| *** baoli has quit IRC | 00:02 | |
| *** naohirot has joined #openstack-ironic | 00:04 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic-python-agent: Add reno for release notes management https://review.openstack.org/258727 | 00:04 |
|---|---|---|
| JayF | reNOYEAH | 00:05 |
| NobodyCam | lol | 00:08 |
| *** harlowja has joined #openstack-ironic | 00:09 | |
| *** ChubYann has quit IRC | 00:09 | |
| openstackgerrit | John L. Villalovos proposed openstack/ironic: genconfig: Debug info for unknown config types https://review.openstack.org/253974 | 00:10 |
| *** penick has joined #openstack-ironic | 00:12 | |
| *** Pespa has joined #openstack-ironic | 00:16 | |
| *** aginwala has quit IRC | 00:22 | |
| *** ChubYann has joined #openstack-ironic | 00:22 | |
| *** aginwala has joined #openstack-ironic | 00:29 | |
| *** dims has quit IRC | 00:29 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic-python-agent: Catch up release notes for Mitaka https://review.openstack.org/258732 | 00:34 |
| jroll | BAM | 00:34 |
| jroll | someone please land at least 258727 so I can backport it to liberty so I can make the jobs for this | 00:35 |
| *** kkoski has quit IRC | 00:38 | |
| NobodyCam | jroll: you really want -W here: https://review.openstack.org/#/c/258727/1/tox.ini | 00:40 |
| NobodyCam | -W Turn warnings into errors. This means that the build stops at the first warning and sphinx-build exits with exit status 1. | 00:41 |
| jroll | NobodyCam: yep, that's what we do in ironic and inspector | 00:44 |
| jroll | once we have the jobs up, that won't be hard to keep track of | 00:44 |
| jroll | and keep us from stupid rendering bugs, hpefully | 00:44 |
| NobodyCam | :) yep. I ask because I was also looking at https://review.openstack.org/#/c/258576/1/tox.ini | 00:45 |
| NobodyCam | :) | 00:45 |
| jroll | NobodyCam: oh, heh | 00:47 |
| jroll | \o/ for bifrost getting reno | 00:47 |
| NobodyCam | lol | 00:48 |
| NobodyCam | two jobs left on 258727 | 00:48 |
| NobodyCam | jroll: BAM | 00:50 |
| NobodyCam | like the .plceholder :) | 00:50 |
| *** rcernin has quit IRC | 00:50 | |
| jroll | NobodyCam: nice, thanks :) | 00:50 |
| * jroll will be backporting that when it lands | 00:50 | |
| NobodyCam | and with that /me run to target... | 00:50 |
| NobodyCam | nice ;) | 00:51 |
| jroll | cool, have a good night! | 00:51 |
| *** alexpilotti has joined #openstack-ironic | 00:52 | |
| *** alexpilotti has quit IRC | 00:56 | |
| *** chlong has joined #openstack-ironic | 01:07 | |
| *** priteau has joined #openstack-ironic | 01:13 | |
| *** priteau has quit IRC | 01:18 | |
| * jroll out for the night, see y'all tomorrow | 01:20 | |
| *** zsmithnyc has quit IRC | 01:27 | |
| *** zsmithnyc has joined #openstack-ironic | 01:27 | |
| *** HimanshuG has joined #openstack-ironic | 01:30 | |
| *** mbound_ has quit IRC | 01:31 | |
| *** HimanshuG has left #openstack-ironic | 01:33 | |
| *** zhenguo has joined #openstack-ironic | 01:38 | |
| *** rvasilets___ has quit IRC | 01:45 | |
| *** Pespa has quit IRC | 01:54 | |
| *** changzhi has joined #openstack-ironic | 01:59 | |
| *** mtanino has joined #openstack-ironic | 01:59 | |
| *** amotoki has joined #openstack-ironic | 02:03 | |
| *** amotoki_ has joined #openstack-ironic | 02:05 | |
| *** praneshp has quit IRC | 02:08 | |
| *** praneshp_ has joined #openstack-ironic | 02:08 | |
| *** amotoki has quit IRC | 02:09 | |
| *** chlong has quit IRC | 02:10 | |
| *** baoli has joined #openstack-ironic | 02:20 | |
| *** baoli has quit IRC | 02:22 | |
| *** chlong has joined #openstack-ironic | 02:22 | |
| *** baoli has joined #openstack-ironic | 02:22 | |
| *** hoangcx has joined #openstack-ironic | 02:27 | |
| *** penick has quit IRC | 02:34 | |
| *** rloo has quit IRC | 02:35 | |
| *** baoli has quit IRC | 02:35 | |
| *** baoli has joined #openstack-ironic | 02:36 | |
| *** praneshp_ has quit IRC | 02:40 | |
| openstackgerrit | Yuiko Takada proposed openstack/ironic: [WIP]Add APIs for State Machine https://review.openstack.org/258764 | 02:42 |
| *** kkoski has joined #openstack-ironic | 02:44 | |
| *** aginwala_ has joined #openstack-ironic | 02:45 | |
| openstackgerrit | Haomeng,Wang proposed openstack/python-ironicclient: support multi-nodes input for node-set-provision-state https://review.openstack.org/242301 | 02:46 |
| *** aginwala has quit IRC | 02:48 | |
| *** hoangcx has left #openstack-ironic | 02:48 | |
| *** praneshp has joined #openstack-ironic | 02:49 | |
| openstackgerrit | Yuiko Takada proposed openstack/ironic: [WIP]Add APIs for State Machine https://review.openstack.org/258764 | 02:53 |
| zer0c00l | How does nova/ironic populate instance_info.root_gb? Does it populate that field from flavor.disk? | 02:54 |
| Haomeng | zer0c00l: yes, it is populated by nova I think | 02:55 |
| Haomeng | zer0c00l: we just copy the info from nova instance | 02:56 |
| *** amotoki_ has quit IRC | 02:57 | |
| *** harlowja has quit IRC | 02:58 | |
| *** baoli has quit IRC | 03:00 | |
| *** priteau has joined #openstack-ironic | 03:13 | |
| *** alexpilotti has joined #openstack-ironic | 03:14 | |
| *** yonglihe has joined #openstack-ironic | 03:14 | |
| *** priteau has quit IRC | 03:17 | |
| *** alexpilotti has quit IRC | 03:18 | |
| openstackgerrit | Merged openstack/ironic: Add grenade plugin https://review.openstack.org/256174 | 03:19 |
| *** jamielennox is now known as jamielennox|away | 03:26 | |
| *** links has joined #openstack-ironic | 03:28 | |
| *** jamielennox|away is now known as jamielennox | 03:31 | |
| openstackgerrit | Shuquan Huang proposed openstack/ironic-python-agent: Replace assertEqual(None, *) with assertIsNone in tests https://review.openstack.org/258778 | 03:32 |
| *** Marga__ has quit IRC | 03:37 | |
| *** Marga_ has joined #openstack-ironic | 03:37 | |
| *** HimanshuGarg has left #openstack-ironic | 03:38 | |
| *** aginwala_ has quit IRC | 03:44 | |
| *** logan- has joined #openstack-ironic | 03:47 | |
| *** amotoki has joined #openstack-ironic | 03:47 | |
| *** amotoki has quit IRC | 03:50 | |
| *** amotoki has joined #openstack-ironic | 03:55 | |
| *** praneshp has quit IRC | 03:56 | |
| *** amotoki_ has joined #openstack-ironic | 04:02 | |
| *** amotoki has quit IRC | 04:05 | |
| *** praneshp has joined #openstack-ironic | 04:07 | |
| *** harshs has joined #openstack-ironic | 04:09 | |
| *** kkoski has quit IRC | 04:19 | |
| *** praneshp has quit IRC | 04:20 | |
| *** harshs_ has joined #openstack-ironic | 04:29 | |
| *** harshs has quit IRC | 04:31 | |
| *** harshs_ is now known as harshs | 04:31 | |
| *** alexpilotti has joined #openstack-ironic | 04:35 | |
| *** alexpilotti has quit IRC | 04:40 | |
| *** Nakato has quit IRC | 04:51 | |
| *** Nakato has joined #openstack-ironic | 04:51 | |
| openstackgerrit | Merged openstack/ironic: Keep the console logs for all boots https://review.openstack.org/257987 | 04:53 |
| *** kirankv has joined #openstack-ironic | 04:53 | |
| *** boris-42_ has quit IRC | 04:53 | |
| *** mtreinish has quit IRC | 05:08 | |
| *** garthb has joined #openstack-ironic | 05:12 | |
| *** harshs has quit IRC | 05:13 | |
| *** harshs_ has joined #openstack-ironic | 05:13 | |
| *** mtreinish has joined #openstack-ironic | 05:17 | |
| *** mtanino has quit IRC | 05:30 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic-specs: Enhance Power Interface for Soft Power Off and Inject NMI https://review.openstack.org/186700 | 05:31 |
| *** yonglihe has quit IRC | 05:31 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic-specs: iRMC OOB rescue mode support https://review.openstack.org/250648 | 05:40 |
| *** harshs has joined #openstack-ironic | 05:50 | |
| *** harshs_ has quit IRC | 05:51 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic-specs: Add 'abort' support for Soft Power Off and Inject NMI https://review.openstack.org/250783 | 05:54 |
| *** aginwala has joined #openstack-ironic | 06:00 | |
| *** aginwala has quit IRC | 06:05 | |
| *** hoangcx has joined #openstack-ironic | 06:05 | |
| *** hoangcx has left #openstack-ironic | 06:06 | |
| *** davidlenwell has quit IRC | 06:12 | |
| *** davidlenwell has joined #openstack-ironic | 06:14 | |
| *** praneshp has joined #openstack-ironic | 06:21 | |
| openstackgerrit | Yuiko Takada proposed openstack/ironic: Migrate Tempest tests into Ironic tree https://review.openstack.org/253982 | 06:23 |
| *** yuikotakada has joined #openstack-ironic | 06:27 | |
| zer0c00l | Haomeng: thanks | 06:40 |
| zer0c00l | Haomeng: i know it gets copied into instance_info by nova | 06:40 |
| zer0c00l | but where does it copy it from? | 06:40 |
| zer0c00l | Haomeng: https://github.com/openstack/nova/blob/master/nova/virt/ironic/patcher.py#L58 | 06:42 |
| zer0c00l | it gets it from instance.root_gb | 06:43 |
| zer0c00l | but i am curious about instance.root_gb | 06:54 |
| zer0c00l | how it is populated | 06:55 |
| Haomeng | zer0c00l: it should be from ironic hypervisor I£ô£ô£ô£ô£ô | 06:56 |
| Haomeng | zer0c00l: it should be from ironic hypervisor I think | 06:57 |
| *** harshs_ has joined #openstack-ironic | 06:59 | |
| *** harshs has quit IRC | 07:00 | |
| *** harshs_ is now known as harshs | 07:00 | |
| Haomeng | zer0c00l: or from flavor info | 07:06 |
| *** alexpilotti has joined #openstack-ironic | 07:12 | |
| *** jaybeale has joined #openstack-ironic | 07:15 | |
| *** alexpilotti has quit IRC | 07:16 | |
| *** harshs has quit IRC | 07:20 | |
| *** harshs has joined #openstack-ironic | 07:24 | |
| *** harshs has quit IRC | 07:24 | |
| *** harshs has joined #openstack-ironic | 07:27 | |
| *** garthb has quit IRC | 07:31 | |
| *** ukalifon has joined #openstack-ironic | 07:33 | |
| *** harshs has quit IRC | 07:35 | |
| *** alexpilotti has joined #openstack-ironic | 07:36 | |
| *** chlong has quit IRC | 07:36 | |
| *** alexpilotti has quit IRC | 07:40 | |
| *** stendulker has joined #openstack-ironic | 07:41 | |
| *** deray has joined #openstack-ironic | 07:44 | |
| deray | hello ironicers!! | 07:45 |
| *** jaybeale has quit IRC | 07:48 | |
| *** hanchao has joined #openstack-ironic | 07:55 | |
| *** praneshp has quit IRC | 07:55 | |
| *** harshs has joined #openstack-ironic | 08:04 | |
| *** jcoufal has joined #openstack-ironic | 08:04 | |
