| jlvillal | JayF: Damn Gerrit restart... | 00:00 |
|---|---|---|
| jlvillal | JayF: I tried to do a +2 | 00:00 |
| JayF | lol | 00:00 |
| JayF | I'm trying to do at least 10% of what I intended to do today | 00:00 |
| JayF | my afternoon got sucked into the gate rabbithole | 00:01 |
| JayF | but as long as tests pass, I'll be one patch deep into the rescue changes (that still need a spec review, if you're lookin' something to do :D) | 00:01 |
| JayF | jlvillal: it's a gerrit backup, to be clear :) | 00:01 |
| jlvillal | JayF: Okay :) | 00:01 |
| *** sdake has quit IRC | 00:10 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Add API and conductor methods for [un]rescue https://review.openstack.org/350831 | 00:13 |
| *** baoli has joined #openstack-ironic | 00:13 | |
| *** mtanino has quit IRC | 00:14 | |
| *** baoli has quit IRC | 00:18 | |
| * JayF leaving for the evening o/ | 00:19 | |
| JayF | I'm seeing more gate failures that look unrelated to ironic but are affecting our patches | 00:20 |
| JayF | I notified infra but as I said, am gone for the day | 00:20 |
| JayF | http://logs.openstack.org/81/119981/33/check/gate-ironic-python-agent-python27-ubuntu-xenial/fc6fb6b/console.html#_2016-08-03_22_46_15_817261 is an example of one | 00:20 |
| JayF | o/ for real | 00:21 |
| *** hoangcx has quit IRC | 00:37 | |
| *** hoangcx has joined #openstack-ironic | 00:37 | |
| *** hoangcx has quit IRC | 00:37 | |
| *** hoangcx has joined #openstack-ironic | 00:45 | |
| *** phuongnh has joined #openstack-ironic | 00:58 | |
| *** wajdi has joined #openstack-ironic | 01:08 | |
| *** rama_y has quit IRC | 01:11 | |
| *** Sukhdev has quit IRC | 01:12 | |
| *** yuanying has quit IRC | 01:14 | |
| *** yuanying has joined #openstack-ironic | 01:14 | |
| *** yuanying has quit IRC | 01:15 | |
| openstackgerrit | Merged openstack/ironic: Migrate to using keystoneauth Sessions https://review.openstack.org/236982 | 01:17 |
| openstackgerrit | Merged openstack/ironic: Fix for "db type could not be determined" error message https://review.openstack.org/350302 | 01:17 |
| *** yuanying has joined #openstack-ironic | 01:18 | |
| *** baoli has joined #openstack-ironic | 01:56 | |
| *** jmiu has quit IRC | 01:57 | |
| *** jmiu_ has joined #openstack-ironic | 01:58 | |
| *** wlfightup has joined #openstack-ironic | 01:58 | |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic-inspector: Allow conditions using values from introspection https://review.openstack.org/345535 | 01:59 |
| *** ChrisAusten has joined #openstack-ironic | 02:01 | |
| *** amotoki has joined #openstack-ironic | 02:10 | |
| *** bfournie has joined #openstack-ironic | 02:16 | |
| *** baoli has quit IRC | 02:26 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/bifrost: Updated from global requirements https://review.openstack.org/350875 | 02:33 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/350097 | 02:35 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-inspector: Updated from global requirements https://review.openstack.org/350669 | 02:35 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-lib: Updated from global requirements https://review.openstack.org/350886 | 02:35 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/350887 | 02:35 |
| *** tiendc has joined #openstack-ironic | 02:36 | |
| *** yuanying has quit IRC | 02:51 | |
| *** yuanying has joined #openstack-ironic | 02:54 | |
| *** yuanying has quit IRC | 02:56 | |
| *** sdake has joined #openstack-ironic | 03:04 | |
| *** sdake has quit IRC | 03:23 | |
| *** persia has quit IRC | 03:30 | |
| *** persia has joined #openstack-ironic | 03:33 | |
| *** jrist has joined #openstack-ironic | 03:34 | |
| *** zhangjl has joined #openstack-ironic | 03:36 | |
| *** sdake has joined #openstack-ironic | 03:50 | |
| *** yuanying has joined #openstack-ironic | 03:51 | |
| *** yuanying has quit IRC | 03:51 | |
| *** yuanying has joined #openstack-ironic | 03:52 | |
| *** amotoki has quit IRC | 03:54 | |
| *** amotoki has joined #openstack-ironic | 03:55 | |
| *** Sukhdev has joined #openstack-ironic | 04:03 | |
| *** Sukhdev has quit IRC | 04:04 | |
| *** Sukhdev has joined #openstack-ironic | 04:04 | |
| *** ayoung has quit IRC | 04:08 | |
| openstackgerrit | Debayan Ray proposed openstack/proliantutils: Adds constructor for InvalidInputError https://review.openstack.org/328904 | 04:16 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Baremetal provisioning in UEFI secure boot mode for iLO drivers https://review.openstack.org/231927 | 04:16 |
| openstackgerrit | Hironori Shiina proposed openstack/ironic: Remove duplicate parameters from local.conf example https://review.openstack.org/350970 | 04:17 |
| *** links has joined #openstack-ironic | 04:28 | |
| *** sdake has quit IRC | 04:37 | |
| *** appu103_ has joined #openstack-ironic | 04:42 | |
| *** appu103_ is now known as aparnav | 04:42 | |
| *** wlfightup has quit IRC | 04:58 | |
| *** moshele has joined #openstack-ironic | 05:00 | |
| *** jraju has joined #openstack-ironic | 05:04 | |
| *** moshele has quit IRC | 05:04 | |
| *** hshiina has joined #openstack-ironic | 05:05 | |
| *** links has quit IRC | 05:06 | |
| openstackgerrit | Merged openstack/python-ironicclient: Make shell main() specify return value in exit code https://review.openstack.org/349196 | 05:11 |
| openstackgerrit | zhangyanxian proposed openstack/bifrost: Fix some spelling mistakes https://review.openstack.org/350981 | 05:14 |
| *** wlfightup has joined #openstack-ironic | 05:16 | |
| *** rcernin has joined #openstack-ironic | 05:16 | |
| openstackgerrit | zhangyanxian proposed openstack/bifrost: Fix some spelling mistakes https://review.openstack.org/350981 | 05:18 |
| *** zhangjl1 has joined #openstack-ironic | 05:20 | |
| *** zhangjl has quit IRC | 05:22 | |
| *** bvandewa has quit IRC | 05:22 | |
| *** sabeen has quit IRC | 05:36 | |
| *** tiendc|afk has joined #openstack-ironic | 05:38 | |
| *** tiendc has quit IRC | 05:41 | |
| *** moshele has joined #openstack-ironic | 05:52 | |
| *** ChrisAusten has quit IRC | 05:54 | |
| *** rbartal has joined #openstack-ironic | 05:57 | |
| openstackgerrit | Naohiro Tamura proposed openstack/python-ironicclient: Add soft reboot/poweroff power states. https://review.openstack.org/247904 | 05:58 |
| *** moshele has quit IRC | 06:01 | |
| *** moshele has joined #openstack-ironic | 06:01 | |
| *** tiendc||afk has joined #openstack-ironic | 06:02 | |
| *** rama_y has joined #openstack-ironic | 06:03 | |
| *** tiendc||afk has quit IRC | 06:03 | |
| *** tiendc|afk has quit IRC | 06:04 | |
| *** tiendc||afk has joined #openstack-ironic | 06:04 | |
| *** zhangjl1 has left #openstack-ironic | 06:04 | |
| *** rama_y has quit IRC | 06:06 | |
| *** rama_y has joined #openstack-ironic | 06:06 | |
| *** jtomasek has quit IRC | 06:09 | |
| *** pcaruana has joined #openstack-ironic | 06:10 | |
| *** rama_y has quit IRC | 06:11 | |
| *** rama_y has joined #openstack-ironic | 06:11 | |
| *** tiendc||afk has quit IRC | 06:33 | |
| *** Sukhdev has quit IRC | 06:38 | |
| *** rama_y has quit IRC | 06:52 | |
| *** tesseract- has joined #openstack-ironic | 06:53 | |
| *** hoangcx has quit IRC | 06:54 | |
| *** hoangcx has joined #openstack-ironic | 06:56 | |
| *** zhangjl1 has joined #openstack-ironic | 07:06 | |
| *** ohamada has joined #openstack-ironic | 07:09 | |
| *** bvandewa has joined #openstack-ironic | 07:14 | |
| *** ccamacho has quit IRC | 07:16 | |
| openstackgerrit | Aparna proposed openstack/proliantutils: Modify error message of InvalidInputException in hpssa https://review.openstack.org/338894 | 07:18 |
| *** ccamacho has joined #openstack-ironic | 07:18 | |
| openstackgerrit | Moshe Levi proposed openstack/ironic: Adding InfiniBand Support https://review.openstack.org/264263 | 07:24 |
| *** openstackgerrit has quit IRC | 07:33 | |
| *** openstackgerrit has joined #openstack-ironic | 07:34 | |
| openstackgerrit | Merged openstack/ironic-inspector: Updated from global requirements https://review.openstack.org/350669 | 07:41 |
| openstackgerrit | Aparna proposed openstack/proliantutils: Add hpssa support only in raid mode https://review.openstack.org/347978 | 07:42 |
| *** vishwanathj has quit IRC | 07:44 | |
| *** e0ne has joined #openstack-ironic | 07:55 | |
| *** e0ne has quit IRC | 07:55 | |
| *** e0ne has joined #openstack-ironic | 07:56 | |
| *** milan has joined #openstack-ironic | 07:57 | |
| *** Lord_Phoenix has joined #openstack-ironic | 07:59 | |
| *** zzzeek has quit IRC | 08:00 | |
| *** bvandewa has quit IRC | 08:00 | |
| *** jpich has joined #openstack-ironic | 08:00 | |
| *** zzzeek has joined #openstack-ironic | 08:01 | |
| *** hoangcx has quit IRC | 08:01 | |
| *** hoangcx has joined #openstack-ironic | 08:04 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/ironic: [WIP] Testing latest u-c https://review.openstack.org/318440 | 08:10 |
| openstackgerrit | Naohiro Tamura proposed openstack/python-ironicclient: Add a new command "ironic node-inject-nmi" https://review.openstack.org/247905 | 08:11 |
| *** e0ne has quit IRC | 08:14 | |
| *** e0ne has joined #openstack-ironic | 08:15 | |
| *** Goneri has joined #openstack-ironic | 08:15 | |
| *** dtantsur|afk is now known as dtantsur | 08:16 | |
| dtantsur | Morning Ironic | 08:16 |
| pas-ha | morning dtantsur and other ironicers :) | 08:19 |
| *** dtantsur is now known as dtantsur|bbl | 08:23 | |
| lucas-dinner | morning pas-ha dtantsur|bbl | 08:24 |
| *** lucas-dinner is now known as lucasagomes | 08:24 | |
| openstackgerrit | Merged openstack/ironic-inspector: Combine multiple warning logs into one in create_ports https://review.openstack.org/346578 | 08:24 |
| *** Lord_Phoenix has quit IRC | 08:25 | |
| aarefiev | morning! | 08:25 |
| aarefiev | morning dtantsur pas-ha lucas-dinner | 08:26 |
| alineb | morning dtantsur|bbl pas-ha lucasagomes aarefiev | 08:27 |
| *** Lord_Phoenix has joined #openstack-ironic | 08:27 | |
| alineb | morning all | 08:27 |
| aarefiev | o/ alineb | 08:27 |
| *** Lizhixin1 has joined #openstack-ironic | 08:33 | |
| lucasagomes | pas-ha, the problem with the node object hash still a thing? | 08:34 |
| *** moshele has quit IRC | 08:41 | |
| *** moshele has joined #openstack-ironic | 08:45 | |
| *** jtomasek has joined #openstack-ironic | 08:53 | |
| *** Romanenko_K has joined #openstack-ironic | 08:57 | |
| *** e0ne has quit IRC | 09:03 | |
| *** e0ne has joined #openstack-ironic | 09:04 | |
| pas-ha | looks like, since the patch is not merged yet https://review.openstack.org/#/c/350679/ | 09:07 |
| vsaienk0 | Morning pas-ha, lucasagomes, dtantsur, alineb, aarefiev, vdrok, jroll, sambetts and all Ironic'ers! | 09:13 |
| *** hshiina has quit IRC | 09:14 | |
| *** hshiina has joined #openstack-ironic | 09:14 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Test that network_interface is explicitly set on POST/PATCH https://review.openstack.org/349520 | 09:14 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add a data migration to fill node.network_interface https://review.openstack.org/349528 | 09:15 |
| jroll | morning y'all | 09:15 |
| * jroll reapproves those rebases | 09:15 | |
| lucasagomes | pas-ha, thanks | 09:16 |
| lucasagomes | jroll, vsaienk0, morning | 09:16 |
| pas-ha | morning jroll | 09:16 |
| jroll | devananda: JayF: yes, there are caveats, I plan to keep hacking on it during ocata :) | 09:18 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Follow up to keystoneauth patch https://review.openstack.org/351076 | 09:18 |
| pas-ha | fixed some typos ^ | 09:19 |
| lucasagomes | someone can get some eyes on: https://review.openstack.org/#/c/344164/ quick review, already has a +2 | 09:22 |
| lucasagomes | cheers | 09:23 |
| *** xek_ has joined #openstack-ironic | 09:23 | |
| *** xek has quit IRC | 09:24 | |
| *** Lizhixin1 has left #openstack-ironic | 09:24 | |
| openstackgerrit | Merged openstack/proliantutils: Adds constructor for InvalidInputError https://review.openstack.org/328904 | 09:28 |
| *** stendulker has joined #openstack-ironic | 09:28 | |
| *** electrofelix has joined #openstack-ironic | 09:35 | |
| openstackgerrit | Moshe Levi proposed openstack/ironic-inspector: Adding InfiniBand Support https://review.openstack.org/264257 | 09:36 |
| openstackgerrit | Moshe Levi proposed openstack/ironic-inspector: Dont review https://review.openstack.org/348457 | 09:36 |
| moshele | dtantsur|bbl: not sure I understand all you comments, but please reexamine my patch https://review.openstack.org/264257 | 09:37 |
| openstackgerrit | Aparna proposed openstack/proliantutils: Modify error message of InvalidInputException in hpssa https://review.openstack.org/338894 | 09:39 |
| *** wlfightup has quit IRC | 09:40 | |
| *** wlfightup has joined #openstack-ironic | 09:40 | |
| *** mgould|afk is now known as mgould | 09:41 | |
| mgould | morning Ironic! | 09:41 |
| *** narasimha_SV has joined #openstack-ironic | 09:43 | |
| vsaienk0 | morning mgould! | 09:45 |
| *** zhangjl has joined #openstack-ironic | 09:46 | |
| vdrok | morning ironic, dtantsur|bbl lucasagomes pas-ha aarefiev alineb mgould jroll | 09:46 |
| vdrok | and morning vsaienk0 :) | 09:46 |
| mgould | morning vsaienk0 vdrok lucasagomes pas-ha aarefiev alineb jroll | 09:46 |
| *** zhangjl1 has quit IRC | 09:46 | |
| lucasagomes | vdrok, mgould morning | 09:48 |
