - Boot from volume fails with 'iscsi' deploy interface (MERGED) | 13:36 |
| patchbot | patch 499926 - ironic - Boot from volume fails with 'iscsi' deploy interface (MERGED) | 13:36 |
| Tengu | dtantsur: a --help showed me it wanted a hostname, not a zone, as argument. now my nova is up and enabled :). | 13:36 |
| Tengu | time for a deploy! | 13:36 |
| dtantsur | \o/ | 13:36 |
| Tengu | have to ensure nodes are actually detected as "enabled" now. | 13:37 |
| Tengu | YES | 13:37 |
| TheJulia | dtantsur: now I REALLY need coffeee | 13:37 |
| dtantsur | TheJulia: yep :) | 13:37 |
| dtantsur | Tengu: nice! was the mistake in nova docs or in my patch? | 13:37 |
| * Tengu starts a deploy.sh after almost a day-long of headaches | 13:37 | |
| TheJulia | dtantsur: could be worse | 13:37 |
| Tengu | dtantsur: not in your update, in another page you pointed: https://docs.openstack.org/nova/latest/admin/services.html | 13:38 |
| dtantsur | aha | 13:38 |
| * TheJulia sets up the espresso machine and hopes that the kitten decides to leave her alone by the time coffee is ready | 13:38 | |
| Tengu | dtantsur: the 4th point is wrong, it states "nova" instead of "compute". | 13:38 |
| dtantsur | kittens have even higher priority than coffee, sorry | 13:38 |
| dtantsur | Tengu: yeah, I was confused by it too, and also consulted --help | 13:38 |
| Tengu | :) | 13:38 |
| Tengu | might be good to update the said page as well. | 13:38 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic stable/pike: Boot from volume fails with 'iscsi' deploy interface https://review.openstack.org/500545 | 13:39 |
| dtantsur | Tengu: if you have a minute, please propose an update | 13:39 |
| Tengu | dtantsur: have to create an account and the rest in fact. | 13:40 |
| dtantsur | Tengu: oh, so you've never contributed to openstack before? it's not too hard, but I can do it, if you're out of time | 13:40 |
| dtantsur | "do it" = "submit an update" | 13:40 |
| TheJulia | dtantsur: but what if least cost multipath is in use? | 13:40 |
| dtantsur | Oo | 13:41 |
| dtantsur | oO | 13:41 |
| * dtantsur makes a shocked owl after hearing "multipath" | 13:41 | |
| * TheJulia wonders if she should giggle, or laugh manically | 13:42 | |
| Tengu | dtantsur: grumblfdfasd. another issue, but that time it's not related to ironic (like, not at all), and probably not to nova either - more something to do with allocated resources to the undercloud: Timed out waiting for messages from Execution (ID: f6aa4256-8f31-4ad7-ac66-01b6a9767fa6, State: RUNNING). The WebSocket timed out before the Workflow completed. - anyway, it passed the "I can't find | 13:42 |
| Tengu | hosts" part. | 13:43 |
| Tengu | dtantsur: thanks a lot :). | 13:43 |
| dtantsur | awesome! you're welcome :) the WebSocker (Zaqar?) problem is better solved on #tripleo | 13:43 |
| dtantsur | Tengu: do you want to try updating https://docs.openstack.org/nova/latest/admin/services.html or should I? | 13:43 |
| Tengu | dtantsur: I'm a bit out of time indeed - I'll check if I can create an account during the week, I'm pretty sure I'll be able to provide some help regarding the doc as I stumble on weird issues :). | 13:44 |
| Tengu | dtantsur: if you have an editor on the page, best to do it asap I think. I'll find other issues in the doc ;) | 13:44 |
| Tengu | I'm pretty sure people will be happy to get a clear doc before I find the time to create my gerrit account :] | 13:44 |
| dtantsur | cool, will do | 13:46 |
| *** hjensas has quit IRC | 13:46 | |
| Tengu | thanks! | 13:46 |
| *** jlabarre has quit IRC | 13:47 | |
| *** jlabarre has joined #openstack-ironic | 13:49 | |
| dtantsur | Tengu: looks good? https://review.openstack.org/#/c/500551/ | 13:50 |
| patchbot | patch 500551 - nova - Correct examples in "Manage Compute services" docu... | 13:50 |