| *** rvasilets___ has joined #openstack-ironic | 08:05 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Add Link-Local-Connection info to ironic port https://review.openstack.org/256365 | 08:11 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: refactor ironic enroll-node code https://review.openstack.org/256364 | 08:11 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Add configure_provision_network function https://review.openstack.org/256367 | 08:11 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Update Ironic VM network connection https://review.openstack.org/256366 | 08:11 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Add Ironic/Neutron integration documentation https://review.openstack.org/258596 | 08:11 |
| *** harshs_ has joined #openstack-ironic | 08:19 | |
| *** harshs has quit IRC | 08:19 | |
| *** harshs_ is now known as harshs | 08:19 | |
| *** ionutbalutoiu has joined #openstack-ironic | 08:23 | |
| *** e0ne has joined #openstack-ironic | 08:23 | |
| *** rcernin has joined #openstack-ironic | 08:26 | |
| *** ifarkas has joined #openstack-ironic | 08:29 | |
| *** ionutbalutoiu has quit IRC | 08:33 | |
| *** MattMan has quit IRC | 08:33 | |
| *** MattMan has joined #openstack-ironic | 08:33 | |
| deray | While running the unit test cases with tox, I am facing this: ImportError: No module named ironic_lib | 08:35 |
| deray | Shoudln't it be faked/mocked as a third party library | 08:35 |
| *** alexpilotti has joined #openstack-ironic | 08:36 | |
| deray | while running the unit tests? | 08:36 |
| deray | Or am i missing something while running the tox targets? | 08:36 |
| *** e0ne has quit IRC | 08:38 | |
| *** e0ne has joined #openstack-ironic | 08:40 | |
| *** alexpilotti has quit IRC | 08:40 | |
| *** rcernin has quit IRC | 08:40 | |
| *** rcernin has joined #openstack-ironic | 08:41 | |
| *** harshs has quit IRC | 08:41 | |
| *** Marga_ has quit IRC | 08:45 | |
| Haomeng | deray: hi, did you follow this guide - http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html | 08:45 |
| Haomeng | deray: run tox -evenv --notest | 08:45 |
| Haomeng | deray: and source .tox/venv/bin/activate | 08:46 |
| *** harshs has joined #openstack-ironic | 08:46 | |
| deray | Haomeng, thx. will go thru it. I have been running just tox from ironic folder (instead of creating venv and sourcing activate). Is this a new development that we can't run just tox anymore? | 08:48 |
| *** priteau has joined #openstack-ironic | 08:51 | |
| *** mkovacik has joined #openstack-ironic | 08:52 | |
| *** e0ne has quit IRC | 08:53 | |
| openstackgerrit | Yuiko Takada proposed openstack/ironic: [WIP]Add APIs for State Machine https://review.openstack.org/258764 | 08:55 |
| *** patrickeast has quit IRC | 08:56 | |
| *** alexpilotti has joined #openstack-ironic | 08:58 | |
| *** pas-ha has joined #openstack-ironic | 09:05 | |
| *** patrickeast has joined #openstack-ironic | 09:06 | |
| *** amotoki_ has quit IRC | 09:06 | |
| *** derekh has joined #openstack-ironic | 09:09 | |
| *** shuquan has joined #openstack-ironic | 09:09 | |
| *** jistr has joined #openstack-ironic | 09:09 | |
| *** ndipanov has joined #openstack-ironic | 09:09 | |
| pas-ha | morning all | 09:10 |
| *** shuquan has quit IRC | 09:11 | |
| *** harshs_ has joined #openstack-ironic | 09:12 | |
| *** harshs has quit IRC | 09:14 | |
| *** harshs_ is now known as harshs | 09:14 | |
| *** patrickeast has quit IRC | 09:15 | |
| *** shuquan has joined #openstack-ironic | 09:15 | |
| *** priteau has quit IRC | 09:15 | |
| TheJulia | Good morning | 09:16 |
| *** betherly has quit IRC | 09:19 | |
| Haomeng | TheJulia: morning:) | 09:19 |
| *** priteau has joined #openstack-ironic | 09:19 | |
| Haomeng | pas-ha: morning:) | 09:19 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Initial support for Reno https://review.openstack.org/258576 | 09:21 |
| *** zigo has quit IRC | 09:25 | |
| *** zigo_ has joined #openstack-ironic | 09:26 | |
| *** shuquan has quit IRC | 09:28 | |
| *** mkovacik has quit IRC | 09:30 | |
| *** amotoki has joined #openstack-ironic | 09:33 | |
| *** romcheg1 has quit IRC | 09:35 | |
| *** betherly has joined #openstack-ironic | 09:38 | |
| *** patrickeast has joined #openstack-ironic | 09:39 | |
| *** harshs has quit IRC | 09:39 | |
| *** shuquan has joined #openstack-ironic | 09:40 | |
| *** harshs has joined #openstack-ironic | 09:43 | |
| *** harshs has quit IRC | 09:44 | |
| *** amotoki has quit IRC | 09:44 | |
| *** mbound_ has joined #openstack-ironic | 09:45 | |
| *** openstackgerrit has quit IRC | 09:47 | |
| *** openstackgerrit has joined #openstack-ironic | 09:48 | |
| *** athomas has joined #openstack-ironic | 09:50 | |
| openstackgerrit | sonu proposed openstack/python-ironic-inspector-client: Removes MANIFEST.in as it is not needed explicitely by PBR https://review.openstack.org/258908 | 09:51 |
| openstackgerrit | Ricardo Carrillo Cruz proposed openstack/bifrost: Allow configuring provisioning IP address when using inventory_dhcp https://review.openstack.org/258909 | 09:51 |
| openstackgerrit | Haomeng,Wang proposed openstack/ironic: adds node name regex filter api support https://review.openstack.org/250112 | 09:52 |
| *** e0ne has joined #openstack-ironic | 09:55 | |
| *** naohirot has quit IRC | 09:57 | |
| *** shuquan has quit IRC | 09:57 | |
| *** shuquan has joined #openstack-ironic | 09:58 | |
| openstackgerrit | Ricardo Carrillo Cruz proposed openstack/bifrost: Allow configuring provisioning IP address when using inventory_dhcp https://review.openstack.org/258909 | 10:00 |
| *** yuikotakada has quit IRC | 10:01 | |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add support for console ports allocation/de-allocation https://review.openstack.org/258915 | 10:04 |
| *** e0ne has quit IRC | 10:05 | |
| *** mbound_ has quit IRC | 10:06 | |
| *** masterbound has joined #openstack-ironic | 10:06 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add bifrost version look-up https://review.openstack.org/258919 | 10:09 |
| *** mbound_ has joined #openstack-ironic | 10:09 | |
| *** masterbound has quit IRC | 10:09 | |
| openstackgerrit | hang cheng proposed openstack/ironic: Modify recoveting to recovering in the comment https://review.openstack.org/258920 | 10:10 |
| hanchao | hello guys, if the node registered in the ironic is in a state of "deploy failed", how can i delete it? | 10:12 |
| *** electrofelix has joined #openstack-ironic | 10:13 | |
| *** shuquan has quit IRC | 10:15 | |
| hanchao | seems i can set the provision state into 'deleted', then i can delete it afterwards. | 10:16 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Initial support for Reno https://review.openstack.org/258576 | 10:16 |
| *** zer0c00l has quit IRC | 10:19 | |
| *** shakamunyi has quit IRC | 10:20 | |
| *** shakamunyi has joined #openstack-ironic | 10:21 | |
| *** superflyy has quit IRC | 10:21 | |
| *** barra204 has joined #openstack-ironic | 10:21 | |
| TheJulia | hanchao: that is exactly what you need to do | 10:21 |
| *** zer0c00l has joined #openstack-ironic | 10:25 | |
| hanchao | TheJulia: :) | 10:25 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Console ports allocation management https://review.openstack.org/249876 | 10:39 |
| openstackgerrit | Arun S A G proposed openstack/ironic-specs: Add spec for deploying OS tarballs https://review.openstack.org/248968 | 10:40 |
| *** romcheg1 has joined #openstack-ironic | 10:43 | |
| *** e0ne has joined #openstack-ironic | 10:43 | |
| *** yolanda has quit IRC | 10:46 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Add spec for deploying OS tarballs https://review.openstack.org/248968 | 10:47 |
| openstackgerrit | Haomeng,Wang proposed openstack/ironic: Fix typo in ironic/conductor/manager.py https://review.openstack.org/258920 | 10:48 |
| *** mbound_ has quit IRC | 10:50 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Mechanism for skipping a clean step on per-node basis https://review.openstack.org/238858 | 10:51 |
| *** yolanda has joined #openstack-ironic | 10:51 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Support multiple deploy mechanisms simultaneously https://review.openstack.org/209821 | 10:53 |
| *** mgould has joined #openstack-ironic | 10:54 | |
| rcarrillocruz | TheJulia: you mean defaulting provisioning_ipv4_address to ipv4_address if not defined in inventory, right? | 10:54 |
| mgould | morning Ironic! | 10:54 |
| rcarrillocruz | (heya folks) | 10:55 |
| *** mbound_ has joined #openstack-ironic | 10:55 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: EFI iSCSI boot for ProLiant hardware https://review.openstack.org/207337 | 10:55 |
| *** mbound_ has quit IRC | 10:59 | |
| TheJulia | rcarrillocruz: yeah | 11:00 |
| rcarrillocruz | ok | 11:00 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Add support for hardware composition https://review.openstack.org/253452 | 11:00 |
| rcarrillocruz | so... you think the logic should be reversed? i.e. static ip addressing should be default | 11:01 |
| TheJulia | seems like a logical thing to do, mainly so we reduce the risk of somebody not defining it | 11:01 |
| rcarrillocruz | i can do that | 11:01 |
| rcarrillocruz | thx | 11:01 |
| rcarrillocruz | the template, yeah, copy-pasta from test machine to my dev machine | 11:01 |
| rcarrillocruz | lacking if | 11:01 |
| TheJulia | So static ip right now is default, but if dhcp is enabled, I would see this being the default | 11:01 |
| TheJulia | at least, use of dhcp for the node | 11:01 |
| TheJulia | Heh, it happens :) | 11:02 |
| TheJulia | no problem, I'm going to try and get some more sleep, be back in a few hours | 11:02 |
| rcarrillocruz | ttyl :-) | 11:02 |
| *** dims has joined #openstack-ironic | 11:07 | |
| *** lucas-dinner is now known as lucasagomes | 11:07 | |
| lucasagomes | morning all | 11:07 |
| sambetts-afk | o;/ lucasagomes | 11:07 |
| *** sambetts-afk is now known as sambetts | 11:08 | |
| *** mkovacik has joined #openstack-ironic | 11:09 | |
| *** Haomeng has quit IRC | 11:10 | |
| *** Haomeng has joined #openstack-ironic | 11:10 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Intel Node Manager policies support https://review.openstack.org/214657 | 11:12 |
| *** mbound_ has joined #openstack-ironic | 11:12 | |
| *** deray has quit IRC | 11:13 | |
| *** changzhi has quit IRC | 11:13 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Add hardware properties API to management interface https://review.openstack.org/209612 | 11:13 |