| *** sambetts|afk is now known as sambetts | 09:55 | |
| sambetts | Morning all | 09:55 |
| sambetts | o/ vsaienk0, mgould lucasagomes jroll?! | 09:56 |
| *** hoangcx has quit IRC | 10:04 | |
| *** dtantsur|bbl is now known as dtantsur | 10:12 | |
| dtantsur | morning sambetts, lucasagomes, vdrok, mgould, vsaienk0, pas-ha, aarefiev, alineb, jroll | 10:13 |
| lucasagomes | hi there | 10:13 |
| dtantsur | wow, so many of us this morning :) | 10:13 |
| mgould | yeah, jroll is either up very early or very late | 10:24 |
| mgould | or maybe BOTH | 10:25 |
| *** chlong has quit IRC | 10:28 | |
| dtantsur | :) | 10:31 |
| *** amotoki has quit IRC | 10:31 | |
| narasimha_SV | https://bugs.launchpad.net/ironic/+bug/1609716 | 10:36 |
| openstack | Launchpad bug 1609716 in Ironic "db sync fails to create node_tags and portgroups table in mysql 5.6" [Undecided,New] | 10:36 |
| narasimha_SV | is there anyway how to resolve this issue? | 10:37 |
| *** chlong has joined #openstack-ironic | 10:41 | |
| *** ohamada has quit IRC | 10:53 | |
| *** ohamada has joined #openstack-ironic | 10:53 | |
| dtantsur | narasimha_SV, I don't know, but I don't think it's a bug in Ironic itself, rather in oslo.db or sqlalchemy. could be helpful to get their folks involved | 10:57 |
| *** tangchen_ has joined #openstack-ironic | 11:01 | |
| *** stendulker has quit IRC | 11:02 | |
| *** wlfightup has quit IRC | 11:02 | |
| tangchen_ | dtantsur: Hi Dmitry, would you please help to take a look at the functional test error here ? https://review.openstack.org/#/c/350365/ | 11:03 |
| tangchen_ | dtantsur: I used "check experimental" to run the functional. But the environment failed to establish. | 11:04 |
| *** wlfightup has joined #openstack-ironic | 11:04 | |
| dtantsur | tangchen_, there is a mistake in project-config for this job, lemme check | 11:04 |
| openstackgerrit | Merged openstack/ironic-lib: Updated from global requirements https://review.openstack.org/350886 | 11:04 |
| tangchen_ | dtantsur: Oh, OK. Thanks. :) | 11:04 |
| openstackgerrit | Merged openstack/ironic: Create a custom StringField that can process functions https://review.openstack.org/350679 | 11:05 |
| narasimha_SV | dtantsur: ok thanks I will raise there | 11:05 |
| openstackgerrit | Merged openstack/ironic: Test that network_interface is explicitly set on POST/PATCH https://review.openstack.org/349520 | 11:06 |
| fxpester | guys, I have a network in separate VLAN (from control nodes), is it possible to use Ironic in that VLAN ? | 11:07 |
| jroll | mgould: woke up early, checked on some things, went back to real life for a bit :) | 11:07 |
| aparnav | dtantsur: Hi | 11:10 |
| aparnav | dtantsur: I have a question about your comment 'need a generic way to configure steps priorities' in the RFE https://bugs.launchpad.net/ironic/+bug/1593194 | 11:11 |
| openstack | Launchpad bug 1593194 in Ironic "[RFE]Add hardware disk erase as cleaning step" [Undecided,Incomplete] - Assigned to Aparna (aparnavtce) | 11:11 |
| dtantsur | narasimha_SV, you need to put `export PROJECTS="openstack/ironic"` before https://github.com/openstack-infra/project-config/blob/master/jenkins/jobs/ironic.yaml#L241 | 11:11 |
| dtantsur | narasimha_SV, similar to https://github.com/openstack-infra/project-config/blob/master/jenkins/jobs/ironic.yaml#L274-L275 | 11:11 |
| dtantsur | aparnav, sure | 11:11 |
| jroll | tangchen_: ^ I think that PROJECTS thing was meant for you :P | 11:11 |
| mat128|afk | good morning everyone | 11:13 |
| *** mat128|afk is now known as mat128 | 11:13 | |
| aparnav | dtantsur: Should I wait for the generic way to configure the step priority to be implemented? What do you suggest? | 11:13 |
| dtantsur | aparnav, not sure about "wait", as nobody is working on it. you're free to suggest such way and implement it. | 11:14 |
| dtantsur | aarefiev, wanna review/approve https://review.openstack.org/#/c/321082/ ? | 11:14 |
| aarefiev | dtantsur: yeah, I've reviewed it, does all depended patches merged? | 11:16 |
| dtantsur | aarefiev, seems so | 11:16 |
| aarefiev | wow, looking | 11:16 |
| tangchen_ | jroll: Hi, jroll, thanks for the reply. But sorry, what PROJECTS thing ? | 11:16 |
| dtantsur | oh yeah, sorry narasimha_SV, my comments were directed at tangchen_ | 11:17 |
| * dtantsur is confusing nicks he does not see too often | 11:17 | |
| *** amotoki has joined #openstack-ironic | 11:17 | |
| dtantsur | morning mat128 | 11:18 |
| aparnav | dtantsur: Is it possible to push this RFE first and then looking into the option for implementation of generic way for clean priority? | 11:18 |
| aparnav | because it a small change | 11:18 |
| dtantsur | aparnav, I'm not in support of adding an option for a clean step which is not even upstream, sorry | 11:18 |
| dtantsur | it's a small change, but imagine every vendors comes with a few such small changes... | 11:18 |
| jroll | tangchen_: what dtantsur said, "you need to put `export PROJECTS="openstack/ironic"` before https://github.com/openstack-infra/project-config/blob/master/jenkins/jobs/ironic.yaml#L241 similar to https://github.com/openstack-infra/project-config/blob/master/jenkins/jobs/ironic.yaml#L274-L275 " | 11:19 |
| tangchen_ | jroll: Oh, I didn't see it. Thanks a lot :) | 11:20 |
| jroll | no problem :) | 11:20 |
| aparnav | dtantsur: yea I understand that. thanks | 11:20 |
| *** lucasagomes is now known as lucas-hungry | 11:21 | |
| tangchen_ | jroll, dtantsur: Hi, guys, I see. Will try it. Thanks. :) | 11:24 |
| *** aparnav has quit IRC | 11:26 | |
| *** chlong has quit IRC | 11:27 | |
| *** chlong has joined #openstack-ironic | 11:28 | |
| fxpester | guys, is ironic support dhcp relay ? it is possible to deploy BM in separate VLAN ? | 11:29 |
| dtantsur | fxpester, ironic is not involved in DHCP directly, and I doubt that neutron supports DHCP relays | 11:30 |
| dtantsur | fxpester, if you figure out the way to make it work - lemme know, we would be also very interested :) | 11:30 |
| dtantsur | worth checking with neutron folks on #openstack-neutron, I think | 11:31 |
| openstackgerrit | Merged openstack/proliantutils: Modify error message of InvalidInputException in hpssa https://review.openstack.org/338894 | 11:33 |
| mgould | dtantsur, fxpester: OTOH, ironic-inspector does deal with DHCP directly | 11:37 |
| dtantsur | yeah, but the question was about ironic :) | 11:37 |
| * mgould doesn't know enough about DHCP relays to know if inspector would work with them, but guesses not | 11:37 | |
| dtantsur | dsneddon tried, I think | 11:38 |
| *** xek__ has joined #openstack-ironic | 11:41 | |
| *** xek_ has quit IRC | 11:42 | |
| openstackgerrit | Merged openstack/ironic: Update devstack plugin with new auth options https://review.openstack.org/297312 | 11:46 |
| jroll | I think you have to configure your switches yourself to do dhcp relay, unless an ML2 plugin will do that | 11:47 |
| TheJulia | Good morning | 11:48 |
| mat128 | jroll: ^ bingo | 11:48 |
| mat128 | I meant fxpester: ^ | 11:48 |
| mat128 | hi TheJulia | 11:48 |
| dtantsur | morning TheJulia | 11:48 |
| jroll | morning TheJulia, mat128 :) | 11:48 |
| dtantsur | jroll, I've heard complaints that dnsmasq does not deal with relayed DHCP requests properly.. | 11:49 |
| vsaienk0 | morning TheJulia, mat128 | 11:49 |
| TheJulia | dtantsur: I've heard the same, but I have no firsthand experience with trying it. | 11:49 |
| jroll | dtantsur: oh hey, another reason not to use dnsmasq :) | 11:49 |
| dtantsur | hehe | 11:49 |
| jroll | that isn't good though | 11:49 |
| jroll | I wonder if it's dnsmasq itself or the way neutron configures it | 11:49 |
| jroll | almost like I need a lab or something... >.> | 11:50 |
| *** amotoki has quit IRC | 11:50 | |
| dtantsur | jroll, our guy didn't succeed with dnsmasq standalone (for ironic-inspector) too | 11:50 |
| jroll | ouch | 11:50 |
| dtantsur | ... and don't get me started on having an actual lab >_< | 11:50 |
| jroll | that's something we need to fix | 11:50 |
| jroll | even if it means contributing isc support to neutron | 11:50 |
| * jroll is going to hate himself for saying that | 11:51 | |
| * fxpester having headache with dnsmask&dhcrelay right now ^^ | 11:51 | |
| jroll | JayF: morgabra: ^^ you'll be interested in this conversation | 11:51 |
| dtantsur | he was thinking about writing his own DHCP server in Python :) | 11:51 |
| mat128 | jroll: thats a big one :) | 11:51 |
| TheJulia | partially | 11:51 |
| jroll | dtantsur: a dhcp server written with twisted has been on my "someday" list for a couple years now | 11:51 |
| TheJulia | a lot of the concepts are the same | 11:51 |
| dtantsur | jroll, oh cool. I used to do a lot of Twisted a couple of years go btw | 11:52 |
| mat128 | TheJulia: can't tell who you are replying to? | 11:52 |
| TheJulia | jroll: If you do that, we may have to question your sanity :) | 11:52 |
| jroll | we've been considering moving to using neutron for dhcp, this would put a big rock in that plan :( | 11:52 |
| TheJulia | mat128: you :) | 11:52 |
| jroll | dtantsur: neat, I worked in the same office as glyph for a while, he's the one trying to convince me to write it | 11:52 |
| dtantsur | jroll, anyway, for now I'm thinking more about neutron-less operation and static DHCP in every L2 domain | 11:52 |
| TheJulia | mat128: Relative sizing differences I guess :) | 11:52 |
| mat128 | TheJulia: ah! so you dont think neutron-isc work is all that big? | 11:52 |
| mat128 | :) | 11:52 |
| jroll | dtantsur: yeah, that's what we do now | 11:53 |
| dtantsur | jroll, that's why I'm thinking about that ;) | 11:53 |
| jroll | dtantsur: there's some problems, like needing to deploy dhcp configs every time you add hardware | 11:53 |
| TheJulia | mat128: I would say medium-ish if a real lab was available >.> | 11:53 |
| jroll | it also means one ramdisk for every node unless you get fancy | 11:53 |
| dtantsur | jroll, we do have one ramdisk in tripleo, so it's not a huge deal | 11:53 |
| mat128 | TheJulia: why do you need a real lab for this? | 11:54 |
| dtantsur | jroll, wdym by "add hardware" here? | 11:54 |
| TheJulia | mat128: testing dhcp relaying of course | 11:54 |
| jroll | dtantsur: adding nodes | 11:54 |
| mat128 | TheJulia: ah.. yeah | 11:54 |
| mat128 | TheJulia: we have one pair of static DHCP servers and a relay in every L2 domain | 11:54 |
| dtantsur | jroll, hmm, so do you use MAC-based configuration for nodes, not just /tftpboot/pxelinux.cfg/default (roughly speaking)? | 11:55 |
| mat128 | TheJulia: I just trust the network equipment to do relaying correctly and blame it on the network guys if it doesn't ;) | 11:55 |
| openstackgerrit | Tang Chen proposed openstack/python-ironicclient: Fix i18n problems in shell.py https://review.openstack.org/349343 | 11:55 |
| jroll | dtantsur: you also need to know the mac address ahead of time, which means convincing your vendor to provide those, and the inevitable typo | 11:55 |
| TheJulia | mat128: lol | 11:55 |
| jroll | dtantsur: well, we use macs for configuring static DHCP :) | 11:55 |
| jroll | you need to know the mac address to do it statically | 11:55 |
| dtantsur | I see. I'm thinking of reusing inspector's catch-all configuration | 11:55 |
| mat128 | jroll: why not a catchall? | 11:56 |
| jroll | ah | 11:56 |
| openstackgerrit | Merged openstack/ironic: Revert "Devstack should use a prebuilt ramdisk by default" https://review.openstack.org/350789 | 11:56 |
| dtantsur | (and local boot for instances, obviously) | 11:56 |
| jroll | we had a reason | 11:56 |
| jroll | now I'm wondering what that was | 11:56 |
| mat128 | different kernel cmdline per machine? | 11:56 |
| jroll | I think it's basically so the agent always has the same IP in the provisioning/cleaning networks | 11:56 |
| dtantsur | maybe root device hints? | 11:56 |
| *** bfournie has quit IRC | 11:56 | |
| jroll | nope | 11:56 |
| dtantsur | ah | 11:56 |
| jroll | we're very homogenous, no need for different kernel cmd line or anything | 11:57 |
| dtantsur | cool, that's what I wanted to confirm | 11:57 |
| jroll | probably a mix of 'same IP always' and "if someone manages to get on the network, they don't get an agent, at least without some mac spoof effort" | 11:57 |
| mat128 | jroll: oh i see | 11:58 |
| *** phuongnh has quit IRC | 11:58 | |
| mat128 | jroll: we made the mistake of considering a similar "management" network as blessed in a previous (circa 2009/2010) product | 11:58 |
| *** lucas-hungry is now known as lucasagomes | 11:58 | |
| dtantsur | makes sense, though knowing MACs is probably not acceptable in hour case | 11:58 |
| mat128 | jroll: we just consider that network as hostile now, and make all decisions accordingly | 11:58 |
| dtantsur | s/hour/our/ of course | 11:58 |
| jroll | mat128: yeah, safe to assume every network is owned | 11:59 |
| mat128 | dtantsur: if ironic reconfigures DHCP based on node config, you don't need to deploy dhcp configs manually | 11:59 |
| mat128 | dtantsur: it would, but dnsmasq :( | 11:59 |
| dtantsur | well, yeah, if we had DHCP relay support, everything would just work | 11:59 |
| mat128 | dtantsur: not sure it's even relay | 11:59 |
| mat128 | dtantsur: you could tell, neutron, please put a dhcp in that network_uuid and make it serve these leases | 12:00 |
| mat128 | dtantsur: you don't really care how it does it (multiple dhcps, one and relays, etc.) | 12:00 |