| TheJulia | ahh coffeeeee | 13:51 |
| dtantsur | :) | 13:51 |
| Tengu | dtantsur: ah, good thing for the --disable-reason part. fine for me | 13:52 |
| dtantsur | good | 13:52 |
| Tengu | dtantsur: and the websocket issue was indeed due to the poor performances of the undercloud VM, running on a poor old server because "well, it's not that important" :). | 13:53 |
| Tengu | but right now: seems to be deploying as expected, once I waited for the load to lower after the nova-compute enable-thing | 13:53 |
| vdrok | morning TheJulia and Tengu | 13:55 |
| Tengu | morning vdrok | 13:55 |
| vdrok | dtantsur: hrm, is there some config option (with default of 10?) to control that service should be down? | 13:55 |
| vdrok | did not know about that | 13:55 |
| Tengu | ah. so now I can re-try the GPT thing for my 2.2To. | 13:55 |
| *** jlabarre has quit IRC | 13:57 | |
| * etingof is wondering how would one get the CI pulling specific patch from a dependency repo? Depends-On does not seem to help | 13:57 | |
| dtantsur | vdrok: there is an option, see #3 at https://docs.openstack.org/ironic/latest/install/configure-compute.html | 13:57 |
| dtantsur | etingof: it should work in the most of cases EXCEPT when you try depending on a library | 13:58 |
| vdrok | dtantsur: ah, thanks! | 13:58 |
| dtantsur | etingof: we consume libraries (ironic-lib, oslo, etc) from pypi | 13:58 |
| etingof | it is indeed the ironic-lib ;) | 13:58 |
| dtantsur | well, then the answer is "you can't", unfortunarely :( | 13:58 |
| *** jlabarre has joined #openstack-ironic | 13:59 | |
| * etingof goes to comment out the dependent tests | 13:59 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic stable/pike: Fix indentation in few of the documentation pages https://review.openstack.org/500557 | 14:02 |
| openstackgerrit | Ilya Etingof proposed openstack/ironic-python-agent master: Report /dev/disk/by-path on inspection https://review.openstack.org/498489 | 14:04 |
| *** udesale has quit IRC | 14:05 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic stable/pike: Update Nova configuration documentation https://review.openstack.org/500562 | 14:05 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic stable/pike: Minor improvements to the resource classes documentation https://review.openstack.org/500563 | 14:05 |
| dtantsur | moar cherry-picks ^^^ | 14:05 |
| TheJulia | yay moar | 14:05 |
| *** udesale has joined #openstack-ironic | 14:07 | |
| Tengu | dtantsur: now another question: if I want to enable GPT support on the overcloud nodes, where should I set the disk_label capability? on the flavor, on the node at import time, or… ? | 14:21 |
| Tengu | I'd like to do that in the right way. | 14:21 |
| dtantsur | Tengu: I don't remember from the top of my head, but I suspect on both flavors and in nodes | 14:22 |
| Tengu | dtantsur: ok, like in the doc then. One said me "ugh, strange". | 14:22 |
| dtantsur | Tengu: https://docs.openstack.org/ironic/latest/install/advanced.html#when-used-with-compute-service | 14:22 |
| dtantsur | yeah | 14:23 |
| Tengu | just wanted to ensure the doc is correct on that. | 14:23 |
| dtantsur | well, this is how capabilities work for ironic currently | 14:23 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic stable/pike: Boot from volume fails with 'iscsi' deploy interface https://review.openstack.org/500566 | 14:23 |
| Tengu | :) | 14:23 |
| Tengu | I was really surprised to have to enable that feature. | 14:23 |
| Tengu | as the disk space is reported, I was hoping it was automagically detected, like "duh, that one has over 2To, let's activate gpt label instead of dos". I was wrong, obviously ,) | 14:24 |
| dtantsur | it's not an unreasonable thing to expect; I don't remember why exactly we did not do it | 14:24 |
| Tengu | time? ;) | 14:25 |