| sambetts | lucasagomes: where is the code for the ipa jobs for ironic, I'd like to have a look for automating my own driver testing | 11:17 |
| sambetts | or pxe jobs | 11:18 |
| sambetts | either really | 11:18 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Add power and thermal limits support https://review.openstack.org/251957 | 11:18 |
| *** stendulker has quit IRC | 11:19 | |
| lucasagomes | sambetts, clena steps? | 11:19 |
| lucasagomes | clean* | 11:19 |
| lucasagomes | or gate jobs? | 11:20 |
| lucasagomes | gate jobs are https://github.com/openstack-infra/project-config/ | 11:20 |
| sambetts | the automated testing | 11:20 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Intel Node Manager policies support https://review.openstack.org/214657 | 11:20 |
| sambetts | I can't find the actual steps, I've found the job template :/ do you know where in project-config that stuff is? | 11:21 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Add power and thermal limits support https://review.openstack.org/251957 | 11:21 |
| sambetts | lucasagomes: or is it in tempest? | 11:23 |
| lucasagomes | sambetts, could be as well... check openstack/tempest | 11:24 |
| sambetts | hmm :/ | 11:25 |
| lucasagomes | but tempest is most API related | 11:25 |
| lucasagomes | sambetts, you are talking about the pxe_ipa, agent_ssh etc jobs? | 11:25 |
| sambetts | yeah | 11:25 |
| lucasagomes | sambetts, those are devstack + project config | 11:25 |
| *** mbound_ has quit IRC | 11:25 | |
| lucasagomes | (and tempest too, /me thinks) | 11:27 |
| vdrok | sambetts, https://github.com/openstack-infra/project-config/blob/master/jenkins/jobs/devstack-gate.yaml#L1108 | 11:28 |
| vdrok | here is where regex passed to tempest is set | 11:28 |
| vdrok | and good morning! | 11:28 |
| lucasagomes | vdrok, morning | 11:28 |
| sambetts | o/ vdrok, is it temepest that runs all the nova boot ... wait for state... etc code? | 11:29 |
| vdrok | sambetts, yep | 11:29 |
| *** pas-ha has quit IRC | 11:29 | |
| vdrok | sambetts, it is this one - https://github.com/openstack/tempest/blob/master/tempest/scenario/test_baremetal_basic_ops.py | 11:29 |
| sambetts | ah thanks! | 11:29 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Use cached copy of hash ring in api service https://review.openstack.org/250692 | 11:31 |
| *** alexpilotti has quit IRC | 11:38 | |
| *** betherly has quit IRC | 11:39 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Baremetal provisioning in UEFI secure boot mode https://review.openstack.org/230274 | 11:40 |
| openstackgerrit | Satoru Moriya proposed openstack/ironic-specs: Volume connection information for Ironic nodes https://review.openstack.org/200496 | 11:41 |
| *** patrickeast has quit IRC | 11:42 | |
| *** betherly has joined #openstack-ironic | 11:44 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Send AMT sensors data to Ceilometer https://review.openstack.org/212450 | 11:46 |
| *** Pespa has joined #openstack-ironic | 11:48 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic-specs: Add console log support https://review.openstack.org/164586 | 11:51 |
| *** betherly has quit IRC | 11:53 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Use the apache-ironic.template from our tree https://review.openstack.org/258967 | 11:54 |
| *** rcernin has quit IRC | 12:00 | |
| *** shuquan has joined #openstack-ironic | 12:05 | |
| *** patrickeast has joined #openstack-ironic | 12:05 | |
| *** caiobo has joined #openstack-ironic | 12:09 | |
| *** shuquan has quit IRC | 12:11 | |
| *** rcernin has joined #openstack-ironic | 12:15 | |
| *** patrickeast has quit IRC | 12:15 | |
| *** raildo-afk is now known as raildo | 12:16 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for providing custom swift endpoint urls https://review.openstack.org/251277 | 12:20 |
| *** shuquan has joined #openstack-ironic | 12:21 | |
| *** dprince has joined #openstack-ironic | 12:22 | |
| sinval | morning Ironic | 12:24 |
| vsaienko | good morning | 12:25 |
| *** rvasilets___ has quit IRC | 12:25 | |
| *** jaypipes has joined #openstack-ironic | 12:26 | |
| *** patrickeast has joined #openstack-ironic | 12:26 | |
| *** shuquan has quit IRC | 12:27 | |
| *** mbound_ has joined #openstack-ironic | 12:29 | |
| openstackgerrit | Ricardo Carrillo Cruz proposed openstack/bifrost: Allow configuring provisioning IP address when using inventory_dhcp https://review.openstack.org/258909 | 12:34 |
| lucasagomes | sinval, vsaienko morning | 12:34 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Devstack: Use [deploy] erase_devices_priority config option https://review.openstack.org/258982 | 12:34 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Devstack: Use [deploy] erase_devices_priority config option https://review.openstack.org/258982 | 12:37 |
| *** Marga_ has joined #openstack-ironic | 12:39 | |
| sambetts | lucasagomes: you still trying to work out whats crashing the gate? | 12:39 |
| jlvillal | Good morning Ironic | 12:41 |
| *** betherly has joined #openstack-ironic | 12:41 | |
| * jlvillal wonders if anyone else could not vote for the 'O' release name... | 12:41 | |
| lucasagomes | sambetts, now we have the console logs, so I'm waiting for some data | 12:42 |
| lucasagomes | sambetts, but yeah, partially on it | 12:42 |
| *** amotoki has joined #openstack-ironic | 12:43 | |
| *** amotoki has quit IRC | 12:46 | |
| *** s-moriya has quit IRC | 12:47 | |
| *** kalpase has joined #openstack-ironic | 12:47 | |
| vdrok | morning jlvillal and sinval | 12:49 |
| sinval | morning jlvillal | 12:52 |
| sinval | vdrok, lucasagomes o/ | 12:52 |
| *** baoli_ has joined #openstack-ironic | 12:54 | |
| sambetts | jlvillal: I managed to vote | 12:55 |
| *** Marga_ has quit IRC | 12:56 | |
| openstackgerrit | Shuquan Huang proposed openstack/python-ironicclient: Replace assertEqual(None, *) with assertIsNone in tests https://review.openstack.org/258988 | 12:57 |
| *** links has quit IRC | 12:58 | |
| openstackgerrit | Anton Arefiev proposed openstack/ironic: Remove downgrade support from migrations https://review.openstack.org/258991 | 13:03 |
| *** hanchao has quit IRC | 13:06 | |
| lucasagomes | I'm still not very used to the new gerrit interface | 13:14 |
| lucasagomes | I think I find the old one cleaner | 13:15 |
| lucasagomes | bigger fonts and so on | 13:15 |
| * lucasagomes check settings | 13:15 | |
| *** rcernin has quit IRC | 13:21 | |
| sambetts | lucasagomes: I agree, things seem a bit scattered around the page | 13:24 |
| lucasagomes | yeah | 13:24 |
| lucasagomes | I did use the zoom in the browser, but it's not ideal | 13:25 |
| * lucasagomes is using firefox | 13:25 | |
| sambetts | I like the new my reviews dashboard with the status, and size etc | 13:26 |
| lucasagomes | and ctrl+f doesn't open a search in the browser with gerrit | 13:27 |
| lucasagomes | oh dear | 13:27 |
| lucasagomes | it opens a shitty javascript regex | 13:27 |
| sambetts | oooh... wtf... thats odd.. | 13:28 |
| lucasagomes | it's horrible search | 13:28 |
| lucasagomes | I mean, feck, bind another key combination | 13:28 |
| lucasagomes | not the one the browsers use | 13:29 |
| sambetts | I don't mind them using the vim shortcuts for that sort of thing e.g. / for search but taking over the browser ones is a bit iffy... | 13:29 |
| *** betherly has quit IRC | 13:29 | |
| *** betherly has joined #openstack-ironic | 13:29 | |
| lucasagomes | sambetts, ++ yeah, leave the "/" that's fine | 13:29 |
| *** patrickeast has quit IRC | 13:30 | |
| *** patrickeast has joined #openstack-ironic | 13:30 | |
| *** betherly_ has joined #openstack-ironic | 13:32 | |
| *** betherly has quit IRC | 13:33 | |
| *** betherly_ is now known as betherly | 13:33 | |
| *** links has joined #openstack-ironic | 13:33 | |
| *** mbound_ has quit IRC | 13:35 | |
| *** trown|outttypeww is now known as trown | 13:37 | |
| jroll | mornin y'all | 13:38 |
| sambetts | morning jroll | 13:38 |
| lucasagomes | jroll, morning | 13:40 |
| lucasagomes | jroll, man, why we deploy the node twice on every gate job? | 13:40 |
| lucasagomes | http://logs.openstack.org/36/255936/4/check/gate-tempest-dsvm-ironic-pxe_ipa/0b8a1c5/logs/screen-ir-cond.txt.gz#_2015-12-17_11_40_18_775 | 13:40 |
| lucasagomes | http://logs.openstack.org/36/255936/4/check/gate-tempest-dsvm-ironic-pxe_ipa/0b8a1c5/logs/screen-ir-cond.txt.gz#_2015-12-17_11_48_51_106 | 13:41 |
| jroll | lucasagomes: yeah, I was wondering about that | 13:41 |
| jroll | OH | 13:41 |
| jroll | you know what, we test rebuild | 13:41 |
| lucasagomes | and the booting time is actually pretty long | 13:41 |
| lucasagomes | 7 minutes | 13:41 |
| lucasagomes | now that we increased the timeout, almost all jobs are failing on cleaning | 13:41 |
| lucasagomes | jroll, do we? /me checks | 13:41 |
| *** betherly_ has joined #openstack-ironic | 13:41 | |
| jroll | last I checked we did | 13:41 |
| lucasagomes | do_node_deploy Calling event: rebuild for node: f835feed-b54d-47d2-9b85-cea755fd0889 do_node_deploy /opt/stack/new/ironic/ironic/conductor/manager.py:551 | 13:41 |
| lucasagomes | yeah hmmmmmmmmmmmmmmmmmmmmm | 13:42 |
| lucasagomes | is a valid case but that's taking a lot of time | 13:42 |
| jroll | are most failing on the timeout fixture I assume? we can bump that some more | 13:42 |
| lucasagomes | at least ~8-10 minutes | 13:42 |
| lucasagomes | jroll, apparently yes | 13:42 |
| jroll | the fixture failure looks like this http://logs.openstack.org/74/256174/1/gate/gate-tempest-dsvm-ironic-pxe_ipa/e4d2ae1/console.html#_2015-12-17_00_21_33_291 | 13:43 |
| jroll | (1200 seconds) | 13:43 |
| lucasagomes | jroll, I wonder if we can disable rebuild and create a better job just for it | 13:43 |
| jroll | the funny part about deploys taking 7-8 minutes, is that's close to real world numbers | 13:43 |
| lucasagomes | something like starting with a node already provisoned and then rebuild from there | 13:43 |
| jroll | so we're truly simulating bare metal :P | 13:43 |
| lucasagomes | jroll, hah yeah | 13:44 |
| jroll | lucasagomes: sure, but how do we start with one provisioned? gotta come from somewhere | 13:44 |
| lucasagomes | but VMS, local network, should be quicker ... | 13:44 |
| jroll | yep | 13:44 |
| jroll | it would be if devstack ran on metal | 13:44 |