| dtantsur | mat128, but how will it work, if neutron is on different network and dnsmasq does not (apparently) support relaying? | 12:00 |
| mat128 | dtantsur: neutron will take care of the dhcp on $deployment_network | 12:00 |
| mat128 | with that upcoming work :) | 12:01 |
| dtantsur | mat128, yes, if it has access to the L2 domain, right? | 12:01 |
| dtantsur | maybe I'm saying stupid things, please bear with me for a second :) | 12:01 |
| mat128 | yes, if you use real hardware then your ML2 plugin has to support configuring this | 12:01 |
| mat128 | the relay is just to make DHCP requests across L2 domains | 12:02 |
| openstackgerrit | Kyrylo Romanenko proposed openstack/ironic-specs: Amend driver section of spec for ironic OSC plugin https://review.openstack.org/351140 | 12:02 |
| mat128 | it basically encapsulates back and forth | 12:02 |
| mat128 | same as having a (distinct) dhcp in every L2 | 12:02 |
| *** clenimar_ has joined #openstack-ironic | 12:02 | |
| dtantsur | we're highly heterogeneous :( so we can't always rely on even present ML2 plugin | 12:02 |
| TheJulia | Shamless plug for the oneview driver team, Their dynamic allocation feature presently has 2x +2s https://review.openstack.org/#/c/286192/ | 12:03 |
| mat128 | dtantsur: then neutron has to have L2 access or you have to configure dhcp relay manually | 12:03 |
| mat128 | dtantsur: no other way of doing network boot | 12:04 |
| dtantsur | yeah, this is my point: we have a case when they don't want an L2 access (sigh) | 12:04 |
| *** zhangjl has quit IRC | 12:04 | |
| mat128 | dtantsur: from a sec. standpoint, it makes sense | 12:04 |
| jroll | looking at dnsmasq man page, seems like there is support for relays, it may just need extra configuration | 12:05 |
| openstackgerrit | Tang Chen proposed openstack/ironic: Split API root unit and functional tests https://review.openstack.org/348965 | 12:05 |
| mat128 | dtantsur: even if dnsmasq had relay support (integrated in neutron), it wouldn't help you dtantsur | 12:05 |
| openstackgerrit | Tang Chen proposed openstack/ironic: Move api driver tests to functional dir https://review.openstack.org/350365 | 12:05 |
| openstackgerrit | Tang Chen proposed openstack/ironic: Prepare for functional tests https://review.openstack.org/349421 | 12:05 |
| jroll | maybe it is just a neutron thing | 12:05 |
| * mat128 mentions people twice... need more coffee | 12:05 | |
| TheJulia | mat128: espresso :) | 12:06 |
| jroll | see the text for --dhcp-range and --dhcp-proxy: http://www.thekelleys.org.uk/dnsmasq/docs/dnsmasq-man.html | 12:06 |
| mat128 | jroll: but then that dnsmasq instance *does* relaying | 12:06 |
| mat128 | you need to run this in the L2 domain where you want dhcp requests to work | 12:06 |
| jroll | mat128: no, that's a different option | 12:06 |
| mgould | morning mat128 TheJulia | 12:06 |
| jroll | not surprisingly called --dhcp-relay | 12:07 |
| *** clenimar_ has quit IRC | 12:07 | |
| TheJulia | good morning mgould | 12:07 |
| mat128 | jroll: my understanding of this option is: relay the first call, let the rest happen with the real (origin) server | 12:07 |
| mat128 | origin as in cdn lingo | 12:07 |
| jroll | mat128: yeah, I think that's how normal dhcp relays work, the request is relayed, the rest is direct | 12:08 |
| mat128 | jroll: something has to have L2 access, either a relay (it could be a physical router) | 12:08 |
| jroll | mat128: so --dhcp-range talks about caveats with that | 12:08 |
| mat128 | jroll: real relays, nope | 12:08 |
| mat128 | see it more like a proxy | 12:08 |
| jroll | mat128: and --dhcp-proxy makes *all* dhcp traffic go through the relay | 12:08 |
| mat128 | eh | 12:08 |
| jroll | which should do what you want here | 12:08 |
| mat128 | jroll: ok I think I got it wrong | 12:08 |
| mat128 | but IRL, a "dhcp relay" is what you describe as a proxy | 12:08 |
| jroll | s/you/dnsmasq/ | 12:09 |
| mat128 | given the original host has no access to the DHCP living in a different L2 domain | 12:09 |
| mat128 | haha | 12:09 |
| lazy_prince | I am trying to setup ironic using devstack and with flat network. if someone has done this before, it could save my time.. | 12:09 |
| mat128 | TheJulia was right, dnsmasq implementation is all contained in "dhcp.py" in neutron. Doesn't look pluggable, but I guess it woul be possible | 12:10 |
| jroll | so I think neutron just needs to know to pass that (or maybe "extra dnsmasq options" are a config thing?) | 12:10 |
| jroll | wait, neutron has things that aren't pluggable? | 12:10 |
| mat128 | jroll: I don't think you can choose your dhcp implementation | 12:10 |
| mat128 | i might be wrong | 12:10 |
| jroll | there's a dhcp_driver config in neutron.conf.agent.dhcp | 12:11 |
| mat128 | jroll: well, if you happen to have a software that has the dnsmasq "api" in all forms and shape, you can simply change "DNSMASQ_SERVICE_NAME = 'dnsmasq'" to something else | 12:11 |
| TheJulia | mat128: I'm full of semi-useless information because I've had to dig through much of the code base debugging issues :) | 12:11 |
| lazy_prince | there are neutron dhcp agents, the default uses dnsmasq, but you could write one which uses isc-dhcp-server too.. | 12:11 |
| jroll | lazy_prince: agree | 12:12 |
| jroll | looks like it'd be pretty easy to add a CONF.dhcp_proxy=<relay_ip> and add it to the dnsmasq implementation | 12:13 |
| jroll | again, I'd need a lab to actually test this | 12:13 |
| mat128 | jroll, lazy_prince: then I'm wrong :) it,s pluggable | 12:13 |
| jroll | and some switch config skills | 12:13 |
| openstackgerrit | Merged openstack/ironic-inspector: Add GenericLocalLinkConnectionHook processing hook https://review.openstack.org/321082 | 12:13 |
| mat128 | jroll: I'm not sure how that would help given you need L2 access | 12:14 |
| mat128 | which is what dtantsur was trying to avoid | 12:14 |
| *** afaranha_ has quit IRC | 12:14 | |
| jroll | mat128: --dhcp-proxy makes everything go through the relay | 12:15 |
| mat128 | but the relay has to listen on that other vlan | 12:15 |
| jroll | which I think means you don't need direct L2, right? | 12:15 |
| jroll | sure? | 12:15 |
| jroll | I mean, I'd have the relay in the TOR switch | 12:15 |
| mat128 | jroll: if you run a dhcp at your place, and you tell it "proxy to mat128's dhcp server" | 12:15 |
| mat128 | jroll: thats whats missing | 12:16 |
| mat128 | the real dhcp server doesn't have to know someone is using it through a relay | 12:16 |
| jroll | I'm confused about the requirements, I guess | 12:16 |
| TheJulia | mat128: I thought dhcp servers needed to be aware of the range and the existence of the relay to work correctly... | 12:17 |
| mat128 | let me check our config real quick | 12:17 |
| jroll | what I'm imagining is [node] <----vlan2----> [switch-with-dhcp-relay] <----vlan3----> [dhcp-server] | 12:17 |
| jroll | which means node and dhcp-server can't talk over l2 | 12:17 |
| TheJulia | I've had to do relaying before and I remember having to do some server side config to make it happy | 12:17 |
| mat128 | that "switch with dhcp relay" is normally a l3 switch/router, but yeah | 12:17 |
| jroll | but if everything is relayed through the switch, you're good | 12:17 |
| TheJulia | but, it was a windows dhcp server | 12:17 |
| *** Goneri has quit IRC | 12:17 | |
| mat128 | jroll, TheJulia: our conf has only the subnet to be served | 12:19 |
| mat128 | and the classic stuff, dns, next-server and such | 12:19 |
| jroll | ditto, with isc | 12:19 |
| *** MattMan has quit IRC | 12:19 | |
| jroll | seems dnsmasq needs extra config | 12:19 |
| *** MattMan has joined #openstack-ironic | 12:19 | |
| mat128 | our L3/router would encapsulate node's request it saw on vlan2, transfer it to vlan3 and encapsulate the reply back | 12:19 |
| mat128 | to be broadcasted on vlan2 | 12:20 |
| *** baoli has joined #openstack-ironic | 12:20 | |
| mat128 | but dtantsur was saying he has no way of configuring that vlan with a relay | 12:20 |
| vdrok | huh, has anyone seen this before? http://logs.openstack.org/99/325599/24/check/gate-ironic-python35-db/2fbac81/console.html | 12:20 |
| mat128 | the altnernative is having a dhcp-server on every segment, which is what neutron does when you use it's dnsmasq impl. | 12:20 |
| *** narasimha_SV has quit IRC | 12:21 | |
| mat128 | vdrok: timeout running the db migration? | 12:21 |
| jroll | mat128: I mean, if you don't have direct connectivity to the dhcp server, you need a relay ofc :P | 12:21 |
| vdrok | yep | 12:21 |
| jroll | the dnsmasq config for the two different types of relays https://gist.githubusercontent.com/jimrollenhagen/cb2ff54fc954235cd129cf5b29a850c0/raw/e14308231f70493407f73a6df8fbe6c1e526822d/gistfile1.txt | 12:21 |
| dtantsur | vdrok, just looked, suspicious.. | 12:21 |
| jroll | that's odd | 12:22 |
| jroll | mat128: I'd love to run the dhcp agent itself on our TORs | 12:22 |
| jroll | I think just knowing they can run python I've been wanting to run something on them heh | 12:23 |
| mat128 | jroll: you can always get your hardware from https://www.cumulusnetworks.com/ | 12:23 |
| mat128 | jroll: I've seen a demo from them in Van where they has the neutron agent running directly on the switch | 12:23 |
| jroll | mat128: yeah, we run arista in v2, same idea, full linux | 12:23 |
| jroll | oh look, they put BGP on the host | 12:24 |
| jroll | \o/ | 12:24 |
| *** Goneri has joined #openstack-ironic | 12:25 | |
| mat128 | jroll: yeah, I've heard it eats 30% of your computing power though, unsure why (it's just announces / routing) | 12:25 |
| jroll | mat128: because quagga :D | 12:25 |
| *** baoli_ has joined #openstack-ironic | 12:25 | |
| mat128 | good old specialized hardware vs generic stuff debate | 12:25 |
| mat128 | yeah | 12:25 |
| *** dprince has joined #openstack-ironic | 12:28 | |
| *** fellypefca has joined #openstack-ironic | 12:28 | |
| *** baoli has quit IRC | 12:29 | |
| *** bfournie has joined #openstack-ironic | 12:31 | |
| *** hshiina has quit IRC | 12:33 | |
| dtantsur | some time ago I was advocating for iPXE as a more stable technology.. I'd like to take my words back >_< | 12:40 |
| lucasagomes | lol | 12:40 |
| lucasagomes | problem is the alternatives are not great either | 12:40 |
| mgould | dtantsur: what do you mean, unstable? The version packaged in RPMs hasn't been updated for three years... | 12:40 |
| dtantsur | I'm using a new one (well, new.. from January) | 12:41 |
| * jroll is curious what's wrong with ipxe | 12:41 | |
| * mgould is starting to think the answer is "just leave your computers turned off, no good can come of booting them" | 12:41 | |
| jroll | +1 | 12:41 |
| dtantsur | it's unstable like hell... | 12:41 |
| lucasagomes | jroll, one thing is their releases | 12:41 |
| dtantsur | mgould++ | 12:41 |
| lucasagomes | which means, no releases heh | 12:41 |
| lucasagomes | the last release of ipxe was in 2010 I think | 12:41 |
| * mgould learned yesterday that ipxe error codes are actually bitmasks | 12:41 | |
| mgould | but the docs don't specify the endianness | 12:42 |
| lazy_prince | I have seen iPXE issues with Emulex when booting in UEFI mode.. | 12:43 |
| lucasagomes | mgould, you usually get an error code no? E.g http://ipxe.org/err/4c1060 | 12:43 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add keystone policy support to Ironic https://review.openstack.org/325599 | 12:43 |
| *** amotoki has joined #openstack-ironic | 12:43 | |
| jroll | rebase to pick up unit tests fix ^ | 12:43 |
| mgould | lucasagomes: yeah, and yesterday I got http://ipxe.org/err/040ee186 | 12:43 |
| mgould | which redirects to 040ee1 | 12:44 |
| jroll | dtantsur: let's get your agent api thing merged today :) | 12:44 |
| mgould | which led to me grepping the source code and finding that the error codes are constructed at runtime by bit-twiddling other things | 12:44 |
| TheJulia | lucasagomes: yeah, last tagged in 2010, but still gets fixes and updates on their mater branch | 12:44 |
| lucasagomes | IPv6 is also disabled by default in iPXE so you have to change the code and rebuild the image (/me added instructions here https://wiki.openstack.org/wiki/Ironic/IPv6) | 12:44 |
| lucasagomes | TheJulia, yeah | 12:44 |
| dtantsur | jroll, let's try, but I have an hour-long meeting in 15 minutes | 12:44 |
| lucasagomes | TheJulia, it's just harder for packaging I guess | 12:44 |
| TheJulia | yeah | 12:44 |
| jroll | dtantsur: if you agree with my 404 comment, I can do the rebase and fix it in a moment | 12:44 |
| jroll | s/moment/little while/ | 12:45 |
| openstackgerrit | Merged openstack/ironic-specs: Amend driver section of spec for ironic OSC plugin https://review.openstack.org/351140 | 12:45 |
| * jroll bbiab | 12:46 | |
| dtantsur | jroll, I agree, please do. however, the rebase might be tricky | 12:46 |
| jroll | dtantsur: well, I'll take a look and if it's hard I'll leave it for you :P | 12:46 |
| mgould | lucasagomes: oh God, the bitmasks aren't even aligned to byte boundaries | 12:48 |
| xavierr | good morning Ironic | 12:50 |
| dtantsur | jroll, lucasagomes, here's the bug, feel free to look at screenshots: https://bugzilla.redhat.com/show_bug.cgi?id=1364079 | 12:50 |
| openstack | bugzilla.redhat.com bug 1364079 in ipxe "iPXE hangs with an infinite stream of different errors" [Unspecified,New] - Assigned to rhos-maint | 12:50 |