| Tengu | might be uneasy, especially with the potential issue bios vs uefi | 14:25 |
| dtantsur | yep :) | 14:25 |
| *** gcb has quit IRC | 14:29 | |
| *** mrch has quit IRC | 14:33 | |
| *** amotoki is now known as amotoki__away | 14:34 | |
| *** udesale has quit IRC | 14:35 | |
| *** ccamacho has quit IRC | 14:36 | |
| TheJulia | dtantsur: I think we were expecting people to just generally use whole disk images | 14:36 |
| * dtantsur le sigh | 14:36 | |
| dtantsur | well, tripleo got early whole disk image support in IIRC pike :) | 14:37 |
| *** ccamacho has joined #openstack-ironic | 14:37 | |
| * TheJulia blinks | 14:37 | |
| * TheJulia ponders making more coffee, and going back to hacking on shade | 14:37 | |
| mordred | I didn't do it | 14:37 |
| dtantsur | this is how one summons mordred :D | 14:38 |
| TheJulia | lol | 14:38 |
| *** links has quit IRC | 14:40 | |
| mordred | dtantsur: I may have an irc ping configured :) | 14:42 |
| Tengu | for "coffee" ? :) | 14:43 |
| TheJulia | Tengu: Most likely that as well | 14:43 |
| Tengu | XD | 14:43 |
| * TheJulia checks the weather to see if the denver area will be as stupidly hot as it was yesterday | 14:44 | |
| * Tengu is happy the weather has changed drastically in switzerland | 14:45 | |
| Tengu | 30°C -> ~20°C - a blessing | 14:45 |
| dtantsur | Tengu: ditto here (CZ) | 14:45 |
| Tengu | uhu | 14:45 |
| TheJulia | dtantsur: so, I didn't explicitly find tons of options for burgers and beer, but I did checkout http://www.caseysbistroandpub.com which is with-in walking distance. Food seems good, bar staff says Monday's are hit or miss for large groups, but there is also a sushi place next door as a post beverage alternate location | 14:47 |
| dtantsur | TheJulia: no reservations for them? | 14:48 |
| TheJulia | it didn't seem like they did them. I'll call them later today and see if they can do them, but in realistically I think descending upon the bar would be best and then kind of organically going from there. The space didn't really lend it to us blocking off a ton of tables | 14:51 |
| dtantsur | okay, I see | 14:51 |
| *** ianychoi_ is now known as ianychoi | 14:52 | |
| dtantsur | TheJulia: thanks for checking this. could you please respond to the appropriate thread with your findings/suggestions? | 14:52 |
| TheJulia | that means I need to open my email :( | 14:52 |
| dtantsur | that's dangerous indeed | 14:53 |
| *** udesale has joined #openstack-ironic | 14:54 | |
| TheJulia | ugh, email | 14:55 |
| *** baoli has quit IRC | 14:57 | |
| TheJulia | dtantsur: appropriate thread? I'm not seeing one :\ | 14:58 |
| dtantsur | TheJulia: http://lists.openstack.org/pipermail/openstack-dev/2017-August/121565.html | 14:58 |
| *** udesale has quit IRC | 14:58 | |
| TheJulia | there we go | 14:59 |
| TheJulia | 'dinner' was the missing key | 14:59 |
| *** aarefiev is now known as aarefiev_afk | 14:59 | |
| *** mjura has quit IRC | 15:06 | |
| *** rpioso has joined #openstack-ironic | 15:29 | |
| *** ccamacho has quit IRC | 15:38 | |
| *** chas has quit IRC | 15:39 | |
| *** chas has joined #openstack-ironic | 15:39 | |
| *** e0ne has quit IRC | 15:41 | |
| *** chas has quit IRC | 15:44 | |
| *** marios has quit IRC | 15:51 | |
| *** jroll has quit IRC | 15:52 | |
| *** jroll has joined #openstack-ironic | 15:53 | |
| TheJulia | wow today is quiet | 15:54 |
| dtantsur | most of the US folks are out.. | 15:54 |
| TheJulia | I'm just here because I'm crazy | 15:54 |
| TheJulia | and I <3 you all | 15:54 |
| dtantsur | =^_^= | 15:55 |
| * TheJulia is unsure if that is good or bad | 15:55 | |
| *** lucasagomes is now known as lucas-afk | 15:58 | |
| *** srart has quit IRC | 15:59 | |
| *** dtantsur is now known as dtantsur|bbl | 16:01 | |
| *** yufei has joined #openstack-ironic | 16:03 | |
| *** yufei has left #openstack-ironic | 16:03 | |