| lucasagomes | jroll, right, yeah we need a node in active state | 13:44 |
| *** patrickeast_ has joined #openstack-ironic | 13:45 | |
| jroll | I'd love to get some bare metal in the gate | 13:45 |
| trown | howdy ironic | 13:45 |
| jroll | morning trown | 13:45 |
| lucasagomes | jroll, ++ tho if the node takes 5 just to pass UEFI/BIOS when linking the hw interruptions and so on | 13:46 |
| lucasagomes | that would be a problem | 13:46 |
| lucasagomes | no rebuild for those jobs! | 13:46 |
| lucasagomes | trown, hey | 13:46 |
| lucasagomes | morning | 13:46 |
| trown | http://git.openstack.org/cgit/openstack/ironic/commit/?id=d489d28d87ce6291d3250c8e55ee51902b51001d is causing drivers to not get loaded in my tripleo env | 13:46 |
| trown | I am still trying to figure out why, but without that commit drivers load fine | 13:46 |
| *** [1]cdearborn has joined #openstack-ironic | 13:47 | |
| jroll | lucasagomes: well, more like bare metal hosts to run devstack in, get rid of the nested virt | 13:47 |
| jroll | trown: out of tree drivers or in tree? | 13:47 |
| trown | jroll: in tree... and even if I go to default of just pxe_ipmitool it still will not show up in ironic driver-list | 13:48 |
| jroll | O_o | 13:48 |
| * jroll tests | 13:48 | |
| lucasagomes | ouch | 13:48 |
| trown | and no traces or anything in conductor logs | 13:48 |
| jroll | upgrade or fresh install? | 13:48 |
| trown | fresh, this is in RDO CI that I am seeing it | 13:49 |
| trown | trying to get trunk CI passing for RDO | 13:49 |
| jroll | hrm | 13:49 |
| lucasagomes | and ironic-lib is installed? | 13:50 |
| jroll | I was trying to actually put this in our staging env yesterday but got stalled on some packaging issues :( | 13:50 |
| trown | lucasagomes: that might be it actually... I think we have not added that as a dependency anywhere... I will try that out | 13:50 |
| lucasagomes | trown, yeah you have to have ironic-lib installed | 13:51 |
| lucasagomes | it's a dependency | 13:51 |
| jroll | aha, that'll do it | 13:51 |
| lucasagomes | tho we should perhaps fail a bit better if it's not installed | 13:51 |
| * jroll makes too many assumptions | 13:51 | |
| trown | lucasagomes: ya, we do not even have ironic lib getting created in the ironic packaging | 13:52 |
| trown | I can fix that | 13:52 |
| lucasagomes | trown, oh :-( | 13:52 |
| lucasagomes | trown, I should have give you a heads up on that change then | 13:52 |
| lucasagomes | sorry | 13:52 |
| lucasagomes | trown, try installing from pip just to see if it works (the drivers load) | 13:52 |
| trown | no worries... I knew it was coming... just ETOOMANYTHINGS | 13:53 |
| TheJulia | rcarrillocruz: will you toss wet cats at me if I point out that your change really needs a release note ? *ducks* | 13:53 |
| lucasagomes | yeah | 13:53 |
| jroll | man, our gate is not nice | 13:53 |
| jroll | sigh | 13:53 |
| trown | CI is hard :P | 13:53 |
| lucasagomes | jroll, :-( yeah I'm trying to investigate too | 13:53 |
| lucasagomes | but urgh... that power on until initial lookup is taking too much time | 13:54 |
| * jroll revives email about onmetal in the gate | 13:54 | |
| lucasagomes | 6~7 min... and we deploy it twice | 13:54 |
| lucasagomes | + cleaning | 13:54 |
| jroll | yeah man | 13:54 |
| jroll | and I'm not sure at all what we can do about it | 13:54 |
| lucasagomes | yeah, looking at the logs see where it's taking that long | 13:55 |
| lucasagomes | is it to download the ramdisk or to actually boot the machine | 13:55 |
| *** alexpilotti has joined #openstack-ironic | 13:55 | |
| lucasagomes | jroll, [ 253.844815] systemd[1]: Startup finished in 33.267s (kernel) + 57.888s (initrd) + 2min 42.132s (userspace) = 4min 13.288s | 13:57 |
| jroll | jeeeez | 13:57 |
| lucasagomes | so it may be taking ~3 to download | 13:58 |
| jroll | needs more tinyipa | 13:58 |
| lucasagomes | + ~4m to actually boot it | 13:58 |
| trown | lucasagomes: confirmed that pip installing ironic-lib resolves the driver loading issue... thanks! | 13:58 |
| TheJulia | yeouch thats a long boot | 13:58 |
| lucasagomes | trown, np! | 13:58 |
| lucasagomes | jroll, yeah tinyipa or what we can do is try to look at the services started in the base OS | 13:59 |
| lucasagomes | and try to disable some | 13:59 |
| jroll | yeah | 13:59 |
| *** alexpilotti has quit IRC | 14:00 | |
| lucasagomes | there's a bunch of "Started Update Engine Stub Timer.", "Watch for a cloud-config at /media/configdrive", "Daily Cleanup of Temporary Directories." | 14:01 |
| lucasagomes | that seems related to coreos that we don't use in a ramdisk | 14:01 |
| lucasagomes | jroll, I will take a look at disabling some of those | 14:01 |
| jroll | cool | 14:01 |
| sambetts | could try tinyipa, would need the modify the devstack plugin to pull my patch and build it | 14:02 |
| lucasagomes | sambetts, that would be good too... because if we can cut the download time it helps a lot | 14:02 |
| *** rloo has joined #openstack-ironic | 14:02 | |
| rcarrillocruz | TheJulia: hah | 14:03 |
| rcarrillocruz | i have no idea what that is about | 14:03 |
| rcarrillocruz | i assume is the 'reno' thing | 14:03 |
| rcarrillocruz | ? | 14:03 |
| rcarrillocruz | if you point me i can look at it sure... | 14:03 |
| lucasagomes | if we cut it from 3 min to 1 min (which is resonable in the local network) that helps, it's almost all the time we spent on the user space stuff | 14:03 |
| lucasagomes | + tinyapi is not coreos so those services won't be there | 14:03 |
| jroll | sambetts: yeah, worth a try | 14:04 |
| lucasagomes | sambetts, do you have any data about tinyapi? When testing locally, how long does it take to boot? | 14:04 |
| lucasagomes | more or less? u know? | 14:04 |
| * lucasagomes tests | 14:04 | |
| TheJulia | rcarrillocruz: yup, reno. The docs are actually good and its actually very simple :) | 14:04 |
| openstackgerrit | Ruby Loo proposed openstack/python-ironicclient: Add 'node-set-provision-state <node> clean' https://review.openstack.org/258694 | 14:06 |
| jroll | lucasagomes: I'm going to bump the overall timeout another 5 minutes, I think that should do it yeah? | 14:08 |
| lucasagomes | jroll, yeah... I think that would do it | 14:08 |
| jroll | maybe 10 to be safe? | 14:08 |
| jroll | the downside of bumping that | 14:08 |
| jroll | is people wait longer on results | 14:08 |
| jroll | but I think that's okay short-term | 14:08 |
| lucasagomes | jroll, right, well we can bump it but continue to work on improving | 14:09 |
| jroll | also, I bugged a guy internally well versed in this stuff... he said we're pretty much hosed performance-wise | 14:09 |
| lucasagomes | jroll, because right now, we have to run the jobs 3 times to get things merged | 14:09 |
| jroll | unless we can convince everyone to enable nested virt | 14:09 |
| lucasagomes | so waiting 10 minutes more is actually better | 14:09 |
| lucasagomes | jroll, yeah, we can start that conversation too | 14:10 |
| lucasagomes | but still, how can it take that long to download the ramdisk to the nested VMs | 14:10 |
| lucasagomes | should be super quick... | 14:10 |
| * lucasagomes can't figure out why | 14:10 | |
| *** pas-ha has joined #openstack-ironic | 14:10 | |
| jroll | well, it has to pass through cpu and ram of the vm | 14:10 |
| jroll | which is super slow running virt in virt | 14:11 |
| jroll | *unless* nested virt is enabled | 14:11 |
| lucasagomes | right, yeah maybe the emulated devices are just slow and that's it | 14:11 |
| jroll | yep | 14:11 |
| jroll | but yeah, probably a couple things here making it worse recently: | 14:12 |
| jroll | * devstack plugin is probably slightly slower (maybe not, dunno) | 14:12 |
| jroll | * newer coreos probably has more services | 14:12 |
| lucasagomes | jroll, one thing you can observe that indicates that the machine has been powered off abruptly is the last console log http://logs.openstack.org/36/255936/4/check/gate-tempest-dsvm-ironic-pxe_ipa/0b8a1c5/logs/ironic-bm-logs/ | 14:12 |
| lucasagomes | that haven't being moved by the libvirt hook yet | 14:12 |
| *** jcoufal_ has joined #openstack-ironic | 14:13 | |
| jroll | yep | 14:13 |
| lucasagomes | meaning that we Ironic didn't call "virsh destroy" | 14:13 |
| jroll | right | 14:13 |
| jroll | so the other option here is disable cleaning completely in the gate until we figure this out | 14:14 |
| jroll | which I don't really like | 14:14 |
| lucasagomes | jroll, yeah, we have cleaning and rebuild | 14:14 |
| lucasagomes | that would also help if we disable | 14:14 |
| jroll | but it's in our repo, instead of begging people to land a patch for the timeout | 14:14 |
| jroll | rebuild is in tempest repo though :/ | 14:14 |
| lucasagomes | jroll, right | 14:14 |
| *** romcheg1 has quit IRC | 14:14 | |
| rloo | hi lucasagomes, jroll | 14:15 |
| lucasagomes | I'm buiilding an env to test the tinyipa, see if I can get some data about it | 14:15 |
| lucasagomes | rloo, morning ruby! | 14:15 |
| rloo | we should increase the timeout so that the tests pass. otherwise, we're retrying and that takes up more resources. | 14:15 |
| lucasagomes | I will grab some quick lunch while it's building | 14:15 |
| jroll | morning rloo :) | 14:16 |
| lucasagomes | rloo, ++ or disable cleaning temporarely as jroll suggested | 14:16 |
| jroll | here's the timeout thing https://review.openstack.org/#/c/259035/ | 14:16 |
| lucasagomes | that will help too | 14:16 |
| rloo | is it 'just' the cleaning/ramdisk that is causing the timeout? | 14:16 |
| *** jcoufal has quit IRC | 14:16 | |
| lucasagomes | rloo, no, it's the overall stuff that adds up | 14:16 |
| jroll | most fails seem to be during cleaning right now | 14:16 |
| jroll | which is the last ramdisk boot | 14:16 |
| lucasagomes | cleaning just happen to be the last boot | 14:16 |
| rloo | we had the cleaning stuff working before. what has caused the slowdown all of a sudden. | 14:16 |
| jroll | but really it's the slow boot time | 14:17 |
| * lucasagomes wonders that too | 14:17 | |
| jroll | rloo: my assumption is newer coreos version does more at startup | 14:17 |
| rloo | is it the overall time from the start of the test, or the time doing this one boot? | 14:17 |
| lucasagomes | I will be right back (people waiting me to lunch) | 14:17 |
| *** lucasagomes is now known as lucas-hungry | 14:17 | |