| mgould | morning xavierr | 12:51 |
| xavierr | jroll: any idea when going to happen the next ironic cut? | 12:51 |
| xavierr | hey mgould | 12:51 |
| * lucasagomes looks | 12:51 | |
| *** dtantsur is now known as dtantsur|mtg | 12:54 | |
| lucasagomes | dtantsur|mtg, is it a virtual machine? | 12:54 |
| openstackgerrit | Merged openstack/ironic: Add metrics for the ipmitool driver https://review.openstack.org/350502 | 12:54 |
| dtantsur|mtg | lucasagomes, no, hw | 12:55 |
| lucasagomes | ack, I can try to investigate after the meeting if needed | 12:55 |
| xavierr | morning dtantsur|mtg lucasagomes o/ | 12:56 |
| lucasagomes | xavierr, hi there :-) | 12:57 |
| xavierr | hey lucasagomes, need you help | 12:58 |
| xavierr | hehehe | 12:58 |
| xavierr | s/you/your/ | 12:58 |
| *** tangchen_ has quit IRC | 12:58 | |
| *** mgould is now known as mgould|mtg | 12:58 | |
| *** tangchen has quit IRC | 12:59 | |
| xavierr | lucasagomes: could you take a look on 286192 and +a if possible? | 12:59 |
| lucasagomes | xavierr, yeah, I've a meeting right now but after that I will take a look | 13:00 |
| xavierr | lucasagomes: awesome lucas, thank you so much | 13:00 |
| dtantsur|mtg | milan, retro? | 13:01 |
| milan | dtantsur|mtg, oops | 13:01 |
| * milan logs in | 13:01 | |
| openstackgerrit | Merged openstack/ironic: Add a data migration to fill node.network_interface https://review.openstack.org/349528 | 13:01 |
| *** jpich has quit IRC | 13:03 | |
| *** Lord_Phoenix has quit IRC | 13:09 | |
| *** sdake has joined #openstack-ironic | 13:11 | |
| jroll | xavierr: I'd like to do 6.1.0 next week and 7.0.0 by the end of the cycle | 13:13 |
| jroll | that patch that just merged is what I was waiting for :D | 13:13 |
| *** permalac has joined #openstack-ironic | 13:13 | |
| permalac | hello guys. | 13:13 |
| permalac | I have a baremetal compute node that I want to discover with ironic, but this compute only has support for uefi. | 13:14 |
| permalac | I'm on kilo and I'm no sure if the ironic has support for uefi discovery . | 13:14 |
| mgould|mtg | permalac: not until Liberty | 13:14 |
| permalac | mgould|mtg, no workarround, not even with something like this ? => http://docs.openstack.org/developer/ironic/kilo/deploy/install-guide.html#pxe-uefi-setup | 13:15 |
| mgould|mtg | permalac: not that's documented, though I guess you could make the changes in https://review.openstack.org/#/c/308271/ manually | 13:18 |
| mgould|mtg | also https://review.openstack.org/#/c/308272/ | 13:19 |
| * mgould|mtg is in a meeting and can't really pay attention, sorry | 13:19 | |
| permalac | mgould|mtg, thanks. I don't really know how to read this change but I'll try. :) | 13:20 |
| *** sdake has quit IRC | 13:24 | |
| *** thiagop has joined #openstack-ironic | 13:25 | |
| thiagop | Good morning folks | 13:26 |
| *** wlfightup has quit IRC | 13:34 | |
| *** Lord_Phoenix has joined #openstack-ironic | 13:34 | |
| jroll | dtantsur|mtg: you might enjoy [openstack-dev] [Nova] Some thoughts on API microversions | 13:36 |
| jroll | :D | 13:36 |
| thiagop | lol | 13:37 |
| *** jvgrant_ has joined #openstack-ironic | 13:38 | |
| *** jpich has joined #openstack-ironic | 13:45 | |
| *** mgoddard_ has joined #openstack-ironic | 13:45 | |
| *** ametts has joined #openstack-ironic | 13:45 | |
| *** jraju has quit IRC | 13:46 | |
| NobodyCam | Good morning ironic'ers :) | 13:47 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Promote agent vendor passthru to core API https://review.openstack.org/330018 | 13:47 |
| jroll | dtantsur|mtg: ^ rebased, fixed my -1s | 13:48 |
| *** mgoddard has quit IRC | 13:48 | |
| *** fellypefca has quit IRC | 13:50 | |
| xavierr | jroll: thank you :) | 13:50 |
| *** ppiela has quit IRC | 13:51 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Promote agent vendor passthru to core API https://review.openstack.org/330018 | 13:53 |
| jroll | and some fixes | 13:54 |
| *** sdake has joined #openstack-ironic | 13:54 | |
| *** PollyZ has joined #openstack-ironic | 13:58 | |
| *** mgoddard_ has quit IRC | 13:58 | |
| *** mgoddard has joined #openstack-ironic | 13:58 | |
| *** baoli_ has quit IRC | 13:59 | |
| *** milan has quit IRC | 14:01 | |
| *** milan has joined #openstack-ironic | 14:01 | |
| *** PollyZ has quit IRC | 14:02 | |
| *** joprovost has joined #openstack-ironic | 14:02 | |
| *** PollyZ has joined #openstack-ironic | 14:03 | |
| jroll | folks, let's see if we can get ^ that one in today | 14:04 |
| jroll | too many rebases | 14:04 |
| mgould|mtg | permalac: copy ipxe.efi to /tftpboot on the machine running inspector, create an inspector.ipxe file based on the template and put it in /httpboot, patch /etc/ironic-inspector/dnsmasq.conf to contain the logic in the other template | 14:04 |
| mgould|mtg | probably some other things | 14:04 |
| mgould|mtg | tl;dr are you really sure you want to do this? :-) | 14:05 |
| mgould|mtg | NobodyCam: morning! | 14:06 |
| mgould|mtg | morning thiagop | 14:06 |
| thiagop | morning NobodyCam mgould|mtg | 14:06 |
| xavierr | mgould|mtg: what does mean 'mtg'? hehe | 14:07 |
| NobodyCam | Morning mgould|mtg, thiagop, xavierr and jroll | 14:08 |
| mat128 | xavierr: I would guess "meeting" | 14:08 |
| mat128 | xavierr: dtantsur has it too ;) | 14:09 |
| xavierr | hey NobodyCam \o/ | 14:09 |
| xavierr | mat128: interesting | 14:09 |
| NobodyCam | Morning mat128 | 14:09 |
| xavierr | ;) | 14:10 |
| NobodyCam | O/ | 14:10 |
| mat128 | hey :) | 14:13 |
| dtantsur|mtg | jroll, thanks for rebasing! | 14:13 |
| dtantsur|mtg | oh microversions, not again...... | 14:13 |
| NobodyCam | Morning dtantsur|mtg ... lol | 14:14 |
| jroll | dtantsur|mtg: it's a relevant discussion | 14:14 |
| *** dtantsur|mtg is now known as dtantsur | 14:14 | |
| dtantsur | okie, I'll check it | 14:14 |
| dtantsur | will take some rest from the meeting first | 14:15 |
| *** mgould|mtg is now known as mgould | 14:15 | |
| jroll | yeah no rush ofc | 14:15 |
| mgould | xavierr: "in a meeting" | 14:15 |
| *** livelace has joined #openstack-ironic | 14:16 | |
| mgould | hence both me and dtantsur having the tag simultaneously :-) | 14:16 |
| xavierr | I will be in a meeting in minutes, so... | 14:16 |
| *** xavierr is now known as xavierr|mtg | 14:17 | |
| * mgould was in meetings before it was cool | 14:17 | |
| xavierr|mtg | o/ | 14:17 |
| xavierr|mtg | hahahaha | 14:17 |
| mgould | o/ | 14:17 |
| NobodyCam | :) | 14:17 |
| *** hoangcx has joined #openstack-ironic | 14:17 | |
| dtantsur | also morning NobodyCam, xavierr|mtg, thiagop and everyone else | 14:17 |
| thiagop | dtantsur: o/ | 14:18 |
| *** moshele has quit IRC | 14:19 | |
| *** syed_ has joined #openstack-ironic | 14:22 | |
| *** PollyZ has quit IRC | 14:22 | |
| *** ppiela has joined #openstack-ironic | 14:22 | |
| *** amotoki has quit IRC | 14:22 | |
| permalac | mgould, I have to do it. | 14:22 |
| mgould | permalac: good luck! | 14:23 |
| *** PollyZ has joined #openstack-ironic | 14:23 | |
| mgould | permalac: do you have iPXE enabled for introspection already? | 14:24 |
| *** joprovost has quit IRC | 14:24 | |
| permalac | mgould, I have not done it myself, but if its default it maybe. | 14:27 |
| mgould | alas, no | 14:28 |
| mgould | instructions here: http://docs.openstack.org/developer/ironic-inspector/install.html#configuring-ipxe | 14:28 |
| mgould | oh, cool, that dnsmasq.conf has UEFI support | 14:28 |
| mgould | yeah, I think those instructions should cover you | 14:29 |
| mgould | you'll need a recent version of iPXE in order to get the file ipxe.efi: you might have to download it from ipxe.org | 14:29 |
| *** mtanino has joined #openstack-ironic | 14:31 | |
| mgould | permalac: if you want to *deploy* to UEFI machines, you'll also need to follow the instructions at http://docs.openstack.org/developer/ironic/deploy/install-guide.html?highlight=elilo#ipxe-setup | 14:31 |
| mgould | pay particular attention to step 6 | 14:31 |
| mgould | [no idea if those instructions work on Kilo, sorry] | 14:31 |
| permalac | mgould, almost got me. :D | 14:32 |
| *** vishwanathj has joined #openstack-ironic | 14:32 | |
| mgould | ? | 14:32 |
| *** catintheroof has joined #openstack-ironic | 14:32 | |
| permalac | I'll have to try very slowly, its my first time with rh director and I go slow. | 14:32 |
| * mgould nods | 14:32 | |
| *** tangchen has joined #openstack-ironic | 14:32 | |
| *** tangchen_ has joined #openstack-ironic | 14:32 | |
| permalac | I'll have to try, I thought you just gave me gold. But then you come with the "not sure if works with kilo" , so you almost got me there. :) | 14:33 |
| mgould | permalac: I've been trying to debug a failing UEFI deployment in our lab for the last 2.5 days, so can confirm this is not an easy place to start | 14:33 |
| *** Lord_Phoenix has quit IRC | 14:33 | |
| fxpester | just build devstack with ironic, tried to start VM by instruction, it is in state "testing | BUILD | spawning | NOSTATE | private=10.1.0.6" for a hour now | 14:34 |
| mgould | permalac: the introspection-with-UEFI instructions should work, but I'm less sure about the deployment ones | 14:34 |
| mgould | I'm not sure when ironic added support for UEFI deployments | 14:35 |
| xavierr|mtg | fxpester: have you already seen the logs | 14:35 |
| xavierr|mtg | ? | 14:35 |
| *** Goneri has quit IRC | 14:35 | |
| permalac | mgould, I'll have to dig deep then. :'/ | 14:35 |
| fxpester | logs: "Command stdout is: "Chassis Power is off" | 14:35 |
| *** joprovost has joined #openstack-ironic | 14:36 | |
| mat128 | fxpester: thats log output from ipmitool, do you know if the machine is doing anything? | 14:37 |
| fxpester | mat128: hm /usr/bin/qemu-system-x86_64 is strted | 14:37 |
| mat128 | virsh list, try to get the graphical / serial console on it | 14:38 |
| mat128 | it may be redirected in a file, depending on your config | 14:38 |
| mgould | permalac: looks like you're out of luck https://bugs.launchpad.net/ironic/+bug/1525989 | 14:38 |
| openstack | Launchpad bug 1525989 in Ironic "iPXE can't be configured with UEFI" [Wishlist,Fix released] - Assigned to Lucas Alvares Gomes (lucasagomes) | 14:38 |
| mgould | anyone: is it possible to deploy to UEFI machines with Kilo? | 14:39 |
| openstackgerrit | Hugo Nicodemos proposed openstack/python-oneviewclient: Workaround to OneView pagination https://review.openstack.org/348532 | 14:40 |
| permalac | mgould, then I have a misconception here, UEFI requires iPXE , its a must , can not be done without iPXE, no other way . But I have a kickstart file to install my uefi machines, this same kind of machines from a tfpt server on another place. | 14:41 |
| openstackgerrit | Akilan Pughazhendi proposed openstack/ironic: Updated Dev quickstart for viewing doc changes https://review.openstack.org/350705 | 14:41 |
| mgould | permalac: I don't know if *ironic* has support for UEFI machines without using iPXE | 14:42 |
| * mgould <- still fairly new here | 14:42 | |
| mat128 | mgould: I don't think it can since only iPXE can boot EFI, afaik | 14:42 |
| mgould | huh, OK | 14:42 |
| mat128 | mgould, permalac: let me rephrase that: given iPXE is required to boot EFI, I don't see any other way Ironic could use to boot EFI without iPXE | 14:43 |
| mat128 | does that make sense? | 14:43 |
| mgould | yep | 14:43 |
| * mat128 is an EFI noob, though | 14:43 | |
| * mgould doesn't really understand it either | 14:43 | |
| *** jcoufal has joined #openstack-ironic | 14:45 | |
| lucasagomes | xavierr|mtg, around? Do you have some time to do a follow up on that oneview patch? | 14:45 |
| permalac | mat128, yes. thanks for the clarification. | 14:46 |
| xavierr|mtg | sure lucasagomes :) | 14:46 |
| permalac | how to you guys do that * thing. Like *I'm in a meeting, * <-- still farily new here. | 14:46 |
| permalac | :) | 14:46 |
| mgould | lucasagomes: is iPXE required to support UEFI? | 14:46 |
| mgould | (a) in general, (b) in Ironic? | 14:46 |
| lucasagomes | xavierr|mtg, cool it should quick, I will +2 it after that | 14:46 |
| mat128 | type /me something | 14:46 |
| lucasagomes | xavierr|mtg, it's looking good btw | 14:46 |
| * mat128 something | 14:47 | |
| mat128 | permalac: ^ this? | 14:47 |
| lucasagomes | mgould, not really, you can use other bootloaders for that | 14:47 |
| mgould | cool, that's what I thought | 14:47 |
| mgould | permalac is trying to configure a Kilo cloud to introspect and deploy to UEFI nodes; is this possible? | 14:48 |
| * xavierr|mtg will do it right now lucasagomes :) | 14:48 | |
| lucasagomes | xavierr|mtg, cool, just ping me when it's done and i will +2 it | 14:49 |
| xavierr|mtg | lucasagomes: thanks | 14:49 |
| * mgould -> lunch, brb | 14:49 | |
| lucasagomes | mgould, late lunch, enjoy :D | 14:50 |
| *** Sylvester has joined #openstack-ironic | 14:51 | |
| jroll | multitenant network nova patch is merged \o/ \o/ \o/ \o/ \o/ | 14:59 |
| jroll | great work everyone | 14:59 |
| thiagop | w00t!!!! | 15:01 |
| JayF | jroll: also b/c we plug both network interfaces into agent/rescue networks | 15:01 |
| *** tangchen has quit IRC | 15:03 | |
| *** tangchen_ has quit IRC | 15:03 | |
| *** wajdi has quit IRC | 15:04 | |
| jlvillal | jroll: Woo hoo! :) | 15:05 |
| sambetts | \o/ | 15:05 |
| fxpester | mat128: done VNC, VM going from `Booting Kernel` state directly to loop X_X | 15:05 |
| mat128 | jroll: party! | 15:06 |