| openstackgerrit | Merged openstack/ironic stable/pike: Boot from volume fails with 'iscsi' deploy interface https://review.openstack.org/499911 | 16:08 |
| *** yangyapeng has quit IRC | 16:09 | |
| *** yangyapeng has joined #openstack-ironic | 16:09 | |
| *** yangyapeng has quit IRC | 16:13 | |
| *** etingof has quit IRC | 16:16 | |
| vdrok | dtantsur|bbl: TheJulia so ironic CLI is already deprecated or it is not yet? | 16:18 |
| *** Goneri has joined #openstack-ironic | 16:18 | |
| * vdrok was not paying attention lately :( | 16:18 | |
| vdrok | wondering do I still need to list commands for both ironic and openstack CLI in a spec | 16:19 |
| *** rpioso has quit IRC | 16:21 | |
| TheJulia | vdrok: deprecated, although I don't think we would reject attempts to maintain parity in the short term. | 16:28 |
| * TheJulia suspects there needs to be more clouds.yaml examples out there to help with osc adoption | 16:28 | |
| vdrok | well, I don't really want to do the feature parity for this spec :) | 16:29 |
| * TheJulia shrugs | 16:29 | |
| vdrok | will leave it out for now, can be added later | 16:29 |
| TheJulia | vdrok: I don't think we can ever really change the template to remove the section short of stripping out the section in all of the existing specs | 16:29 |
| *** etingof has joined #openstack-ironic | 16:31 | |
| *** kaifeng has joined #openstack-ironic | 16:39 | |
| *** stendulker has joined #openstack-ironic | 16:42 | |
| *** tesseract has quit IRC | 16:48 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs master: Owner-aware baremetal nodes https://review.openstack.org/415512 | 16:51 |
| TheJulia | ohhhh | 16:52 |
| *** dtantsur|bbl is now known as dtantsur | 16:54 | |
| dtantsur | vdrok: it's not officially deprecated yet, but we're on the way to it | 16:55 |
| *** milan has joined #openstack-ironic | 16:55 | |
| openstackgerrit | Ilya Etingof proposed openstack/ironic-specs master: IPA to report node activity https://review.openstack.org/492702 | 16:55 |
| vdrok | a huge update to https://review.openstack.org/415512 | 16:57 |
| patchbot | patch 415512 - ironic-specs - Owner-aware baremetal nodes | 16:57 |
| *** derekh has quit IRC | 16:57 | |
| vdrok | dtantsur: meeting? :) | 17:00 |
| *** rpioso has joined #openstack-ironic | 17:01 | |
| *** stendulker_ has joined #openstack-ironic | 17:01 | |
| *** Goneri has quit IRC | 17:02 | |
| *** stendulker has quit IRC | 17:02 | |
| *** mgoddard has quit IRC | 17:05 | |
| *** stendulker_ has quit IRC | 17:07 | |
| *** stendulker_ has joined #openstack-ironic | 17:07 | |
| *** ricardoas has joined #openstack-ironic | 17:07 | |
| *** ralonsoh has quit IRC | 17:08 | |
| *** jkilpatr has quit IRC | 17:17 | |
| *** chas has joined #openstack-ironic | 17:40 | |
| *** chas has quit IRC | 17:45 | |
| *** stendulker_ has quit IRC | 17:49 | |
| *** kaifeng has quit IRC | 17:49 | |
| TheJulia | So i missed one thing... something aside from zombie related tv shows to play in the background would be totally awesome..... | 17:54 |
| TheJulia | in addition to code and coffee | 17:54 |
| dtantsur | hah | 17:54 |
| rpioso | I have a question about ManagementInterface.set_boot_device(). Must the implementation of that method configure the hardware to be consistent with what the caller's requested? Or can that be deferred until the next reboot of the node? | 17:56 |
| * dtantsur has to go now | 17:57 | |
| dtantsur | see you tomorrow | 17:57 |
| rpioso | dtantsur: o/ | 17:57 |
| milan | o/ | 17:57 |
| *** rcernin has joined #openstack-ironic | 17:57 | |
| *** dtantsur is now known as dtantsur|afk | 17:57 | |
| TheJulia | rpioso: could you give us some context? | 17:58 |
| *** mat128 has joined #openstack-ironic | 17:59 | |
| TheJulia | rpioso: as in, what raises the question in your mind, that way we can understand and give you the best possible thoughts | 17:59 |