| jroll | it's all boots | 14:17 |
| jroll | we do 3 ramdisk boots, and two cirros boots | 14:17 |
| jroll | cirros is quick | 14:17 |
| rloo | and the timeout is for the sum of the times for all the boots? | 14:18 |
| jroll | ~30 seconds for cirros | 14:18 |
| jroll | yeah, so there's three timeouts | 14:18 |
| jroll | provision timeout (boot -> ACTIVE), unprovision timeout (delete -> AVAILABLE), and tempest test timeout (full test time) | 14:18 |
| jroll | the last one is the only one failing now | 14:18 |
| rloo | it would be nice to increase the timeout so nothing times out, and then collect the actual amounts of time the boots actually take. | 14:19 |
| jroll | right, so lucas made us not overwrite boot logs now | 14:19 |
| jroll | so we have more data | 14:19 |
| jroll | e.g. http://logs.openstack.org/36/255936/4/check/gate-tempest-dsvm-ironic-pxe_ipa/0b8a1c5/logs/ironic-bm-logs/ | 14:19 |
| jroll | one ramdisk boot: http://logs.openstack.org/36/255936/4/check/gate-tempest-dsvm-ironic-pxe_ipa/0b8a1c5/logs/ironic-bm-logs/baremetalbrbm_0_console_17-12-2015-11:40:39.txt.gz | 14:19 |
| jroll | [ 253.844815] systemd[1]: Startup finished in 33.267s (kernel) + 57.888s (initrd) + 2min 42.132s (userspace) = 4min 13.288s. | 14:19 |
| jroll | so we're investigating maybe moving to tinyipa to eliminate all the junk coreos does that we don't use here | 14:20 |
| rloo | good, now how to automate collecting those numbers :) | 14:20 |
| jroll | and also figuring out how to disable some of coreos' fancy stuff we don't use | 14:20 |
| jroll | hehe | 14:20 |
| * jroll wonders if these get to logstash | 14:20 | |
| *** _degorenko|afk is now known as degorenko | 14:20 | |
| rloo | my other thought was whether this was only with certain test machines. | 14:21 |
| jroll | I've wondered that too | 14:21 |
| jroll | it seems to be ovh every time I think to look | 14:21 |
| jroll | but I don't have data on provider vs fail rate | 14:21 |
| rloo | jroll: well, it would be useful info but i suppose we need to fix it regardless | 14:22 |
| jroll | mhm | 14:22 |
| *** kirankv has quit IRC | 14:23 | |
| rloo | jroll: the latest coreos merged on Monday, and it started timing out a lot last Wed-Thurs. | 14:25 |
| * rloo wonders what's wrong with assertEqual(None, *) anyway and why we care to change it to assertIsNone(). | 14:26 | |
| jroll | rloo: I agree http://tinyurl.com/p8t8osb | 14:27 |
| jroll | rloo: on that, I assume it's a "== None" vs "is None" thing, the latter is more "right" but I'm not sure if it matters | 14:27 |
| rloo | jroll: yeah, that's what i mean. it wasn't wrong in the first place. seems like there are lots of other things to be fixing. anyway. | 14:28 |
| *** mbound_ has joined #openstack-ironic | 14:28 | |
| jroll | rloo: we prefer easy inconsequential work, it seems | 14:28 |
| rloo | jroll: so yeah, the coreos probably just made things that much worse. i guess an accumulation of stuff... | 14:28 |
| rloo | jroll: the royal 'we' that doesn't include me :) | 14:28 |
| jroll | rloo: indeed :) | 14:29 |
| rloo | lucas-hungry: when you get back, what do you want me to do wrt your comments on https://review.openstack.org/#/c/257662/? | 14:30 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Don't clean nodes by default in devstack https://review.openstack.org/259046 | 14:32 |
| jroll | there's the disable cleaning patch ^ | 14:32 |
| jroll | I'm going to put it through a few rechecks to see how things go, before I'm okay with landing it | 14:32 |
| *** jcook has quit IRC | 14:32 | |
| *** thiagop has joined #openstack-ironic | 14:34 | |
| *** alexpilotti has joined #openstack-ironic | 14:34 | |
| *** jcook has joined #openstack-ironic | 14:36 | |
| *** alexpilotti has quit IRC | 14:37 | |
| thiagop | morning folks | 14:39 |
| *** alexpilotti has joined #openstack-ironic | 14:40 | |
| rloo | jroll: if the cleaning is causing the timeout, why would it timeout waiting for node to reach active? I'd think it would be 'available'. | 14:40 |
| jroll | rloo: it's the "full test timeout" thing | 14:40 |
| jroll | tempest has a per-test timeout | 14:41 |
| jroll | the baremetal test does deploy, rebuild, delete | 14:41 |
| rloo | jroll: OH. Yikes, we should bump it up to 1 hour then. | 14:41 |
| jroll | so that cleaning on the tail end is where tempest nukes it from orbit | 14:41 |
| jroll | lol | 14:41 |
| jroll | rloo: we try to keep it reasonable so that we can't hog gate resources and return results in a reasonable time | 14:42 |
| *** ChuckC has joined #openstack-ironic | 14:42 | |
| *** romcheg1 has joined #openstack-ironic | 14:42 | |
| jroll | sambetts: can you add a patch for tinyipa today so we can play with that? :) | 14:42 |
| sambetts | a patch, a rebase? | 14:44 |
| thiagop | sambetts: are we using the tinyipa image to deploy agent? The coreos image is still not working? | 14:46 |
| jroll | sambetts: a patch to ironic/devstack/ | 14:46 |
| sambetts | ooh, yeah I'll look into it | 14:47 |
| jroll | sweet, thank you | 14:50 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Document the process of proposing new features https://review.openstack.org/259055 | 14:53 |
| *** betherly_ has quit IRC | 14:54 | |
| *** patrickeast_ has quit IRC | 14:54 | |
| jroll | vdrok: that was on my todo list for today, you rock :) | 14:54 |
| vdrok | lucas-hungry, jroll rloo ^^ please take a look, there is some copypaste from neutron guide :) | 14:55 |
| vdrok | jroll, :) | 14:55 |
| rloo | thx vdrok, it wasn't on my list :) | 14:55 |
| sambetts | jroll: does the gate default to building its own deploy ramdisk? | 15:04 |
| *** rvasilets___ has joined #openstack-ironic | 15:05 | |
| sambetts | jroll: also the function build_ipa_coreos_ramdisk uses $IRONIC_PYTHON_AGENT_REPO and I don't see that defined anywhere | 15:06 |
| *** caiobo has quit IRC | 15:06 | |
| jroll | sambetts: everything except the *-src job uses the ramdisk from tarballs.o.o | 15:06 |
| jroll | code/devstack/stackrc | 15:07 |
| jroll | 521:IRONIC_PYTHON_AGENT_REPO=${IRONIC_PYTHON_AGENT_REPO:-${GIT_BASE}/openstack/ironic-python-agent.git} | 15:07 |
| jroll | ^ that's set in devstack currently | 15:07 |
| sambetts | ah ok :) | 15:07 |
| jroll | that's only used in IPA jobs afaik | 15:08 |
| sambetts | hmm.. ok so I'll have to override more than I though then | 15:08 |
| jroll | where at? | 15:09 |
| sambetts | upload_baremetal_ironic_deploy | 15:09 |
| jroll | sambetts: if you want to just start with blowing things away to build tinyipa instead, that's fine | 15:09 |
| jroll | I just want to see some gate jobs running on it | 15:10 |
| *** lucas-hungry is now known as lucasagomes | 15:13 | |
| lucasagomes | rloo, hi there, I think u can remove that test from availble to cleaning... or maybe adding a generic test to ensure 'done' is can't be set by any state | 15:13 |
| *** hemna has joined #openstack-ironic | 15:13 | |
| lucasagomes | vdrok, looking | 15:14 |
| rloo | lucasagomes: ok. i'll just remove it. thx. | 15:14 |
| jlvillal | mordred, Has anyone else reported issues voting for the 'O' release name? I get 'Your voter key is invalid. You should have received a correct URL by email.' | 15:15 |
| jlvillal | My 'N' release name voting worked. | 15:15 |
| rloo | if anyone knows how to get rid of the yellow background for comments, in new gerrit, i'll buy them a beer | 15:17 |
| NobodyCam | good morning Ironicers | 15:18 |
| NobodyCam | good morning jroll jlvillal rloo JayF sinval vdrok sambetts devananda TheJulia thiagop mgould and anyone else not listed here :) | 15:19 |
| rloo | g'morning NobodyCam | 15:19 |
| thiagop | hey NobodyCam, morning | 15:19 |
| jroll | heya NobodyCam :) | 15:19 |
| jlvillal | NobodyCam, morning | 15:19 |
| NobodyCam | :) | 15:20 |
| NobodyCam | morning | 15:20 |
| jlvillal | Good morning: rloo, thiagop jroll sambetts vdrok and everyone else! | 15:20 |
| NobodyCam | wow I actually sletp in a bit this morning :) | 15:20 |
| sambetts | o/ NobodyCam jlvillal | 15:20 |
| thiagop | jlvillal: o/ | 15:20 |
| NobodyCam | :) | 15:20 |
| * jlvillal ponders if any PacktPub e-books are worth $5. Thinking about the Neutron book by the racker James Denton :) | 15:21 | |
| jroll | jlvillal: I hear that's a great book :) | 15:21 |
| rloo | morning jlvillal | 15:22 |
| jlvillal | jroll, :) | 15:22 |
| jlvillal | rloo, :) | 15:22 |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Prevent 'manage' from CLEANING https://review.openstack.org/257662 | 15:23 |
| jroll | jlvillal: there's print copies and such too, if the question was about the quality of the books packtpub publishes | 15:24 |
| jlvillal | jroll, I was wondering about quality of the content of the books. I think I will buy the Neutron book :) | 15:25 |
| *** alexpilotti has quit IRC | 15:25 | |
| *** daemontool has joined #openstack-ironic | 15:27 | |
| *** alexpilotti has joined #openstack-ironic | 15:32 | |
| *** raildo is now known as raildo-afk | 15:33 | |
| *** eggmaster has joined #openstack-ironic | 15:34 | |
| *** mtanino has joined #openstack-ironic | 15:36 | |
| *** kkoski has joined #openstack-ironic | 15:36 | |
| *** kkoski has quit IRC | 15:38 | |
| *** raildo-afk is now known as raildo | 15:38 | |
| *** kalpase has quit IRC | 15:39 | |
| vdrok | morning NobodyCam :) | 15:50 |
| *** links has quit IRC | 15:50 | |
| *** david-lyle_ has joined #openstack-ironic | 15:55 | |
| sambetts | jroll: how do I make sure that a non-python requirement is installed? | 15:55 |
| sambetts | in devstack | 15:55 |
| jroll | sambetts: devstack/files/{apts,rpms} | 15:56 |
| NobodyCam | morning :) | 15:56 |
| sambetts | jroll: thanks :) | 15:56 |
| * jroll bbiab | 15:57 | |
| *** david-lyle_ has quit IRC | 15:57 | |
| * lucasagomes hates that ctrl+f bind in the new gerrit interface | 15:58 | |
| lucasagomes | damn that's garbage! | 15:58 |
| *** david-lyle has quit IRC | 15:59 | |
| openstackgerrit | Sam Betts proposed openstack/ironic: Enabled tinyipa for devstack Ironic https://review.openstack.org/259089 | 15:59 |
| sambetts | No idea if thats going to work ^ | 15:59 |
| *** ChuckC has quit IRC | 15:59 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Remove conditional preventing iPXE to be configured with UEFI https://review.openstack.org/257497 | 16:00 |
| sambetts | jroll, jlvillal: I have that neutron book, its been extremely useful | 16:00 |