| mat128 | fxpester: what is ironic node-show saying? | 15:06 |
| mat128 | fxpester: in paste.openstack.org preferably | 15:06 |
| dtantsur | wow wow wow | 15:07 |
| openstackgerrit | Michael Turek proposed openstack/python-ironicclient: Add possibility to work with portgroups https://review.openstack.org/335964 | 15:07 |
| fxpester | mat128: http://paste.openstack.org/show/549247/ | 15:08 |
| mat128 | fxpester: that machine is available but it has an instance_uuid | 15:08 |
| *** rcernin has quit IRC | 15:09 | |
| mat128 | fxpester: can you grep the node uuid in ironic-conductor logs (ir-cond) ? | 15:09 |
| mat128 | fxpester: I wonder what happened that brought it back to availalbe | 15:09 |
| fxpester | http://paste.openstack.org/show/549250/ | 15:10 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Fix tempest realted exceptions during docs build https://review.openstack.org/351239 | 15:10 |
| mat128 | fxpester: this one is cleaning, maybe the deployment failed? | 15:11 |
| *** moshele has joined #openstack-ironic | 15:13 | |
| fxpester | mat128: yesterday I deployed Devstack ok, the onl things I changed resources, now its - 512 mem and 2gb HDD - is this enought for devstack VM images ? | 15:14 |
| *** e0ne has quit IRC | 15:15 | |
| mat128 | fxpester: 512mb is fine for TinyIPA | 15:15 |
| mat128 | fxpester: can you share your local.conf through paste? | 15:15 |
| sambetts | current tinyipa build will run at 256mb :) | 15:15 |
| mat128 | :D | 15:15 |
| fxpester | local.conf - http://paste.openstack.org/show/549252/ | 15:16 |
| *** ayoung has joined #openstack-ironic | 15:17 | |
| mat128 | fxpester: from what I can tell, you should be fine, tinyipa is the default (https://github.com/openstack/ironic/blob/master/devstack/lib/ironic#L138) and 512M is def. enough | 15:18 |
| JayF | as long as you're deploying the cirros images | 15:18 |
| mat128 | JayF: true.. | 15:19 |
| fxpester | yes, default devstack workflow, it was ok yesterday... X_X | 15:19 |
| fxpester | cirros. | 15:19 |
| *** baoli has joined #openstack-ironic | 15:19 | |
| mat128 | JayF: I found out qemu-img can stream images from http/https, but can't do checksumming :( | 15:19 |
| JayF | Yeah; we do stream raw images in the agent | 15:19 |
| *** rama_y has joined #openstack-ironic | 15:19 | |
| mat128 | JayF: I forgot "qcow2" in my sentence =) | 15:20 |
| JayF | I'm not sure even the added benefit is worth the lost reliability | 15:20 |
| * JayF has seen those checksum checks prevent bad images in the past, in production, multiple times | 15:20 | |
| openstackgerrit | Moshe Levi proposed openstack/ironic-inspector: Adding InfiniBand Support https://review.openstack.org/264257 | 15:21 |
| openstackgerrit | Moshe Levi proposed openstack/ironic-inspector: Dont review https://review.openstack.org/348457 | 15:21 |
| mat128 | JayF: thats why I don't want / can't go forward with that idea | 15:21 |
| mat128 | but having 20GB windows images running on 4GB machines is problematic without streaming to disk | 15:21 |
| JayF | it'd be nice to add that feature to qemu-image | 15:21 |
| mat128 | yes | 15:21 |
| sambetts | JayF: couldn't we take a chacksum on the ironic side / glance generate one and then we check it once we've written the whole image? | 15:21 |
| sambetts | mat128: ^ | 15:21 |
| mat128 | JayF: we shrunk windows images down a lot, but they still exceed the min RAM they could live with running | 15:21 |
| mat128 | sambetts: 2 issues with this | 15:22 |
| mat128 | a) the data changes (so does the checksum) | 15:22 |
| mat128 | and reading the whole disk to do a checksum will be slow | 15:22 |
| mat128 | and the checksum will be different based on disk size | 15:22 |
| JayF | and if you can't read the data into ram to do an image | 15:22 |
| JayF | can you read the data into ram to do a checksum? | 15:22 |
| JayF | you'd also end up having to write some kind of crazy streaming checksum | 15:22 |
| mat128 | you need the same amount of space | 15:22 |
| sambetts | ah, yup grok | 15:23 |
| mat128 | or you will have to download it twice | 15:23 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Fix __all__ module attributes https://review.openstack.org/351246 | 15:23 |
| mat128 | and the second download might be different | 15:23 |
| JayF | which would also be subject to hardware failures (i.e. if the disk didn't give you back all the bits identically) | 15:23 |
| mat128 | so adding that streaming checksum feature we have in python in the raw mode directly in qemu-img would be beneficial | 15:23 |
| *** baoli_ has joined #openstack-ironic | 15:23 | |
| mat128 | like, qemu-convert with an "expected checksum" | 15:23 |
| mat128 | bail out if the checksum turned out different | 15:23 |
| *** tangchen has joined #openstack-ironic | 15:23 | |
| *** tangchen_ has joined #openstack-ironic | 15:23 | |
| *** baoli has quit IRC | 15:23 | |
| mat128 | can't say my C skills are all that great, but def. a feature to add to qemu-img | 15:24 |
| * mgould thought most checksum algorithms operated on streams... | 15:24 | |
| JayF | mgould: you may be right | 15:25 |
| mat128 | as it is in our downstream backlog: "I want be able to install an OS image bigger than the amount of RAM installed" | 15:25 |
| *** catintheroof has quit IRC | 15:25 | |
| mat128 | mgould: they do | 15:25 |
| mat128 | but not the tooling around | 15:25 |
| mgould | \o/ | 15:25 |
| mgould | ah, OK | 15:25 |
| mat128 | qemu-img is unable to take it's qcow data from stdin | 15:25 |
| mat128 | it has to read a file / http stream itself | 15:25 |
| jroll | JayF: python does streaming checksums, that's how we stream raw images today | 15:25 |
| JayF | jroll: yeah; but we implemented it | 15:26 |
| jroll | I mean, kinda | 15:26 |
| JayF | jroll: I was more curious about md5sum, etc | 15:26 |
| jroll | it has it built in | 15:26 |
| lucasagomes | mat128, it needs random access to the data | 15:26 |
| JayF | jroll: like mat128 said, the tooling around it | 15:26 |
| jroll | you initialize it and call update() on the object | 15:26 |
| jroll | ah | 15:26 |
| mat128 | lucasagomes: alas, yes | 15:26 |
| mgould | JayF: the md5 algo operates on a sequence of 512-bit chunks | 15:26 |
| mat128 | best of both worlds would have been | 15:26 |
| lucasagomes | we can stream raw images but not qcow2 | 15:26 |
| *** Sylvester has quit IRC | 15:27 | |
| mat128 | md5sum that works like pv, dumps md5 once done on stderr | 15:27 |
| mat128 | and qemu-img that takes data from stdin | 15:27 |
| mat128 | ;) | 15:27 |
| lucasagomes | because to convert it with qemu-img it needs to access random locations of the file | 15:27 |
| lucasagomes | so can't be done in a stream | 15:27 |
| mat128 | lucasagomes: but qemu-convert can take an http/https URL | 15:27 |
| *** wajdi has joined #openstack-ironic | 15:27 | |
| mat128 | I guess it does range requests | 15:27 |
| mgould | https://en.wikipedia.org/wiki/MD5#Algorithm describes padding happening first, but you could do it when you reach the last chunk instead | 15:27 |
| lucasagomes | mat128, yeah that could be it | 15:27 |
| mat128 | but that would save writing the image to ram | 15:27 |
| lucasagomes | mat128, for IPA I've implemented raw images streaming | 15:28 |
| mat128 | mgould: the agent already contains a streaming md5 impl | 15:28 |
| mgould | cool | 15:28 |
| lucasagomes | and it does calculate the md5 as it goes as well | 15:28 |
| mat128 | lucasagomes: yes I was referring to this, but we only have qcow2 images | 15:28 |
| *** baoli_ has quit IRC | 15:29 | |
| lucasagomes | mat128, I see, so yeah, if qemu-convert supports we can probably use it for the qcow2 images | 15:29 |
| mat128 | lucasagomes: essentially, I would need to reimplement your checksumming directly in qemu-img, but that might be a problem if it only does random requests | 15:29 |
| mat128 | crap :( | 15:29 |
| lucasagomes | true :-/ | 15:29 |
| *** baoli has joined #openstack-ironic | 15:29 | |
| mat128 | or make qemu-convert take qcow2 data from stdin ;) | 15:30 |
| mgould | wait, what problem are we trying to solve? turn qemu-img into a Unix filter? | 15:30 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Promote agent vendor passthru to core API https://review.openstack.org/330018 | 15:30 |
| dtantsur | jroll, ^^ | 15:30 |
| mat128 | pretty much | 15:30 |
| *** mgoddard_ has joined #openstack-ironic | 15:30 | |
| dtantsur | now the IPA part | 15:31 |
| *** baoli_ has joined #openstack-ironic | 15:31 | |
| mat128 | the problem I need to solve is, as it is in our (downstream) backlog: "I want be able to install a qcow2 image bigger than the amount of RAM installed" | 15:31 |
| *** moshele has quit IRC | 15:31 | |
| dtantsur | JayF, I haven't seen your comments before posting the patch, but it seems to address them.. | 15:31 |
| *** rama_y has quit IRC | 15:31 | |
| JayF | dtantsur: I added nothing new | 15:32 |
| lucasagomes | mat128, as an workaround at least for now converting it to raw before putting in glance shouldn't be a big deal | 15:32 |
| lucasagomes | is it? | 15:32 |
| JayF | dtantsur: so you're probably good | 15:32 |
| mat128 | lucasagomes: then I download 20G instead of 5-6G for windows :( | 15:32 |
| mat128 | but I guess I could have raw copies of images | 15:32 |
| mat128 | but my end users would have to choose the right ones, so I guess I need to have support for multiple copies of the same image | 15:32 |
| lucasagomes | true yeah :-/ it penalizes the network | 15:32 |
| mat128 | referenced by the same uuid | 15:32 |
| mgould | mat128: raw.gz? | 15:33 |
| mat128 | with nova deciding which "version" of the image would best fit | 15:33 |
| mat128 | mgould: hmm | 15:33 |
| mat128 | mgould: I guess the agent could do gunzipping before running the checksum and writing to disk | 15:33 |
| mat128 | interesting | 15:33 |
| lucasagomes | yeah | 15:33 |
| *** mgoddard has quit IRC | 15:34 | |
| mgould | mat128: I was thinking of gunzipping in streaming mode, and checksumming and writing as you go | 15:34 |
| *** baoli has quit IRC | 15:34 | |
| mat128 | that could work for getting that 20G down to something reasonable | 15:35 |
| *** amotoki has joined #openstack-ironic | 15:35 | |
| mat128 | still have that "multiple copies of the same image in the same region" problem though | 15:35 |
| mgould | worth checking how big the file is after gzipping, at least | 15:35 |
| mat128 | yeah | 15:35 |
| mat128 | mgould: checking now, qcow2 is 6.2G | 15:36 |
| mat128 | downloading | 15:36 |
| mgould | another thought: if qcow2 requires random access, could you create a swap partition at the end of the disk, use that to do the conversion, then write the results starting at the beginning? | 15:37 |
| JayF | dtantsur: your tests are going to fail, at least I hope | 15:37 |
| mgould | you'd be limited to images < 1/2 the size of the disk, but that's probably more acceptable | 15:37 |
| mat128 | mgould: full disk images :( | 15:37 |
| JayF | dtantsur: it looks like you didn't add metrics config to the unit tests | 15:37 |
| mat128 | mgould: partition table | 15:37 |
| mgould | mat128: bollocks! | 15:37 |
| mat128 | but interesting | 15:37 |
| lucasagomes | mat128, you may find that interesting too https://rwmj.wordpress.com/tag/qemu-img/ | 15:38 |
| mgould | NFS-mounted swap? :-) | 15:38 |
| mgould | for Real Ultimate Performance | 15:38 |
| dtantsur | JayF, I think I've changed the tests to not verify the whole configuration dictionary | 15:39 |
| mat128 | lucasagomes: interesting | 15:39 |
| mat128 | nbd | 15:39 |
| mat128 | another idea I had was mount the remote file with something like httpfs over fuse | 15:39 |
| mat128 | but performance was abysmal | 15:39 |
| mgould | mat128: I am shocked, shocked | 15:40 |
| mgould | lucasagomes: that solves the "streaming output" problem, but not the "need random access to a file that's larger than RAM and can't touch the disk" problem | 15:41 |
| *** priteau has joined #openstack-ironic | 15:41 | |
| mat128 | mgould: maybe qemu-img can use nbd input too? | 15:42 |
| mat128 | but then nbd would need access to the qcow file | 15:42 |
| JayF | dtantsur: so we don't have any tests validating the metrics configs are being sent via the new api then? | 15:42 |
| *** sabeen has joined #openstack-ironic | 15:43 | |
| dtantsur | JayF, it's like testing one's ability to copy-paste.. I don't really see point in it. | 15:43 |
| JayF | vgadiraj: https://bugs.launchpad.net/bugs/1609573 | 15:43 |
| openstack | Launchpad bug 1609573 in Ironic "Ironic gate jobs should not pass configs through devstack-gate when possible" [Undecided,New] | 15:43 |
| JayF | dtantsur: I guess I'd ask why heartbeat timeout is tested and not metrics; they seem different | 15:44 |
| JayF | dtantsur: that's the only reason I ask | 15:44 |
| dtantsur | JayF, my plan was to ensure that some configuration is actually sent | 15:44 |
| lucasagomes | mat128, right... do you know if other formats may support converting in a stream? | 15:44 |
| lucasagomes | vmdk maybe? | 15:44 |
| lucasagomes | vdi | 15:44 |
| mat128 | lucasagomes: well, qemu-img does not take any stream (stdin) for it's convert command, but maybe nbd does | 15:45 |
| vgadiraj | JayF: thanks! | 15:45 |
| lucasagomes | mat128, it seems that "stream optimized vmdk" is a thing | 15:46 |
| dtantsur | fg | 15:46 |