| TheJulia | greetings mat128 | 17:59 |
| mat128 | hey! been a while | 18:00 |
| TheJulia | that it has! How has life been treating you? | 18:00 |
| rpioso | BMCs, such as the iDRAC, can require a good deal of time, ~5 minutes, to affect the boot device setting on the server. | 18:00 |
| rpioso | TheJulia: ^^^ | 18:01 |
| rpioso | TheJulia: That method is exposed by the ironic API. | 18:01 |
| TheJulia | rpioso: is it certain that upon the next power cycle, as long as the bmc has completed that task, that the machine will boot? | 18:02 |
| TheJulia | boot using that device that is | 18:02 |
| mat128 | TheJulia: not too bad actually, still overloaded by work | 18:02 |
| mat128 | TheJulia: and you? | 18:02 |
| rpioso | TheJulia: Setting it on the server synchronously within ironic Conductor and the driver would make the API hang. | 18:02 |
| TheJulia | mat128: It goes alright, just trying to keep remaining sanity | 18:02 |
| TheJulia | rpioso: yeah, and at the same time I feel like we offer the perception that it is done when the method returns | 18:03 |
| rpioso | TheJulia: That's the issue. | 18:03 |
| TheJulia | rpioso: perhaps with idrac power management, it could have some built-in wait, but on sync calls it would still hang, unless it errored until set | 18:03 |
| TheJulia | actually, that is not a tooooo horrible way to handle it.... although that raises the error handling to the end user, which is kind of equally bad | 18:04 |
| *** milan has quit IRC | 18:04 | |
| TheJulia | rpioso: is there any difference between "default" and "next boot"? | 18:04 |
| TheJulia | next boot being a single boot in that thought | 18:05 |
| *** jlabarre has quit IRC | 18:09 | |
| rpioso | TheJulia: There's no difference in the amount of time required +/- | 18:09 |
| *** jlabarre has joined #openstack-ironic | 18:11 | |
| rpioso | TheJulia: Presently, the h/w config of the boot device is deferred until a power state change is requested. | 18:11 |
| rpioso | TheJulia: The drac's ManagementInterface.set_boot_device() caches the requested boot device setting on the ironic node object and returns success. | 18:13 |
| TheJulia | That kind of seems like the only somewhat reasonable way to handle it :( | 18:13 |
| rpioso | TheJulia: :( | 18:13 |
| rpioso | TheJulia: And what should ManagementInterface.get_boot_device() return? Fetching it from the h/w is "fast". | 18:14 |
| TheJulia | rpioso: does the setting also take several minutes to take effect if one is using ipmi? | 18:14 |
| rpioso | TheJulia: I don't know. | 18:16 |
| *** yangyapeng has joined #openstack-ironic | 18:17 | |
| rpioso | TheJulia: The drac driver uses a WS-Man interface. | 18:17 |
| TheJulia | rpioso: what should be returned, I guess it should be the current device, and doc could just tell the user "hey, this takes a while internally to the bmc" | 18:18 |
| rpioso | "the current device" configured on the ironic node, if any; otherwise, whatever's configured on the h/w? | 18:20 |
| rpioso | TheJulia: ^^^ | 18:20 |
| *** yangyapeng has quit IRC | 18:21 | |
| TheJulia | I was thinking the other way around, but that could work as well, the key is consistency for experience and minimizing impact. | 18:21 |
| TheJulia | The reconfiguration task in the idrac taking that long is.... not great. :( | 18:22 |
| rpioso | TheJulia: I'm looking into adding UEFI support to the drac driver. Hopefully, that's better :) | 18:25 |
| TheJulia | That makes a little more sense! | 18:26 |
| rpioso | TheJulia: What's the purpose of node.instance_info? Is there a doc about that? | 18:26 |
| TheJulia | rpioso: it contains information about the instance that has been requested, such as image to be written | 18:27 |
| TheJulia | explicit doc, I don't think so | 18:27 |