| *** mbound_ has quit IRC | 16:01 | |
| jlvillal | sambetts, Thanks! I did buy it :) | 16:01 |
| *** mbound_ has joined #openstack-ironic | 16:03 | |
| *** jrist has joined #openstack-ironic | 16:04 | |
| *** kkoski has joined #openstack-ironic | 16:04 | |
| *** david-lyle has joined #openstack-ironic | 16:05 | |
| *** jrist has quit IRC | 16:05 | |
| *** jrist has joined #openstack-ironic | 16:06 | |
| *** vishwana_ has joined #openstack-ironic | 16:07 | |
| *** vishwanathj has quit IRC | 16:09 | |
| eggmaster | sambetts: which neutron book? thx | 16:10 |
| *** boris-42_ has joined #openstack-ironic | 16:11 | |
| sambetts | eggmaster: https://www.packtpub.com/virtualization-and-cloud/learning-openstack-networking-neutron-second-edition | 16:11 |
| *** vinm213 has joined #openstack-ironic | 16:21 | |
| vinm213 | Hi all | 16:23 |
| NobodyCam | good morning vinm213 | 16:23 |
| vinm213 | Good morning NobodyCam | 16:24 |
| vinm213 | Actually it's night here | 16:24 |
| *** mbound_ has quit IRC | 16:25 | |
| vinm213 | I want to discuss about a recent fix that lucas has proposed | 16:25 |
| *** romcheg1 has left #openstack-ironic | 16:25 | |
| *** mbound_ has joined #openstack-ironic | 16:25 | |
| vinm213 | https://review.openstack.org/#/c/257497/ | 16:25 |
| vinm213 | uefi - ipxe support for ironic | 16:25 |
| vinm213 | I was trying to test it on my set u | 16:26 |
| vinm213 | What would be the config template we use in this case? | 16:27 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Document the process of proposing new features https://review.openstack.org/259055 | 16:29 |
| rloo | vinm213: you mean the ipxe config template? it is mentioned in the install guide -- in that patch. | 16:30 |
| *** alexpilotti has quit IRC | 16:32 | |
| lucasagomes | vinm213, yeah you should use the iPXE template... but there's a lot of moving parts there | 16:32 |
| lucasagomes | if you are using local boot the images should have grub-efi installed and so on | 16:32 |
| vinm213 | yes. but if it is ipxe we use "pxe_config_template=$pybasedir/drivers/modules/ipxe_config.template" | 16:32 |
| * lucasagomes is looking for an env to test the end to end | 16:32 | |
| lucasagomes | vinm213, right | 16:33 |
| *** ukalifon has quit IRC | 16:33 | |
| vinm213 | And if it is uefi it is "uefi_pxe_config_template=$pybasedir/drivers/modules/pxe_grub_config.template" for grub bootloader | 16:33 |
| *** alexpilotti has joined #openstack-ironic | 16:33 | |
| sambetts | lucasagomes: My devstack patch is literally a test to see if tinyipa even works in the gate, because the normal jobs won't build the image from scratch | 16:33 |
| vinm213 | now which should i go for ? | 16:33 |
| lucasagomes | vinm213, depend on the ROM you boot right? if you are using that bootx64 that's the grub | 16:34 |
| lucasagomes | if you use ipxe.efi you should use the ipxe one | 16:34 |
| vinm213 | it's chain loading | 16:34 |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Show transitions initiated by API requests https://review.openstack.org/259099 | 16:34 |
| lucasagomes | tho it's all in tests yet | 16:34 |
| rloo | jlvillal: you wondered what I was going to do with that FSM.is_stable() method ^^. Didn't give me as much mileage as I had hoped but... | 16:35 |
| vinm213 | ok so the parameter "uefi_pxe_config_template" will be ignored then? | 16:35 |
| * lucasagomes checks | 16:35 | |
| lucasagomes | vinm213, lemme check | 16:35 |
| jlvillal | rloo, Thanks for the pointer. | 16:36 |
| rloo | jlvillal: if you're curious :) it is all document-related. | 16:36 |
| jlvillal | rloo, Looking now :) | 16:37 |
| lucasagomes | vinm213, https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/pxe.py#L479-L483 | 16:37 |
| rloo | jlvillal: better to look after jenkins generates the diagram | 16:37 |
| lucasagomes | vinm213, you have to set the ipxe template to uefI_pxe_config_template | 16:37 |
| * lucasagomes needs an env to test it all asap | 16:37 | |
| rloo | jlvillal: oh wait, i generated the diagram. better to look after jenkins generates the docs :) | 16:38 |
| *** alexpilotti has quit IRC | 16:38 | |
| jlvillal | rloo, Yeah I looked at the SVG file and was thinking, "I wish I knew how to see the picture" :) | 16:38 |
| rloo | vdrok: I don't have answers to some of those questions in 259055, will leave it for our PTL :) | 16:38 |
| rloo | jlvillal: yeah, I know. Can't just do a 'diff' on that! | 16:39 |
| lucasagomes | vinm213, can you set it and tell me if it works? | 16:39 |
| vdrok | It seems that now gerrit posts a link to the change only with closes-bug? | 16:39 |
| vinm213 | sure i will | 16:39 |
| lucasagomes | vinm213, uefi_pxe_config_template=$pybasedir/drivers/modules/ipxe_config.template | 16:39 |
| vinm213 | yes. i got it | 16:39 |
| lucasagomes | thanks | 16:39 |
| vdrok | rloo, you changed to partial-bug in cleaning patches and it did not appear in rfe bug? | 16:40 |
| rloo | vdrok: oh. that's the problem with bugs. if you open a bug *after* the patch exists, the bug won't get updated. so sad. | 16:40 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Remove conditional preventing iPXE to be configured with UEFI https://review.openstack.org/257497 | 16:40 |
| vinm213 | Thanks | 16:40 |
| rloo | vdrok: switching to rfe bugs is going to be a bit of a pain. | 16:40 |
| vdrok | rloo, hmm, but it seems to work with closes-bug | 16:41 |
| rloo | vdrok: oh really? i didn't think it did that. well, so it isn't totally broken then. | 16:41 |
| vdrok | i'll check, i think i saw that | 16:42 |
| * lucasagomes forgot to add a way to remove the qemu script even if IRONIC_LOG_ROTATE is false :-( | 16:42 | |
| * lucasagomes cooks a patch | 16:42 | |
| lucasagomes | qemu hook* | 16:42 |
| -openstackstatus- NOTICE: Zuul is moving in very slow motion since roughly 13:30 UTC; the Infra team is investigating. | 16:42 | |
| *** jaybeale has joined #openstack-ironic | 16:44 | |
| vdrok | rloo, yep, you're right, it appears only if the patch is proposed after the bug :( | 16:46 |
| *** mkovacik has quit IRC | 16:47 | |
| rloo | vdrok: :-( At least for that one, I had copy/pasted from the BP, so the link to that patch is there. | 16:48 |
| *** degorenko is now known as _degorenko|afk | 16:49 | |
| vdrok | rloo, yes, so it seems that for the ones that are moved we need to not to forget to add the links when the bug goes to fix released | 16:50 |
| *** davideag_ has joined #openstack-ironic | 16:50 | |
| rloo | vdrok: so much overhead. who's great idea was this anyway? :) | 16:50 |
| vdrok | :) | 16:50 |
| *** davideagnello has quit IRC | 16:51 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Remove QEMU hook if IRONIC_VM_LOG_ROTATE is set to False https://review.openstack.org/259104 | 16:53 |
| *** davideag_ has quit IRC | 16:56 | |
| *** jaybeale has quit IRC | 17:05 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic-specs: Add indexable JSON fields spec https://review.openstack.org/253605 | 17:05 |
| jlvillal | aarefiev, Question on https://review.openstack.org/#/c/252946/3..6/specs/ironic-node-inb-auto-discovery.rst | 17:08 |
| jlvillal | sambetts, You commented on: https://review.openstack.org/#/c/252946/3..6/specs/ironic-node-inb-auto-discovery.rst Maybe you can answer my question. | 17:09 |
| jlvillal | :) | 17:09 |
| aarefiev | jlvillal: hi, listening | 17:09 |
| jlvillal | aarefiev, Great | 17:09 |
| jlvillal | It mentions extending 'set-attribute'. Where does that come from? Is that an inspector thing? Ironic thing? | 17:10 |
| aarefiev | jlvillal: it's existing action in inspector, it just update node attributes | 17:11 |
| sambetts | jlvillal: set-attribute is a ironic-inspector rule action | 17:11 |
| aarefiev | :) | 17:11 |
| jlvillal | aarefiev, sambetts Okay. I don't really see the spec talking much about inspector. At least spelling out that this is about inspector. | 17:11 |
| *** Marga_ has joined #openstack-ironic | 17:11 | |
| jlvillal | I will comment to that effect :) | 17:11 |
| JayF | lucasagomes: I'm going to push up a patch today adding a dont-poweroff-after-cleanfail option | 17:12 |
| JayF | lucasagomes: I know youv've wanted that in the past, so wanted to give you a heads up | 17:12 |
| lucasagomes | JayF, ++ | 17:12 |
| lucasagomes | JayF, I want one for DEPLOYFAIL too | 17:12 |
| sambetts | jlvillal: Its in the ironic-inspector-specs repo do we need to specifically say in the spec that its for inspector? | 17:12 |
| lucasagomes | JayF, that's very useful to not power off stuff | 17:12 |
| JayF | lucasagomes: The cleanfail one is a downstream patch | 17:12 |
| jlvillal | sambetts, Ah duh! | 17:12 |
| JayF | lucasagomes: so was going to just copy it upstream | 17:12 |
| lucasagomes | JayF, ++ | 17:13 |
| * jlvillal feels foolish... | 17:13 | |
| sambetts | easy mistakes to make :) | 17:13 |
| aarefiev | jlvillal: http://docs.openstack.org/developer/ironic-inspector/usage.html#introspection-rules | 17:13 |
| lucasagomes | JayF, btw https://review.openstack.org/#/c/248832/ | 17:13 |
| JayF | lucasagomes: on my review queue now | 17:14 |
| lucasagomes | JayF, this also helps, I plan to collect the system logs in case of DEPLOYFAIL and upload it to swift | 17:14 |
| lucasagomes | thx | 17:14 |
| * JayF needs to make his downstream hwms less log-noisy if he's going to enable that | 17:14 | |
| JayF | I think we print the output of dmidecode about a half dozen times | 17:14 |
| JayF | lol | 17:14 |
| lucasagomes | hah | 17:15 |
| lucasagomes | when it comes to troubleshoot everything helps | 17:16 |
| rloo | hey lucasagomes. wrt bashate. is it supposed to complete? i just set up a fresh dev, and it fails on ironic/devstack/upgrade/upgrade.sh : L52 | 17:18 |
| *** Marga_ has quit IRC | 17:18 | |
| *** jistr has quit IRC | 17:19 | |
| jlvillal | aarefiev, Left a review. I for sure don't know that much about inspector. But hopefully some of it is useful :) | 17:20 |
| lucasagomes | rloo, to complete? I guess so | 17:20 |
| lucasagomes | rloo, lemme take a look | 17:20 |
| rloo | lucasagomes: yeah, i mean successful, not fail. | 17:20 |
| jroll | lucasagomes: a couple small things inline on the properties json spec, LGTM though overall \o/ | 17:20 |
| lucasagomes | jroll, cheers! | 17:20 |
| rloo | lucasagomes: unless there is something wrong with my env. | 17:20 |