| JayF | dtantsur: +2'd with that one comment we just chatted about in here. TY \o/ | 15:46 |
| JayF | dtantsur: ^z | 15:46 |
| dtantsur | thanks! | 15:46 |
| JayF | :P | 15:46 |
| dtantsur | lol, yeah | 15:46 |
| mgould | mat128: isn't NBD just a "block device over network" protocol? | 15:47 |
| mat128 | mgould: yes but maybe it supports reading from a stream? | 15:47 |
| mat128 | oh god | 15:47 |
| mat128 | thats impossible | 15:47 |
| mat128 | nvm | 15:47 |
| mat128 | it's just as if you had a fuse thing for qcow | 15:47 |
| mat128 | it would need to read the bytes when you access something | 15:47 |
| mgould | mat128: basically, yes | 15:47 |
| mat128 | making it random access | 15:47 |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Add openstack baremetal driver commands https://review.openstack.org/350050 | 15:47 |
| mgould | mat128: well, it would need space bounded by the maximum seek distance | 15:48 |
| mgould | which might be smaller than the file length | 15:48 |
| mgould | but might not, idk | 15:48 |
| mgould | depends on your access patterns :-) | 15:48 |
| mgould | brb | 15:50 |
| lucasagomes | folks can I get some eyes on https://review.openstack.org/#/c/336102/ when you have time? Thanks | 15:50 |
| *** tangchen_ has quit IRC | 15:51 | |
| *** tangchen has quit IRC | 15:51 | |
| *** uggla_ has quit IRC | 15:52 | |
| devananda | morning, all | 15:53 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-python-agent: Use new agent API if available https://review.openstack.org/334523 | 15:54 |
| dtantsur | this might be it ^^^ | 15:54 |
| dtantsur | morning devananda | 15:54 |
| lucasagomes | devananda, hi there | 15:56 |
| *** ametts has quit IRC | 15:57 | |
| openstackgerrit | Merged openstack/ironic: Add keystone policy support to Ironic https://review.openstack.org/325599 | 15:59 |
| jroll | ^^^^ wooooooo | 15:59 |
| TheJulia | excellent | 15:59 |
| jroll | so much good stuff this week | 15:59 |
| mat128 | lucasagomes: comments inline | 15:59 |
| JayF | \o/ \o/ \o/ | 15:59 |
| jroll | devananda: morning, wdyt about 6.1.0 early next week? | 16:00 |
| lucasagomes | mat128, thanks a lot! | 16:00 |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Add volume_connector table to save connector information https://review.openstack.org/200983 | 16:00 |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Introduce VolumeConnector object https://review.openstack.org/214584 | 16:00 |
| devananda | woot!! | 16:01 |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Introduce new RPCs to support volume connector operation https://review.openstack.org/214585 | 16:01 |
| devananda | yea yea! policy and keystoneauth merged! | 16:01 |
| lucasagomes | w00t | 16:01 |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Add volume_targets table to save target information https://review.openstack.org/285218 | 16:01 |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Introduce VolumeTarget object https://review.openstack.org/285219 | 16:01 |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Introduce new RPCs to support volume target operations https://review.openstack.org/285220 | 16:02 |
| *** tangchen_ has joined #openstack-ironic | 16:04 | |
| *** mgoddard_ has quit IRC | 16:04 | |
| *** tangchen has joined #openstack-ironic | 16:04 | |
| *** mgoddard has joined #openstack-ironic | 16:04 | |
| *** ametts has joined #openstack-ironic | 16:04 | |
| devananda | jroll: seems reasonable to me. lemme have a look at what else is in the pipe real quick, though | 16:07 |
| jroll | JayF: https://review.openstack.org/#/c/330018/31/ironic/conf/api.py@56 | 16:07 |
| jroll | devananda: yeah, I still need to do that as well | 16:08 |
| mat128 | jroll: I interpreted it the same was a Jay | 16:08 |
| *** baoli_ has quit IRC | 16:08 | |
| mat128 | s/was/way | 16:08 |
| jroll | I did too, until I clicked the link :P | 16:09 |
| jroll | idk, guess we could ask vsaienk0 which way he meant | 16:09 |
| mat128 | jroll: well it makes sense for devstack not to use the old location | 16:10 |
| jroll | :) | 16:10 |
| openstackgerrit | Thiago Paiva Brito proposed openstack/ironic: Add Dynamic Allocation feature for the OneView drivers https://review.openstack.org/286192 | 16:11 |
| devananda | jroll: wdyt of getting the vendor api promotion in 6.1 ? | 16:11 |
| jroll | devananda: yeah, I'd like to - my hope is to get that merged today | 16:11 |
| jroll | it has a +2 | 16:11 |
| jroll | feel free to review it :D | 16:11 |
| devananda | cool. I'll take a look. | 16:11 |
| devananda | it probably requires a change for policy now | 16:12 |
| jroll | well, no, there's no auth on those endpoints | 16:12 |
| jroll | unless we want to put policy on it with a rule of * | 16:12 |
| *** ohamada has quit IRC | 16:12 | |
| xavierr|mtg | lucasagomes: all comments done! ^^^ | 16:12 |
| lucasagomes | xavierr|mtg, awesome, give me few min and I will re-review | 16:13 |
| xavierr|mtg | lucasagomes: o/ | 16:13 |
| *** baoli has joined #openstack-ironic | 16:14 | |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Add test to ensure policy is always authorized https://review.openstack.org/350177 | 16:14 |
| dtantsur | devananda, you're right, I've totally forgot about policies... | 16:15 |
| devananda | jroll: there is a policy check on them in my (now landed) patch | 16:15 |
| *** mat128 is now known as mat128|lunch | 16:15 | |
| jroll | devananda: oh? | 16:15 |
| dtantsur | I have to run now, so if something could pick it up, that would be awesome | 16:15 |
| devananda | dtantsur: also, I believe ^ will address your (quite valid) concern about ensuring that new API endpoints are checked | 16:15 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Collect deployment logs from IPA https://review.openstack.org/336102 | 16:15 |
| lucasagomes | mat128|lunch, ^ thanks for the comments | 16:15 |
| fxpester | mat128: changed my devstack back to 1gb RAM - everything ok now, looks like CoreOS is too memory hungry | 16:15 |
| devananda | however it's failing some tests and I'm not sure why :( | 16:15 |
| mat128|lunch | fxpester: you are on CoreOS? | 16:15 |
| mat128|lunch | fxpester: it wasnt set in your local.conf and the default is TinyIPA... confused :X | 16:16 |
| dtantsur | devananda, cool, I'll check it when I come back (probably tomorrow though) | 16:16 |
| devananda | dtantsur: cool. have a good night! | 16:16 |
| fxpester | mat128: I`m on Devstack defaults, and after debug I see CoreOS before CirrOS | 16:16 |
| jroll | dtantsur: we'll talk it out, I can update if needed :) | 16:16 |
| dtantsur | g'night! | 16:16 |
| dtantsur | thanks | 16:16 |
| mat128|lunch | lucasagomes: that sample file grew :) | 16:16 |
| *** dtantsur is now known as dtantsur|afk | 16:16 | |
| mat128|lunch | lucasagomes: are you sure you ran genconfig on the right branch? | 16:16 |
| mat128|lunch | lucasagomes: woha, metrics stuff | 16:17 |
| mat128|lunch | lucasagomes: poke me when you're done fixing the rebase, I,ll check it out | 16:17 |
| mat128|lunch | lucasagomes: nvm, just my habit of checking 10..11 | 16:17 |
| lucasagomes | mat128|lunch, yeah I need to run genconfig again :-/ forgot | 16:17 |
| mat128|lunch | it looks fine | 16:18 |
| mat128|lunch | I was looking at 10..11 but what i wanted is (base..10)..(base..11) | 16:18 |
| mat128|lunch | but thats not possible | 16:18 |
| lucasagomes | oh right yeah | 16:18 |
| lucasagomes | that's gerrit | 16:18 |
| *** Romanenko_K has quit IRC | 16:19 | |
| mat128|lunch | lucasagomes: perfect, thanks for this work | 16:19 |
| mat128|lunch | lucasagomes: it will really help us out in production :) | 16:19 |
| lucasagomes | mat128|lunch, ty for reviewing it! | 16:19 |
| * mat128|lunch off to lunch for real this time =) | 16:20 | |
| mat128|lunch | np | 16:20 |
| lucasagomes | yeah, hope it does cause we have the problem of people reporting bugs without collecting logs etc... hope that helps with it | 16:20 |
| jroll | devananda: so, you have policy on vendor passthru stuff, but I believe it's skipped due to https://github.com/openstack/ironic/blob/master/ironic/api/config.py#L33 | 16:20 |
| jroll | devananda: I'm thinking we don't need it on the ramdisk endpoints | 16:20 |
| lucasagomes | xavierr|mtg, +2 | 16:20 |
| xavierr|mtg | lucasagomes: awesome, thanks again \o/ | 16:21 |
| lucasagomes | yw :-) | 16:21 |
| *** ipukha has quit IRC | 16:21 | |
| *** e0ne has joined #openstack-ironic | 16:23 | |
| devananda | jroll: it totally is enforced -- https://gist.github.com/devananda/185622799d0d9667bdb3cbcf90444a2d | 16:29 |
| jroll | devananda: not for lookup/heartbeat | 16:29 |
| devananda | oh! brain... | 16:29 |
| jroll | :) | 16:29 |
| TheJulia | thiagop: +2 applied | 16:30 |
| TheJulia | NobodyCam: https://review.openstack.org/#/c/286192 has been revised, if you want to take a quick look | 16:30 |
| xavierr|mtg | thank you TheJulia :D | 16:31 |
| *** kro_focused is now known as krotscheck | 16:32 | |
| TheJulia | xavierr|mtg: no problem :) | 16:33 |
| lucasagomes | I'm calling it a day | 16:38 |
| lucasagomes | have a great evening all! See y'all tomorrow | 16:38 |
| xavierr|mtg | lucasagomes: "what a day what a lovely day" | 16:38 |
| lucasagomes | :D | 16:38 |
| lucasagomes | see ya | 16:39 |
| *** lucasagomes is now known as lucas|dinner | 16:39 | |
| jlvillal | lucas|dinner: Ciao! | 16:39 |
| devananda | jroll: confirmed - you're correct | 16:40 |
| devananda | even changing the policy config to "baremetal:node:ipa_heartbeat": "rule:is_admin", it is still not enforced | 16:40 |
| NobodyCam | morning lucas|dinner .. have a good night.. morning jlvillal and devananda :) | 16:41 |
| anteaya | NobodyCam: hey there | 16:41 |
| NobodyCam | hey morning antenna :) | 16:41 |
| anteaya | it occurs to me that last week some time I said hello NobodyCam and asked you how you were, then disappeared | 16:41 |
| anteaya | NobodyCam: I'm sorry about that | 16:41 |
| anteaya | and good morning | 16:41 |
| jroll | devananda: yeah, I wonder if we should add a comment in the code "this doesn't have a policy because" | 16:42 |
| jlvillal | NobodyCam: Morning :) | 16:42 |
| jroll | devananda: which I can do quickly | 16:42 |
| devananda | jroll: I think it should have policy checked | 16:42 |
| devananda | and there is actually a policy in place for it - it's just skipped higher up in the evaluation | 16:43 |
| jroll | devananda: against what? there's no auth on those paths | 16:43 |
| devananda | IMO it's a bug in my policy cod | 16:43 |
| *** e0ne has quit IRC | 16:43 | |
| NobodyCam | devananda: you have a policy fish? | 16:44 |
| NobodyCam | lol ... /me *ducks* | 16:44 |
| jroll | lol | 16:44 |
| jroll | devananda: I'm confused what you're going to check against, IPA doesn't send a token | 16:44 |
| *** rbartal has quit IRC | 16:45 | |
| devananda | NobodyCam: lol :) | 16:49 |
| devananda | I totally have a policy fish! | 16:50 |
| NobodyCam | :) Nice !!!! | 16:50 |
| *** Sukhdev has joined #openstack-ironic | 16:50 | |
| devananda | jroll: but just imagine if it did ;) | 16:51 |
| JayF | That sounds like a fish we should scale when we get that feature done then :D | 16:52 |
| NobodyCam | JayF: lol.. and good morning :) | 16:53 |
| *** david-lyle has quit IRC | 16:53 | |
| JayF | although I know how much our lack of IPA auth ofFINds you | 16:53 |
| *** david-lyle has joined #openstack-ironic | 16:53 | |
| jroll | devananda: that'd be fantastic, but no way I'm blocking this patch on it :) | 16:54 |
| *** jtomasek has quit IRC | 16:55 | |
| *** jpich has quit IRC | 16:57 | |
| *** mgould is now known as mgould|afk | 16:57 | |
| *** rbartal has joined #openstack-ironic | 16:58 | |
| *** anush_ has quit IRC | 17:00 | |
| * sambetts -> home | 17:02 | |
| sambetts | night all | 17:03 |
| *** sambetts is now known as sambetts|afk | 17:03 | |
| NobodyCam | night sambetts|afk | 17:03 |
| sambetts|afk | cya NobodyCam | 17:03 |
| *** tangchen has quit IRC | 17:11 | |
| *** permalac has quit IRC | 17:11 | |
| *** PollyZ has quit IRC | 17:18 | |
| *** rama_y has joined #openstack-ironic | 17:19 | |
| *** bfournie has quit IRC | 17:21 | |
| devananda | interesting | 17:22 |
| *** mat128|lunch is now known as mat128 | 17:23 | |
| *** jcoufal_ has joined #openstack-ironic | 17:24 | |
| *** dprince has quit IRC | 17:24 | |
| *** tangchen has joined #openstack-ironic | 17:24 | |
| * devananda pokes at the agent vendor api patch for a while | 17:26 | |
| *** jcoufal has quit IRC | 17:26 | |
| *** hoangcx has quit IRC | 17:26 | |
| *** daemontool has joined #openstack-ironic | 17:28 | |
| *** bvandewa has joined #openstack-ironic | 17:34 | |
| vdrok | jroll: iirc you said that osc folks don't want projects to use one letter arguments? | 17:36 |
| vdrok | we have -l for local_link_connection now in client | 17:37 |
| *** hoangcx has joined #openstack-ironic | 17:37 | |
| jroll | vdrok: I didn't say that, someone else did, idk the rules | 17:39 |
| jroll | in ironic cli it's okay | 17:39 |
| *** bvandewa has quit IRC | 17:39 | |
| vdrok | jroll: I mean openstack baremetal port create -l switch_id=blah | 17:39 |
| jroll | oh | 17:39 |
| jroll | :{ | 17:39 |
| vdrok | we allow both -l and --local-link-connection | 17:39 |
| jroll | I have to run for now but could you file a bug? | 17:40 |
| vdrok | yup, sure | 17:40 |
| jroll | thanks | 17:40 |
| vdrok | I can fix it too I guess :) just another backwards compat issue :( | 17:40 |