| *** yee379 has joined #openstack-ironic | 18:27 | |
| *** alezil_ has joined #openstack-ironic | 18:29 | |
| *** pas-ha_ has joined #openstack-ironic | 18:30 | |
| *** igordc_ has joined #openstack-ironic | 18:32 | |
| rpioso | TheJulia: What's the dfference between properties.capabilities and instance_info.capabilities? | 18:33 |
| *** leifmadsen_ has joined #openstack-ironic | 18:36 | |
| *** yee37926 has quit IRC | 18:37 | |
| *** fried_rice has quit IRC | 18:37 | |
| *** igordc has quit IRC | 18:37 | |
| *** pas-ha has quit IRC | 18:37 | |
| *** leifmadsen has quit IRC | 18:37 | |
| *** alezil has quit IRC | 18:37 | |
| *** pas-ha_ is now known as pas-ha | 18:37 | |
| *** alezil_ is now known as alezil | 18:37 | |
| TheJulia | rpioso: properties.capabilities are supposed to be scheduled upon, what is available, what defaults may be asserted to the hardware that are operationally the context of the node. Where as instance_info.capabilities are what the node has been scheduled upon. instance_info.capabilities might not actually be asserted by all users since there is the stand-alone use case. Also, a lot of this is up in the air | 18:41 |
| TheJulia | with resource classes and all | 18:41 |
| *** trown has quit IRC | 18:42 | |
| *** trown has joined #openstack-ironic | 18:42 | |
| *** dims has quit IRC | 18:44 | |
| *** pc_m has quit IRC | 18:44 | |
| *** pc_m has joined #openstack-ironic | 18:44 | |
| *** fried_rice has joined #openstack-ironic | 18:45 | |
| *** oomichi has quit IRC | 18:45 | |
| *** oomichi has joined #openstack-ironic | 18:46 | |
| *** mat128 has quit IRC | 18:47 | |
| *** dims has joined #openstack-ironic | 18:47 | |
| rpioso | TheJulia: That helps. Thank you! I need to get up to speed on resource classes :) | 18:48 |
| ricardoas | TheJulia rpioso: oneview management interface only apply the change if the machine is powered off... otherwise is stores (at driver_internal_info iirc) to be applied the next time it reboots | 18:51 |
| TheJulia | ricardoas: good point | 18:52 |
| rpioso | TheJulia: Where do /etc/ironic/ironic.conf's settings fit into properties.capabilities and instance_info.capabilities? | 18:52 |
| ricardoas | but even if the machine is powered off it can take a while... it would be good if at least it hangs on a timeout | 18:52 |
| ricardoas | not sure if it is the best way, though | 18:53 |
| TheJulia | rpioso: like default boot mode? | 18:53 |
| rpioso | TheJulia: Yes, that one :) | 18:55 |
| TheJulia | rpioso: ironic.conf is the overall default, specific settings can be asserted on a node level | 18:55 |
| rpioso | TheJulia: So it's used if properties.capabilities.boot_node is not set? | 18:56 |
| *** baoli has joined #openstack-ironic | 18:59 | |
| rpioso | ricardoas: Thank you for sharing that. I just pulled up the oneview code. Its approach looks the same as drac's. | 19:01 |
| *** baoli has quit IRC | 19:03 | |
| rpioso | ricardoas: Good to know that the drac is not alone :) | 19:07 |
| TheJulia | rpioso: yes, if memory serves. Right now it defaults to netboot, We need to change that. I tossed up a patch to see what would break, just haven't gotten back to it yet | 19:11 |
| rpioso | TheJulia: Are you familiar with the code mechanics to deprecate a hardware type's supported interface? | 19:16 |
| *** fellypefca has quit IRC | 19:20 | |
| rpioso | TheJulia: Is it as simple as overriding the __setattr__ method to look for the property being set to a deprecated value, and, if so determined, emitting a warning-level deprecation log message? | 19:22 |
| rpioso | Has anyone successfully used UEFI boot with agent deploy? | 19:35 |
| *** chas has joined #openstack-ironic | 19:41 | |
| *** chas has quit IRC | 19:45 | |
| *** racedo has quit IRC | 20:02 | |
| *** yangyapeng has joined #openstack-ironic | 20:04 | |