| lucasagomes | rloo, yes it's suppose to successfully pass | 17:20 |
| *** mannidi has joined #openstack-ironic | 17:21 | |
| lucasagomes | rloo, lemme check | 17:21 |
| lucasagomes | are you running it as part of the pep8 command? or bashate? | 17:21 |
| rloo | lucasagomes: i think pep8 cuz i just did 'tox' | 17:21 |
| lucasagomes | rloo, ack | 17:21 |
| lucasagomes | yeah because the last patch in the series moves it to the pep8 | 17:21 |
| lucasagomes | rloo, setting up a tox here to test | 17:21 |
| aarefiev | jlvillal: thank you, sir | 17:21 |
| jroll | rloo: lucasagomes: that file is from the grenade plugin, I bet the patch just needs a rebase and some fixes there | 17:21 |
| jroll | the grenade plugin merged last night | 17:22 |
| lucasagomes | ohh | 17:22 |
| lucasagomes | ok, lemme test and I will rebase | 17:22 |
| jroll | I can update it, where's the bashate review? | 17:22 |
| sambetts | jroll: https://review.openstack.org/#/c/259089 waiting for the results to see if its worked | 17:22 |
| jroll | sambetts: \o/ thank you | 17:23 |
| *** bradjones has quit IRC | 17:23 | |
| *** trown is now known as trown|lunch | 17:23 | |
| lucasagomes | jroll, https://review.openstack.org/#/c/258506/ | 17:23 |
| *** bradjones has joined #openstack-ironic | 17:24 | |
| *** bradjones has quit IRC | 17:24 | |
| *** bradjones has joined #openstack-ironic | 17:24 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Config to stop powering off nodes on cleanfail https://review.openstack.org/259119 | 17:24 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Document the process of proposing new features https://review.openstack.org/259055 | 17:24 |
| JayF | lucasagomes: ^ literally 100% identical to this patch we're running in prod | 17:24 |
| jroll | lucasagomes: cool, I'll update | 17:25 |
| lucasagomes | JayF, awesome! I will take a look | 17:25 |
| rloo | biab | 17:25 |
| lucasagomes | jroll, cool, just run it locally see if it need to fix some more syntax errors there | 17:25 |
| jroll | yep | 17:25 |
| *** rloo is now known as rloo_afk | 17:25 | |
| *** [1]cdearborn has quit IRC | 17:25 | |
| lucasagomes | my env is almost finished (slower conn here in brazil than I have in ire) | 17:25 |
| *** [1]cdearborn has joined #openstack-ironic | 17:26 | |
| lucasagomes | jroll, it's failing :-( | 17:26 |
| jroll | lucasagomes: yeah, probably more indentation problems, I'll take care of it | 17:27 |
| lucasagomes | jroll, thanks a lot | 17:27 |
| jroll | np | 17:27 |
| jroll | I introduced the bug, I fix the bug :P | 17:27 |
| *** jcoufal_ has quit IRC | 17:28 | |
| lucasagomes | JayF, question... even when you set power_off_after_failure to True | 17:28 |
| lucasagomes | JayF, nova will power it off anyway no? Cause it will call destroy() to the instance | 17:29 |
| lucasagomes | JayF, do you have something in nova preventing it? | 17:29 |
| JayF | lucasagomes: this patch is for cleanfail, not for deployfail | 17:29 |
| * lucasagomes long time I don't look at it | 17:29 | |
| lucasagomes | https://review.openstack.org/#/c/259119/1/ironic/drivers/modules/deploy_utils.py | 17:29 |
| lucasagomes | JayF, ^ | 17:29 |
| lucasagomes | maybe you squashed both patches there then | 17:29 |
| JayF | hmm. Yeah I see. | 17:29 |
| JayF | We do *not* quash that behavior in Nova | 17:30 |
| *** Marga_ has joined #openstack-ironic | 17:30 | |
| JayF | so that's essentially useless. | 17:30 |
| JayF | Want me to limit it to just cleanfail then? | 17:30 |
| lucasagomes | JayF, I think so, or update the commit message to reflect on that change too | 17:30 |
| jroll | I'd do both and update commit message | 17:31 |
| JayF | k | 17:31 |
| jroll | it's still valuable, even if nova stomps on it | 17:31 |
| lucasagomes | JayF, I would love to have that for deploy fail too btw, but we need to also change something in nova | 17:31 |
| lucasagomes | jroll, right | 17:31 |
| JayF | lucasagomes: but it does have immediate usefulness for folks using bifrost | 17:31 |
| sambetts | I mean you could add it so that it works in standalone | 17:31 |
| lucasagomes | JayF, totally, I'm totally in favor of having it | 17:31 |
| lucasagomes | just mention it in the commit message, otherwise it's a bit sneaky | 17:31 |
| *** mbound_ has quit IRC | 17:32 | |
| jroll | lucasagomes: I'm getting this locally with tox -r -epep8, unclear why ./ironic/drivers/base.py:1028:21: F821 undefined name '_' | 17:32 |
| jroll | unless flake just released and is about to break our gate or something | 17:32 |
| lucasagomes | jroll, right, yeah that's flake cause the bashate just looks at /devstack | 17:33 |
| jroll | right | 17:33 |
| lucasagomes | bash -c "find {toxinidir}/devstack | 17:33 |
| *** cemason has joined #openstack-ironic | 17:33 | |
| jroll | coming from all over this method https://github.com/openstack/ironic/blob/master/ironic/drivers/base.py#L1004 | 17:33 |
| lucasagomes | odd, I'm running flake8 no problem here | 17:34 |
| lucasagomes | and also _( we do a gettext install | 17:35 |
| lucasagomes | so should be global | 17:35 |
| jroll | I don't believe we do anymore | 17:35 |
| jroll | anyway, I'll fix that later | 17:35 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Config to stop powering off nodes on failure https://review.openstack.org/259119 | 17:35 |
| lucasagomes | thanks | 17:35 |
| JayF | ^ commit message updated | 17:35 |
| lucasagomes | jroll, maybe oslo.i18n do it... not sure | 17:36 |
| lucasagomes | or should do it | 17:36 |
| sambetts | lucasagomes: there hasn't been any merges into i18n for 13 days | 17:36 |
| lucasagomes | right, yeah idk... gotta investigate | 17:36 |
| jroll | it's fine if it's not breaking the gate | 17:37 |
| jroll | could be a flake version thing too | 17:37 |
| jroll | [E] E042: local declaration hides errors: ' local old_driver=$(source $BASE_DEVSTACK_DIR/functions; source $BASE_DEVSTACK_DIR/localrc; echo $VIRT_DRIVER)' | 17:37 |
| jroll | that doesn't tell me much... | 17:38 |
| sambetts | I've recently had to do this https://review.openstack.org/#/c/258641/ for networking-cisco to get rid of a whole done of i18n deprecation errors, unsure if related | 17:38 |
| JayF | jroll: you don't handle the case where echo fails | 17:38 |
| JayF | jroll: that var could be empty when you're done | 17:38 |
| jroll | JayF: yeah, I figured that was the case | 17:40 |
| jroll | I think the word "local" misdirected me | 17:40 |
| *** ifarkas has quit IRC | 17:40 | |
| lucasagomes | jroll, yeah that's an odd command as well... cause old_driver == VIRT_DRIVER no? | 17:40 |
| lucasagomes | maybe it changes when you source stuff. /me doesn't know much bash | 17:41 |
| jroll | JayF: making it non-local fixes it though :/ | 17:41 |
| jroll | lucasagomes: right, localrc could change it | 17:41 |
| JayF | OH | 17:41 |
| lucasagomes | I see | 17:41 |
| lucasagomes | shell dark magic | 17:42 |
| JayF | jroll: because if you have local blah=`blah`; echo $?; it returns the $? of local | 17:42 |
| JayF | jroll: blah=`blah`; echo $?; it returns the $? of `blah` | 17:42 |
| jroll | JayF: ah, yeah | 17:42 |
| BadCub | moring folks | 17:44 |
| jlvillal | lucasagomes, On the bashate stuff. Would it make any sense to maybe create a new bash script which runs bashate? Instead of sticking it inside tox.ini? | 17:44 |
| jlvillal | BadCub, morning! | 17:46 |
| *** alexpilotti has joined #openstack-ironic | 17:46 | |
| lucasagomes | jlvillal, hmm becuase that bash -c "find ..." command is too ugly? | 17:46 |
| lucasagomes | I'm good either way, don't have a strong opinion on it | 17:47 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Run bashate as part of the pep8 command https://review.openstack.org/258506 | 17:47 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Fix syntax errors in the shell scripts under devstack/tools https://review.openstack.org/258133 | 17:47 |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Fix bashate errors in grenade plugin https://review.openstack.org/259129 | 17:47 |
| jroll | ^ fixed and rebased | 17:47 |
| jlvillal | lucasagomes, sort of ugly. But I just thought it might allow it to become more flexible. Heck maybe even a Python script :) | 17:48 |
| jlvillal | lucasagomes, But either way works | 17:48 |
| lucasagomes | jroll, cheers | 17:48 |
| jroll | np | 17:49 |
| jroll | jlvillal: let's do it when we feel it's needed, rather than just in case | 17:49 |
| *** vinm213 has quit IRC | 17:49 | |
| jlvillal | jroll, Okay :) | 17:50 |
| *** praneshp has joined #openstack-ironic | 17:51 | |
| *** alexpilotti has quit IRC | 17:51 | |
| *** electrofelix has quit IRC | 17:51 | |
| jroll | jlvillal: fwiw, I tend to think that with most changes | 17:51 |
| jroll | too much to do to worry about things like that | 17:51 |
| lucasagomes | jroll, tested with the new patch wfm | 17:52 |
| lucasagomes | would be good to get that chain in so we avoid more syntax error stuff in the bash scripts | 17:52 |
| lucasagomes | specially now that people are working hard on tests | 17:52 |
| jroll | lucasagomes: yeah +1 | 17:52 |
| jroll | lucasagomes: do you want to address this or leave it for now? https://review.openstack.org/#/c/258133/3/tox.ini | 17:53 |
| *** e0ne has quit IRC | 17:53 | |
| BadCub | hey hey jlvillal :) | 17:53 |
| lucasagomes | jroll, to rename it? | 17:54 |
| lucasagomes | ok... lemme quickly do it | 17:54 |
| jroll | yeah | 17:54 |
| jroll | I +2d anyway | 17:54 |
| jroll | just wanted to ask | 17:54 |
| *** ukalifon1 has joined #openstack-ironic | 17:54 | |
| jroll | later patch is fine too | 17:54 |
| lucasagomes | jroll, ah, so go for it... we can do it at the end | 17:54 |
| lucasagomes | cause I +2 the other patch after that as well | 17:54 |
| * lucasagomes adds to the end of the chain | 17:54 | |
| jroll | lucasagomes: yeah, do it at the end and also flake8 it | 17:54 |
| lucasagomes | jroll, ++ fair enough | 17:55 |
| * lucasagomes is on it | 17:55 | |
| jroll | k, +2 on the 2 you posted | 17:55 |
| lucasagomes | thx | 17:55 |
| * jroll runs off to a quick meeting | 17:55 | |
| *** [1]cdearborn has quit IRC | 17:56 | |
| thiagop | folks, going away for the holidays | 18:00 |
| thiagop | May you have a wonderful christmas and a happy new year! | 18:00 |
| *** derekh has quit IRC | 18:00 | |
| thiagop | See you on the other side ;) | 18:00 |