| *** dprince has joined #openstack-ironic | 17:41 | |
| *** Sukhdev has quit IRC | 17:45 | |
| *** vishwanathj has quit IRC | 17:48 | |
| *** e0ne has joined #openstack-ironic | 17:53 | |
| *** tesseract- has quit IRC | 17:53 | |
| *** maurosr has quit IRC | 17:54 | |
| *** bfournie has joined #openstack-ironic | 17:56 | |
| *** bvandewa has joined #openstack-ironic | 17:56 | |
| *** joprovost1 has joined #openstack-ironic | 17:58 | |
| *** joprovost has quit IRC | 17:59 | |
| *** joprovost has joined #openstack-ironic | 18:00 | |
| *** ccamacho has quit IRC | 18:01 | |
| *** mgoddard_ has joined #openstack-ironic | 18:01 | |
| *** joprovost1 has quit IRC | 18:02 | |
| *** syed_ has quit IRC | 18:02 | |
| *** mgoddard has quit IRC | 18:04 | |
| *** mgoddard_ has quit IRC | 18:06 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Operator documentation for multitenancy https://review.openstack.org/228496 | 18:07 |
| vdrok | a super-easy change if someone has time https://review.openstack.org/351246 | 18:09 |
| *** joprovost1 has joined #openstack-ironic | 18:14 | |
| *** joprovost has quit IRC | 18:15 | |
| *** joprovost1 is now known as joprovost | 18:15 | |
| vdrok | thank you JayF :) and good night everyone! | 18:16 |
| JayF | Always a fan of an opportunity to learn a new thing :D | 18:16 |
| JayF | o/ | 18:17 |
| *** PollyZ has joined #openstack-ironic | 18:18 | |
| *** Sukhdev has joined #openstack-ironic | 18:22 | |
| *** PollyZ has quit IRC | 18:24 | |
| *** praneshp has joined #openstack-ironic | 18:24 | |
| *** hoangcx has quit IRC | 18:30 | |
| *** syed_ has joined #openstack-ironic | 18:31 | |
| *** bvandewa has quit IRC | 18:33 | |
| *** ayoung has quit IRC | 18:38 | |
| *** hoangcx has joined #openstack-ironic | 18:40 | |
| *** maurosr has joined #openstack-ironic | 18:41 | |
| *** ChubYann has joined #openstack-ironic | 18:45 | |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Add volume_connector table to save connector information https://review.openstack.org/200983 | 18:45 |
| *** milan has quit IRC | 18:46 | |
| *** Goneri has joined #openstack-ironic | 18:47 | |
| *** milan has joined #openstack-ironic | 18:47 | |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Introduce VolumeConnector object https://review.openstack.org/214584 | 18:48 |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Introduce new RPCs to support volume connector operation https://review.openstack.org/214585 | 18:48 |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Add volume_targets table to save target information https://review.openstack.org/285218 | 18:49 |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Introduce VolumeTarget object https://review.openstack.org/285219 | 18:49 |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: Introduce new RPCs to support volume target operations https://review.openstack.org/285220 | 18:49 |
| *** daemontool has quit IRC | 18:51 | |
| *** daemontool has joined #openstack-ironic | 18:52 | |
| *** joprovost has quit IRC | 19:04 | |
| *** ccamacho has joined #openstack-ironic | 19:05 | |
| *** xavierr|mtg is now known as xavierr | 19:06 | |
| *** bvandewa has joined #openstack-ironic | 19:07 | |
| *** bvandewa has quit IRC | 19:07 | |
| *** bvandewa has joined #openstack-ironic | 19:07 | |
| *** Goneri has quit IRC | 19:08 | |
| *** joprovost has joined #openstack-ironic | 19:10 | |
| *** daemontool_ has joined #openstack-ironic | 19:11 | |
| *** sdake_ has joined #openstack-ironic | 19:13 | |
| openstackgerrit | Stephanie Miller proposed openstack/ironic: [WIP] Add storage_interface DB field and object https://review.openstack.org/348005 | 19:13 |
| *** daemontool has quit IRC | 19:15 | |
| *** sdake has quit IRC | 19:16 | |
| *** vishwanathj has joined #openstack-ironic | 19:19 | |
| *** vishwanathj has quit IRC | 19:21 | |
| *** vishwanathj has joined #openstack-ironic | 19:21 | |
| *** praneshp has quit IRC | 19:28 | |
| *** vishwanathj has quit IRC | 19:31 | |
| *** vishwanathj has joined #openstack-ironic | 19:31 | |
| *** bvandewa has quit IRC | 19:32 | |
| *** vishwanathj has quit IRC | 19:33 | |
| *** vishwanathj has joined #openstack-ironic | 19:33 | |
| *** PollyZ has joined #openstack-ironic | 19:36 | |
| *** baoli has quit IRC | 19:38 | |
| *** Sukhdev has quit IRC | 19:39 | |
| *** bvandewa has joined #openstack-ironic | 19:39 | |
| *** baoli has joined #openstack-ironic | 19:42 | |
| *** jrist has quit IRC | 19:43 | |
| *** ijw has joined #openstack-ironic | 19:46 | |
| *** rama_y has quit IRC | 19:46 | |
| *** rama_y has joined #openstack-ironic | 19:46 | |
| *** M4g1c5t0rM has joined #openstack-ironic | 19:52 | |
| *** ayoung has joined #openstack-ironic | 19:56 | |
| *** jrist has joined #openstack-ironic | 19:57 | |
| *** vishwanathj has quit IRC | 19:59 | |
| *** vishwanathj has joined #openstack-ironic | 19:59 | |
| *** moshele has joined #openstack-ironic | 20:03 | |
| *** rama_y has quit IRC | 20:05 | |
| *** ametts has quit IRC | 20:05 | |
| *** rama_y has joined #openstack-ironic | 20:05 | |
| *** ametts has joined #openstack-ironic | 20:10 | |
| *** rama_y has quit IRC | 20:11 | |
| *** ijw has quit IRC | 20:11 | |
| *** rama_y has joined #openstack-ironic | 20:12 | |
| *** baoli has quit IRC | 20:13 | |
| *** baoli has joined #openstack-ironic | 20:14 | |
| *** vishwanathj has quit IRC | 20:14 | |
| *** vishwanathj has joined #openstack-ironic | 20:14 | |
| openstackgerrit | Michael Turek proposed openstack/ironic: Prevent URL collisions with sub-controllers: nodes/ports https://review.openstack.org/314514 | 20:21 |
| *** livelace has quit IRC | 20:23 | |
| *** PollyZ has quit IRC | 20:29 | |
| *** catintheroof has joined #openstack-ironic | 20:30 | |
| *** e0ne has quit IRC | 20:31 | |
| *** e0ne has joined #openstack-ironic | 20:32 | |
| NobodyCam | thiagop: you around? | 20:33 |
| thiagop | NobodyCam:yup | 20:33 |
| NobodyCam | :) just a quick question... looking at lines 203 and 290 of https://review.openstack.org/#/c/286192/31/ironic/drivers/modules/oneview/common.py and reason they don't raise OneViewInvalidNodeParameter ? | 20:35 |
| thiagop | NobodyCam: in fact, we should have removed the try-catches as we removed on power and management.py... | 20:38 |
| *** jrist has quit IRC | 20:38 | |
| *** vishwanathj has quit IRC | 20:38 | |
| NobodyCam | ahh so there will be another rev ?? :p | 20:38 |
| *** vishwanathj has joined #openstack-ironic | 20:39 | |
| thiagop | NobodyCam: we haven't noted that... Can we fix it in a follow up or do you think it's a huge blocker? | 20:40 |
| *** vishwanathj has quit IRC | 20:40 | |
| *** vishwanathj has joined #openstack-ironic | 20:41 | |
| NobodyCam | na but wanted to check with you | 20:43 |
| *** e0ne has quit IRC | 20:44 | |
| thiagop | NobodyCam: :D jenkins votes are beautiful when all-green like that | 20:45 |
| *** jrist has joined #openstack-ironic | 20:50 | |
| openstackgerrit | Thiago Paiva Brito proposed openstack/python-oneviewclient: Comply with iLO model of usage for one-time boot https://review.openstack.org/350060 | 20:53 |
| *** praneshp has joined #openstack-ironic | 20:58 | |
| *** rama_y has quit IRC | 20:59 | |
| *** praneshp has quit IRC | 21:00 | |
| *** rama_y has joined #openstack-ironic | 21:00 | |
| *** bvandewa_ has joined #openstack-ironic | 21:00 | |
| *** ijw has joined #openstack-ironic | 21:01 | |
| *** baoli_ has joined #openstack-ironic | 21:02 | |
| openstackgerrit | Merged openstack/ironic: Fix __all__ module attributes https://review.openstack.org/351246 | 21:04 |
| *** baoli has quit IRC | 21:04 | |
| *** bvandewa has quit IRC | 21:04 | |
| *** ijw has quit IRC | 21:04 | |
| *** ijw has joined #openstack-ironic | 21:04 | |
| *** vishwanathj has quit IRC | 21:05 | |
| *** vishwanathj has joined #openstack-ironic | 21:06 | |
| NobodyCam | thiagop: better start writing that follow up :p | 21:09 |
| NobodyCam | also do we need another version of https://review.openstack.org/#/c/305441? | 21:09 |
| *** vishwanathj has quit IRC | 21:11 | |
| thiagop | NobodyCam: in fact, I'm on it right now | 21:11 |
| *** vishwanathj has joined #openstack-ironic | 21:11 | |
| *** PollyZ has joined #openstack-ironic | 21:12 | |
| NobodyCam | :) | 21:12 |
| *** baoli_ has quit IRC | 21:12 | |
| *** baoli has joined #openstack-ironic | 21:12 | |
| thiagop | NobodyCam: yes, we have, it's pretty out of date. We'll start working on it on monday (tomorrow is a holiday in our state) | 21:13 |
| *** vishwanathj has quit IRC | 21:13 | |
| thiagop | NobodyCam: ^about docs | 21:13 |
| *** vishwanathj has joined #openstack-ironic | 21:13 | |
| NobodyCam | okay because now with dynamic allocation approved we'll need docs on how to use it quickly :) | 21:14 |
| thiagop | NobodyCam: jroll said that he's planning on releasing 6.1 next week. We'll do whatever we can to make these docs before that | 21:16 |
| jroll | s/planning/hoping/ | 21:16 |
| NobodyCam | :) | 21:16 |
| *** vishwanathj has quit IRC | 21:16 | |
| jroll | that said, this is why I prefer docs with code | 21:16 |
| jroll | so we don't accidentally release something without docs | 21:16 |
| NobodyCam | jroll: +++ | 21:16 |
| *** vishwanathj has joined #openstack-ironic | 21:16 | |
| thiagop | jroll: we'll make sure that won't happen. But I totally agree | 21:16 |
| *** baoli has quit IRC | 21:18 | |
| *** baoli has joined #openstack-ironic | 21:18 | |
| *** Sukhdev has joined #openstack-ironic | 21:18 | |
| *** rbrud has joined #openstack-ironic | 21:20 | |
| *** rbrud has quit IRC | 21:21 | |
| *** rbrud has joined #openstack-ironic | 21:24 | |
| *** vishwanathj has quit IRC | 21:24 | |
| *** vishwanathj has joined #openstack-ironic | 21:24 | |
| mrda | Morning Ironic | 21:25 |
| NobodyCam | good morning mrda | 21:25 |
| *** baoli has quit IRC | 21:26 | |
| *** ametts has quit IRC | 21:26 | |
| thiagop | morning mrda | 21:26 |
| *** amotoki has quit IRC | 21:27 | |
| *** priteau has quit IRC | 21:28 | |
| mrda | o/ | 21:30 |
| jlvillal | Morning mrda | 21:35 |
| *** sabeen has quit IRC | 21:35 | |
| *** vishwanathj has quit IRC | 21:37 | |
| *** ijw has quit IRC | 21:37 | |
| *** vishwanathj has joined #openstack-ironic | 21:37 | |
| *** vishwanathj has quit IRC | 21:38 | |
| *** vishwanathj has joined #openstack-ironic | 21:39 | |
| *** bvandewa has joined #openstack-ironic | 21:39 | |
| *** ijw has joined #openstack-ironic | 21:40 | |
| *** bvandewa_ has quit IRC | 21:43 | |
| *** ayoung has quit IRC | 21:43 | |
| *** ijw has quit IRC | 21:44 | |
| *** vishwanathj has quit IRC | 21:44 | |
| *** vishwanathj has joined #openstack-ironic | 21:45 | |
| *** vishwanathj has quit IRC | 21:49 | |
| *** vishwanathj has joined #openstack-ironic | 21:50 | |
| jroll | mrda: morning! | 21:51 |
| *** vishwanathj has quit IRC | 21:53 | |
| *** amotoki has joined #openstack-ironic | 21:53 | |
| *** vishwanathj has joined #openstack-ironic | 21:54 | |
| mrda | hey jroll | 21:57 |
| openstackgerrit | Thiago Paiva Brito proposed openstack/ironic: OneView driver docs explaining Dynamic Allocation https://review.openstack.org/305441 | 21:57 |
| thiagop | NobodyCam: ^not much to update in a first (and tired) glance | 21:58 |
| JayF | thiagop: reviewing now | 21:58 |
| NobodyCam | just an off the wall question anyone had a particularly good experience with any burn in software? | 21:58 |
| *** dprince has quit IRC | 21:59 | |
| JayF | I mean, you need a more clear use statement than that to get good software | 21:59 |
| JayF | like what are you trying to determine, NobodyCam | 21:59 |
| jlvillal | NobodyCam: Burn-in of what? | 21:59 |
| jlvillal | Like a brand new computer? Stress test? | 21:59 |
| * jroll prefers burn-outs | 21:59 | |
| thiagop | lol | 21:59 |
| * jlvillal has no experience with such software | 21:59 | |
| NobodyCam | new hardware readiness stuff memory, cpu, disk | 21:59 |
| NobodyCam | lol | 22:00 |
| jlvillal | NobodyCam: My only experience tends to be that memory test software has been relatively useless for me. | 22:00 |
| jlvillal | NobodyCam: Had a system that ran for two days with MemTest86+ and passed. | 22:00 |
| jlvillal | NobodyCam: Booted into Ubuntu 14.04 and would crash within 2 minutes. | 22:01 |
| NobodyCam | :/ | 22:01 |
| JayF | NobodyCam: basically three things: 1) get it hot (usually best way is to get all CPUs to 100% at the same time with the disks spinning too) | 22:01 |
| jlvillal | NobodyCam: Swapped out the DIMMs and it worked... | 22:01 |
| JayF | NobodyCam: 2) Run a memory tester multiple times | 22:01 |
| JayF | NobodyCam: 3) Run bonnie++ | 22:01 |
| jlvillal | JayF: That's a pretty program ;) | 22:01 |
| * jlvillal realizes that might be a pun hard to understand. http://www.merriam-webster.com/dictionary/bonnie | 22:02 | |
| JayF | I never have heard the term bonny used that way, neat | 22:02 |
| jlvillal | JayF: I think they use it in Scotland, but not sure. mgould|afk should know :) | 22:03 |
| NobodyCam | :) | 22:03 |
| *** PollyZ has quit IRC | 22:05 | |
| *** rbrud has quit IRC | 22:06 | |
| JayF | thiagop: made a few suggestions on that documentation patch, curious what you think | 22:11 |
| *** vishwanathj has quit IRC | 22:14 | |
| *** vishwanathj has joined #openstack-ironic | 22:15 | |
| thiagop | JayF: checking now... | 22:17 |
| *** PollyZ has joined #openstack-ironic | 22:17 | |