| *** yangyapeng has quit IRC | 20:11 | |
| TheJulia | rpioso: no, I think irmc just put a reno in that they were going to pull an interface in queens. | 20:19 |
| ricardoas | rpioso: :D | 20:19 |
| TheJulia | rpioso: so, I seem to remember we did so at hp, but it took some steps and custom built images | 20:21 |
| *** crushil has quit IRC | 20:27 | |
| *** crushil has joined #openstack-ironic | 20:27 | |
| * TheJulia digs back into the keystoneauth code | 20:32 | |
| *** chas has joined #openstack-ironic | 20:44 | |
| rpioso | TheJulia: I couldn't find the irmc reno :( I'll ask again in channel. | 20:44 |
| *** rbudden has quit IRC | 20:45 | |
| rpioso | TheJulia: Regarding agent deploy's support of UEFI boot, could you recommend someone specific to ask? | 20:45 |
| *** rcernin has quit IRC | 20:45 | |
| rpioso | TheJulia: Thank you so much for your assistance. | 20:46 |
| *** chas has quit IRC | 20:47 | |
| TheJulia | rpioso: agent boto in general or agent deployment of a uefi image? | 20:47 |
| ricardoas | has anybody run opestack ansible aio with ironic for pike lately? I can´t find apache2 at ironic-api container :( | 20:47 |
| TheJulia | ricardoas: I have not, I've never messed with OSA sadly | 20:54 |
| ricardoas | TheJulia: it seems like it was replaced by wsgi stuff but i have no idea of what it is | 20:57 |
| ricardoas | apache2 was replaced* | 20:57 |
| TheJulia | ahh, yeah, that was a community goal to go to wsgi. I'm wondering if their roles are just expecting something that is not present because of the deployment settings. | 20:59 |
| *** jaypipes has quit IRC | 21:01 | |
| ricardoas | TheJulia: got it! ironic api container is missing a lib... | 21:01 |
| * ricardoas is going to dig into deployment settings... | 21:02 | |
| *** yangyapeng has joined #openstack-ironic | 21:15 | |
| *** yangyapeng has quit IRC | 21:20 | |
| *** pcaruana has quit IRC | 21:20 | |
| *** baoli has joined #openstack-ironic | 21:21 | |
| *** baoli has quit IRC | 21:25 | |
| rpioso | TheJulia: The ironic.drivers.agent.AgentDeploy interface's support of UEFI boot mode. | 21:33 |
| *** b4rti has joined #openstack-ironic | 21:37 | |
| TheJulia | rpioso: so agent deploy is for whole disk images, so it is dependent upon the disk image that is requested | 21:37 |
| *** jlabarre has quit IRC | 21:38 | |
| rpioso | TheJulia: Doesn't the server need to be placed into the proper boot mode? | 21:39 |
| *** jlabarre has joined #openstack-ironic | 21:41 | |
| TheJulia | so, realistically it should be the boot interface | 21:43 |
| TheJulia | deploy calls prepare_ramdisk and prepare_instance in the boot interface | 21:44 |
| *** jtomasek has quit IRC | 21:44 | |
| rpioso | TheJulia: Exactly. I'm wondering if anyone has successfully used UEFI boot mode with AgentDeploy. | 21:46 |
| * rpioso is going to eat dinner | 21:47 | |
| *** rpioso is now known as rpioso|eat | 21:48 | |
| *** yangyapeng has joined #openstack-ironic | 22:03 | |
| TheJulia | https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/ilo/boot.py#L331 and prepare_instance in that same file | 22:05 |
| *** rpioso|eat has quit IRC | 22:06 | |
| *** yangyapeng has quit IRC | 22:08 | |
| *** jlabarre has quit IRC | 22:28 | |
| *** jlabarre has joined #openstack-ironic | 22:29 | |
| *** chas has joined #openstack-ironic | 22:47 | |
| *** chas has quit IRC | 22:52 | |
| *** rwsu has joined #openstack-ironic | 22:59 | |
| *** ppiela has joined #openstack-ironic | 23:21 | |
| *** rama_y has joined #openstack-ironic | 23:21 | |
| *** ijw has joined #openstack-ironic | 23:33 | |
| *** karimb has quit IRC | 23:41 | |
| *** ijw has quit IRC | 23:44 | |
| *** ijw has joined #openstack-ironic | 23:51 | |
| *** baoli has joined #openstack-ironic | 23:51 | |
| *** baoli has quit IRC | 23:56 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!