| *** jcoufal_ has quit IRC | 22:19 | |
| *** ayoung has joined #openstack-ironic | 22:26 | |
| *** jabroney has joined #openstack-ironic | 22:27 | |
| *** wajdi_ has joined #openstack-ironic | 22:29 | |
| openstackgerrit | Thiago Paiva Brito proposed openstack/ironic: OneView driver docs explaining Dynamic Allocation https://review.openstack.org/305441 | 22:31 |
| thiagop | JayF: tyvm! ^ | 22:31 |
| JayF | np | 22:32 |
| JayF | thiagop: btw, the non-capitalized ironic thing, I think is a little silly and it looks strange, but it's the rules so *shrug* | 22:32 |
| *** wajdi has quit IRC | 22:33 | |
| *** wajdi_ has quit IRC | 22:33 | |
| *** ijw has joined #openstack-ironic | 22:33 | |
| NobodyCam | JayF: Yea. I still see it as a proper name but. meh... :p | 22:34 |
| thiagop | JayF: understood. No problem at all. I just leave the name of the Third Party services capitalized since... well... they're ours | 22:34 |
| JayF | lol | 22:34 |
| JayF | OneView vs ironic | 22:34 |
| *** PollyZ has quit IRC | 22:34 | |
| JayF | makes us feel small, man :D | 22:34 |
| thiagop | in fact, I see a bunch of docs from HPE that says OneView, other bunch with Oneview... | 22:35 |
| * NobodyCam thinks about a BIG tent comment.. but doesn't say anything :P | 22:35 | |
| *** PollyZ has joined #openstack-ironic | 22:36 | |
| JayF | thiagop: +2 with a couple more nits, if you fix 'em real quick I'll restore my vote (or you can go do something else :D) | 22:36 |
| * thiagop suddenly realizes that should have been drinking a beer at this time at night | 22:37 | |
| thiagop | but... I'm here just for that | 22:37 |
| JayF | Advantage of PST: nobody here to keep me stuck to IRC when I should be doing !code, lol | 22:38 |
| thiagop | lol | 22:38 |
| *** ijw has quit IRC | 22:40 | |
| thiagop | JayF: @304, you mean the "Command Line Interface" part, right? | 22:40 |
| JayF | thiagop: it should be "The ironic-oneview CLI" | 22:41 |
| JayF | thiagop: and below that, "The ironic-oneviewd is a" | 22:41 |
| JayF | well, actually, that should probably be "The ironic-oneviewd daemon monitors the ..." | 22:42 |
| JayF | but it's not a big deal, these docs are great even if the grammar is a little off when they land | 22:42 |
| openstackgerrit | Merged openstack/ironic: Add Dynamic Allocation feature for the OneView drivers https://review.openstack.org/286192 | 22:43 |
| thiagop | JayF: Found a couple more "Pre-Allocation" and "Dynamic Allocation" camel cased, fixing | 22:44 |
| *** jvgrant_ has quit IRC | 22:46 | |
| openstackgerrit | Thiago Paiva Brito proposed openstack/ironic: OneView driver docs explaining Dynamic Allocation https://review.openstack.org/305441 | 22:48 |
| thiagop | JayF: done^ | 22:48 |
| JayF | NobodyCam: ^ you wanna give that a quick review and land it? | 22:48 |
| JayF | NobodyCam: and then we'll only have been missing docs for about, oh, 15 minutes | 22:49 |
| NobodyCam | looking now ;) | 22:49 |
| JayF | jroll: ^ we're about to land those oneview docs to go with the code, so fears remain unrealized :P | 22:49 |
| jroll | yay | 22:50 |
| thiagop | and /me will have a greater motive to drink tonight | 22:50 |
| NobodyCam | lol onveclient min version bumped from 2.2.0 to 2.4.0 :p | 22:52 |
| openstackgerrit | Thiago Paiva Brito proposed openstack/ironic: Follow-up 286192: refactoring exception treatment https://review.openstack.org/351422 | 22:52 |
| *** syed_ has quit IRC | 22:52 | |
| thiagop | NobodyCam: in fact, there is this too -> https://review.openstack.org/#/c/340596/ | 22:53 |
| jlvillal | thiagop: I did the, "I'm on a plane and don't have Internet access" complaint on that commit message :) | 22:53 |
| *** bvandewa has quit IRC | 22:53 | |
| jroll | we shouldn't have "follow up on $ref" for anything anyway | 22:54 |
| jroll | just say what the commit does | 22:54 |
| thiagop | 340596 is the real motive to bump to 2.4.0 since it allows for quicker boots | 22:54 |
| jroll | following up on a thing is not a change, cleaning up typos or whatever is | 22:54 |
| JayF | I -1'd for almost identical reason as jlvillal | 22:54 |
| jlvillal | JayF: Thanks. | 22:55 |
| thiagop | jlvillal lol, ok | 22:56 |
| thiagop | jroll: roger | 22:56 |
| *** ijw has joined #openstack-ironic | 22:56 | |
| NobodyCam | thiagop: +2 happy to hit +a unless others are looking at it | 22:57 |
| * jlvillal wonders if something broke python-ironicclient functional tests and then that job got made voting after the change... | 22:57 | |
| JayF | NobodyCam: gogogogo | 22:58 |
| NobodyCam | lol | 22:58 |
| *** lintan has joined #openstack-ironic | 22:58 | |
| *** syed_ has joined #openstack-ironic | 22:59 | |
| * jlvillal hopes just transient failure, but looks some more... | 22:59 | |
| NobodyCam | omg is it really "National Chocolate Chip Day" ??? | 23:00 |
| JayF | NobodyCam: IDK, but I'm going to go tell my wife it is now | 23:00 |
| jlvillal | JayF: Any idea about this? http://logs.openstack.org/73/348673/4/gate/gate-ironicclient-dsvm-functional/9e52c4c/logs/devstacklog.txt.gz#_2016-08-04_22_32_29_646 | 23:00 |
| JayF | NobodyCam: maybe I'll get some cookies out of the deal | 23:00 |
| thiagop | NobodyCam: lol, "M&M's" day | 23:00 |
| JayF | jlvillal: That's a new one on me for sure | 23:01 |
| *** ijw has quit IRC | 23:01 | |
| jroll | oh, I bet policy code broke that | 23:02 |
| jroll | it re-worked the users we use | 23:02 |
| jroll | I bet that uses our devstack plugin, but without nova :( | 23:02 |
| openstackgerrit | Merged openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/350887 | 23:02 |
| jlvillal | jroll: Ah, thanks | 23:02 |
| jlvillal | devananda: ^^^^ FYI | 23:02 |
| jroll | jlvillal: totally random guess though | 23:02 |
| jlvillal | jroll: Less random than my guess of, "I hope it is a transient issue, should I just do a recheck??" :) | 23:03 |
| JayF | jroll: I don't think devstack is using a non-admin user now | 23:03 |
| JayF | jroll: I'd be more convinced it was the keystoneauth stuff but *shrug* | 23:04 |
| jroll | JayF: no, but it's using a nova user | 23:04 |
| jroll | 2016-08-04 22:32:28.438 | ++ /opt/stack/new/ironic/devstack/lib/ironic:create_ironic_accounts:748 : get_or_add_user_project_role baremetal_admin nova service | 23:04 |
| jroll | wait, that should add it thuogh | 23:04 |
| JayF | ah, baremetal_admin != admin-admin | 23:04 |
| JayF | but if it's trying to do a thing with nova, it's only a baremetal_admin, right? | 23:04 |
| jroll | maybe that adds the admin role to the nova user | 23:05 |
| jroll | yeah I think it's trying to add baremetal_admin to nova | 23:05 |
| jroll | nova user* | 23:05 |
| jroll | but nova user doesn't exist | 23:05 |
| jroll | make sense? | 23:05 |
| jroll | probably needs a get_or_create_user, assuming that exists | 23:05 |
| *** ayoung has quit IRC | 23:06 | |
| JayF | jroll: I'm just curious as to why that would be configured differently than normal devstack runs | 23:07 |
| * JayF looks at p-c | 23:07 | |
| jroll | JayF: I assume client functional tests don't need nova and friends | 23:07 |
| jroll | they just talk to ironic | 23:07 |
| JayF | https://github.com/openstack-infra/project-config/blob/master/jenkins/jobs/ironic.yaml#L269 | 23:09 |
| JayF | ayup | 23:09 |
| JayF | undeclared dependency was undeclared | 23:09 |
| openstackgerrit | Thiago Paiva Brito proposed openstack/ironic: Replacing generic OneViewError w/ InvalidNodeParameter https://review.openstack.org/351422 | 23:09 |
| thiagop | maybe it's better now jlvillal JayF jroll ^ | 23:10 |
| jlvillal | JayF: jroll: I updated the whiteboard. https://etherpad.openstack.org/p/IronicWhiteBoard Feel free to correct any errors | 23:10 |
| JayF | thiagop: already +2'd :D | 23:10 |
| jlvillal | JayF: Do you think we need to add 'nova' to that list? | 23:11 |
| * thiagop needs to fix my VIM validations | 23:11 | |
| jroll | thanks to both of you, I'm not looking because I just spent hours on nova tests so I'm out | 23:11 |
| *** ijw has joined #openstack-ironic | 23:11 | |
| JayF | jlvillal: I think that's the heavy-handed way to fix it | 23:11 |
| jlvillal | jroll: Have a great night. Thanks for all the work on the Nova stuff to make Ironic better :) | 23:11 |
| JayF | jlvillal: probably an easier way just in our plugin | 23:12 |
| jroll | jlvillal: totes. you too :) | 23:12 |
| thiagop | jlvillal: good night | 23:12 |
| jlvillal | thiagop: good night | 23:12 |
| *** catintheroof has quit IRC | 23:13 | |
| thiagop | you guys did rock with this nova + multitenant work | 23:13 |
| JayF | jlvillal: https://github.com/openstack/ironic/blob/master/devstack/lib/ironic#L748 pretty sure if you add a get_or-create-user nova | 23:14 |
| jlvillal | JayF: So any interest in doing some pair stuff on the python-ironicclient gate issue? | 23:14 |
| JayF | er | 23:14 |
| JayF | jlvillal: https://github.com/openstack/ironic/blob/master/devstack/lib/ironic#L748 pretty sure if you add a get_or_create_user nova | 23:14 |
| JayF | here | 23:14 |
| JayF | it should fix it | 23:14 |
| JayF | but we should ensure that passes tests on both ironic and ironicclient | 23:14 |
| JayF | jlvillal: nope, because I'm already done trobuleshooting \o/ | 23:15 |
| JayF | :D | 23:15 |
| jlvillal | JayF: heh | 23:15 |
| jlvillal | JayF: You want me to do the patch? I'm happy too. | 23:15 |
| jlvillal | Or happy to review if you do it | 23:15 |
| JayF | I was offering it to you so I could still review it :) | 23:16 |
| *** ijw has quit IRC | 23:16 | |
| jlvillal | JayF: Okay. | 23:17 |
| JayF | jlvillal: you'll want a dummy change in py-ironicclient depending on your devstack fix so we can verify it fixes a thing there | 23:18 |
| *** bvandewa has joined #openstack-ironic | 23:18 | |
| jlvillal | Yep | 23:18 |
| jlvillal | JayF: I think I should use "create_service_user", as Nova does | 23:21 |
| jlvillal | Which comes from the devstack/lib/keystone file | 23:21 |
| JayF | jlvillal: no | 23:21 |
| JayF | jlvillal: if you look inside that; there's a get_or_create_user in there | 23:21 |
| jlvillal | get_or_create_user "$1" "$SERVICE_PASSWORD" "$SERVICE_DOMAIN_NAME" | 23:22 |
| jlvillal | get_or_add_user_project_role "$role" "$1" "$SERVICE_PROJECT_NAME" "$SERVICE_DOMAIN_NAME" "$SERVICE_DOMAIN_NAME" | 23:22 |
| thiagop | Folks, I'm calling it a (good) day | 23:22 |
| JayF | jlvillal: I'm afraid if we use create_service_user we're more likely to conflict with the existing one, right? | 23:22 |
| jlvillal | So you don't think we need the project role part? | 23:22 |
| thiagop | gotta get some sleep | 23:22 |
| JayF | oh, you're saying if we use create_service_user we can do it in one command? | 23:22 |
| jlvillal | thiagop: Ciao | 23:22 |
| thiagop | see you on monday | 23:22 |
| thiagop | Thanks JayF jlvillal jroll and NobodyCam | 23:22 |
| JayF | let me look again, I had that code up 1s ago | 23:22 |
| jlvillal | JayF: Well I was worried we needed both commands. | 23:22 |
| JayF | jlvillal: I think it's better to first just use get_or_create_user, if they pass, great | 23:23 |
| *** thiagop has quit IRC | 23:23 | |
| JayF | jlvillal: if not, then do the heavier thing | 23:23 |
| jlvillal | JayF: Okay, I'm willing to try that. | 23:23 |
| JayF | jlvillal: I'd rather our tests be using the least privledge possible so we know if something changes | 23:23 |
| JayF | jlvillal: but honestly I don't know much about the semantics of the "service" role | 23:24 |
| JayF | so that's just sorta, general ops-patterning | 23:24 |
| jlvillal | JayF: me either | 23:24 |
| JayF | so I think do less; if in the morning the jobs didn't pass, see how they failed and probably do the create_service_user | 23:24 |
| openstackgerrit | John L. Villalovos proposed openstack/ironic: Create the 'nova' user if not existing https://review.openstack.org/351430 | 23:27 |
| *** amotoki has quit IRC | 23:28 | |
| *** PollyZ has quit IRC | 23:28 | |
| JayF | jlvillal: quick -1, we shoudl file a bug + add it there | 23:29 |
| *** ijw has joined #openstack-ironic | 23:31 | |
| jlvillal | JayF: Yeah, you are right. I was being lazy ;) | 23:31 |
| * JayF fun police, wee-ooo-wee-oo | 23:31 | |
| jlvillal | JayF: Doing it now | 23:31 |
| *** bvandewa_ has joined #openstack-ironic | 23:32 | |
| openstackgerrit | John L. Villalovos proposed openstack/ironic: Create the 'nova' user if not existing https://review.openstack.org/351430 | 23:35 |
| *** bvandewa has quit IRC | 23:36 | |
| JayF | jlvillal: cool; if you don't mind linking the dummy-intermediate-change in the comments there, I'll make sure to look first thing and +2 if it fixes | 23:37 |
| *** hoangcx2 has joined #openstack-ironic | 23:39 | |
| jlvillal | JayF: working on dummy change now | 23:39 |
| *** hoangcx has quit IRC | 23:40 | |
| openstackgerrit | John L. Villalovos proposed openstack/python-ironicclient: NOMERGE: Testing fix for functional testing https://review.openstack.org/351432 | 23:42 |
| jlvillal | JayF: Patch pushed and comment added | 23:42 |
| JayF | ty | 23:42 |
| JayF | we'll see in the 'mornin | 23:42 |
| jlvillal | JayF: Yep. Hopefully EU or PRC will have it all done before then :) | 23:43 |
| *** sdake has joined #openstack-ironic | 23:44 | |
| *** catintheroof has joined #openstack-ironic | 23:46 | |
| *** sdake_ has quit IRC | 23:46 | |
| *** vishwanathj has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!