| *** thiagop has quit IRC | 18:01 | |
| *** bizarrochristy has joined #openstack-ironic | 18:01 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Devstack: Remove QEMU hook at ./unstack https://review.openstack.org/259104 | 18:03 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Support reboot_requested bool on agent clean_steps https://review.openstack.org/259135 | 18:03 |
| *** ndipanov has quit IRC | 18:03 | |
| *** athomas has quit IRC | 18:04 | |
| NobodyCam | see ya next year thiagop | 18:05 |
| *** cemason has quit IRC | 18:05 | |
| *** rajathagasthya has joined #openstack-ironic | 18:06 | |
| *** penick has joined #openstack-ironic | 18:07 | |
| *** davideagnello has joined #openstack-ironic | 18:11 | |
| *** praneshp has quit IRC | 18:14 | |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Streamline pip installs https://review.openstack.org/259138 | 18:16 |
| *** alexpilotti has joined #openstack-ironic | 18:17 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Add extensions to the scripts at devstack/tools/ironic/scripts https://review.openstack.org/259139 | 18:18 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Run flake8 against the python scripts under tools/ and devstack/tools https://review.openstack.org/259140 | 18:18 |
| lucasagomes | jroll, tan<tab> ^ | 18:18 |
| *** alexpilotti has quit IRC | 18:22 | |
| *** ukalifon1 has quit IRC | 18:25 | |
| *** harlowja has joined #openstack-ironic | 18:27 | |
| *** pas-ha has quit IRC | 18:31 | |
| jroll | lucasagomes: it's lintan :P | 18:33 |
| lucasagomes | oh | 18:33 |
| jroll | +2 on both, thanks man | 18:35 |
| * mgould -> home; good night! | 18:35 | |
| NobodyCam | night mgould | 18:36 |
| *** mgould has quit IRC | 18:36 | |
| *** devananda has quit IRC | 18:36 | |
| *** Marga_ has quit IRC | 18:36 | |
| *** aginwala has joined #openstack-ironic | 18:36 | |
| *** mkovacik has joined #openstack-ironic | 18:37 | |
| lucasagomes | ty! | 18:38 |
| JayF | I just had an idea w/r/t the gate | 18:38 |
| JayF | Have we checked download times pre and post stream_raw_images? | 18:38 |
| JayF | I'd imagine VM disks are pretty damn slow too | 18:38 |
| JayF | although that wouldn't impact boot times for the ramdisk, so probably a bad idea | 18:39 |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Streamline pip installs https://review.openstack.org/259138 | 18:41 |
| *** sambetts is now known as sambetts-afk | 18:41 | |
| jroll | yeah, that's something to look at though | 18:41 |
| jroll | although vm ram is also slow | 18:41 |
| jroll | I imagine streaming would be faster, either way you have to write it to disk | 18:42 |
| lucasagomes | JayF, right, something to investigate yeah... tho we are talking about a 25M image | 18:42 |
| JayF | yeah; the ipa ramdisk is much larger | 18:42 |
| jroll | :( | 18:42 |
| lucasagomes | yeah | 18:42 |
| * jroll watches tinyipa patch | 18:42 | |
| lucasagomes | JayF, jroll we have to start biulding and distributing tiny ipa images | 18:43 |
| jroll | "tiny ipa" or "tinyipa" | 18:44 |
| jroll | :P | 18:44 |
| JayF | +1 | 18:44 |
| lucasagomes | so we don't have to build them in gate (unless necessary, like -src jobs) | 18:44 |
| lucasagomes | jroll, haha idk, tinyipa | 18:44 |
| JayF | we don't build anything in gate ouside of -src jobs | 18:44 |
| lucasagomes | ++ | 18:44 |
| *** praneshp has joined #openstack-ironic | 18:44 | |
| JayF | the ironic gates use the prebuilt images, I thought? | 18:44 |
| JayF | unless that changed | 18:44 |
| lucasagomes | JayF, it does, in a post job right? | 18:44 |
| jroll | lucasagomes: fwiw on sam's patch I just wanted to see what gate jobs look like with it, it's probably a WIP | 18:44 |
| jroll | JayF: they do | 18:44 |
| lucasagomes | but I think it only builds the coreos one | 18:44 |
| JayF | yep | 18:44 |
| lucasagomes | jroll, yeah | 18:45 |
| JayF | I built all that stuff, just didn't know if it had been replaced | 18:45 |
| *** rloo_afk is now known as rloo | 18:45 | |
| *** e0ne has joined #openstack-ironic | 18:46 | |
| jroll | oh, this exists, cool https://review.openstack.org/#/c/234902/ | 18:46 |
| *** trown|lunch is now known as trown | 18:47 | |
| *** devananda has joined #openstack-ironic | 18:52 | |
| *** Marga_ has joined #openstack-ironic | 18:53 | |
| lucasagomes | JayF, I nitpick'ed the clean/deploy power off test patch... (and it's missing tests for the power_off_after_failure=False) | 18:53 |
| lucasagomes | other than that it LGTM | 18:54 |
| JayF | lucasagomes: I literally just did patch -p1 .../patches/that_patch.patch and pushed it up :) No complaints from me if it still needs a little work | 18:54 |
| lucasagomes | cool | 18:54 |
| JayF | ^ < | 18:54 |
| lucasagomes | yeah missing that test... also would be good to the config option to not power off if deploy fails to indicate that it's about deployment in the name | 18:55 |
| lucasagomes | and not a generic failure | 18:55 |
| *** aginwala has quit IRC | 18:56 | |
| JayF | what do you mean by that? | 18:57 |
| *** Marga_ has quit IRC | 18:57 | |
| *** aginwala has joined #openstack-ironic | 18:59 | |
| lucasagomes | JayF, I mean that, instead of calling it power_off_after_failure we should call it power_off_after_deploy_failure | 18:59 |
| lucasagomes | same of the help string | 18:59 |
| lucasagomes | to meantion deployment there, because set_failed_state() is to set fail state for a deployment failure (deploy_utils.py) | 19:00 |
| JayF | if you get all that on the review, I'll update the patch | 19:00 |
| *** bizarrochristy has quit IRC | 19:00 | |
| * JayF gone for a couple of hours | 19:00 | |
| lucasagomes | JayF, it's there | 19:00 |
| lucasagomes | JayF, no rush | 19:00 |
| lucasagomes | for the help* | 19:01 |
| openstackgerrit | Merged openstack/python-ironicclient: Removes MANIFEST.in as it is not needed explicitely by PBR https://review.openstack.org/258387 | 19:02 |
| *** aginwala has quit IRC | 19:04 | |
| *** cemason has joined #openstack-ironic | 19:07 | |
| *** cemason has quit IRC | 19:12 | |
| *** aginwala has joined #openstack-ironic | 19:14 | |
| *** absubram has joined #openstack-ironic | 19:15 | |
| *** aginwala has quit IRC | 19:17 | |
| *** aginwala has joined #openstack-ironic | 19:18 | |
| *** rajathagasthya has quit IRC | 19:18 | |
| *** mannidi has quit IRC | 19:20 | |
| eggmaster | I'm trying to recreate an error I saw on a recent deployment where a customer's instack.json file had a couple MAC addresses transposed. It caused PXE booting to fail only when not all the nodes were used in a deployment. The twist is that I'm trying to recreate it in a virthost-based environment, i.e. the ironic nodes are VMs. The instackenv.json file I have has identical pm_addr for all the nodes. I'm trying to understand how pxe_ssh | 19:32 |
| eggmaster | driver works and if I'm on a wild goose chase (i.e. maybe this is something I have to do with BM). TIA. | 19:32 |
| *** e0ne has quit IRC | 19:34 | |
| *** ukalifon1 has joined #openstack-ironic | 19:37 | |
| lucasagomes | eggmaster, https://github.com/openstack/os-cloud-config/blob/master/os_cloud_config/nodes.py#L50-L53 | 19:42 |
| lucasagomes | pm_addr will be the ip of the host | 19:43 |
| eggmaster | oh, the virthost. | 19:43 |
| eggmaster | I gets it now. | 19:43 |
| lucasagomes | pm_user the user name to use in the ssh command, and pm_password is the key contents (which is weird because we support ssh_password) | 19:43 |
| lucasagomes | but anyway | 19:43 |
| lucasagomes | eggmaster, cool | 19:44 |
| eggmaster | yeah so, baremetal it is. | 19:44 |
| lucasagomes | eggmaster, I've seem a similar problem to what you described | 19:45 |
| lucasagomes | the reason was because the mac addresses registered in the nodes belonged to other nodes | 19:45 |
| eggmaster | lucasagomes: restating to make sure I understand. pxe_ssh will just ssh to the virthost to bring the ironic nodes up/down, as opposed to impi on baremetal which will actually bring stuff up/down and then wait for them to pxe boot. | 19:45 |
| lucasagomes | so when all machines where being deployed the DHCP server had rules to all MACs | 19:45 |
| eggmaster | lucasagomes: yes that's exactly it. | 19:46 |
| lucasagomes | but when only few were deployed some would fail, because the mac configured wasn't exactly the mac address of those nodes booting | 19:46 |
| eggmaster | lucasagomes: when we tried to debug with 1,1 deploy, it was saying 'no leases available' | 19:46 |
| lucasagomes | eggmaster, that understading is correct | 19:46 |
| lucasagomes | eggmaster, and the reason why we have a ssh driver like that is because we don't test on real machines in gate | 19:46 |
| lucasagomes | eggmaster, so we need to "mock" ipmi somehow | 19:47 |
| eggmaster | lucasagomes: I asked for RFE to network validation script to check for this b/c we wasted so much time debugging it onsite, but I also wanted to create some CI to verify that validate script caught that type of error | 19:47 |
| lucasagomes | eggmaster, ++ | 19:47 |
| eggmaster | that's my motivation here. seems like baremetal is only way to go on this. | 19:47 |
| lucasagomes | those validation tools like that are external things | 19:48 |
| lucasagomes | we won't deal that with Ironic itself | 19:48 |
| jroll | yeah, you have to make very sure all MACs are accurate and match the same server as BMC IP, we even go as far as having datacenter people walk through and validate mac addresses just to be sure | 19:48 |
| lucasagomes | eggmaster, you may want to talk to athomas about it | 19:48 |
| eggmaster | lucasagomes: he's aware of it | 19:48 |
| lucasagomes | eggmaster, cool | 19:48 |
| openstackgerrit | Milan Kováčik proposed openstack/ironic-inspector-specs: High Availability for Ironic Inspector https://review.openstack.org/253675 | 19:49 |
| lucasagomes | eggmaster, one problem you will have with the virtual setup is that, we use the MAC address to identify which machine to power on and off | 19:49 |
| *** piet has joined #openstack-ironic | 19:50 | |
| lucasagomes | so in the virsh setup it will power on the wrong node if the mac is incorrect | 19:50 |
| lucasagomes | you will want to use baremetal (or hack the virsh driver to use the VM name?!) instead | 19:50 |
| *** ukalifon1 has quit IRC | 19:51 | |
| eggmaster | ack | 19:52 |
| lucasagomes | jroll, when you have time, any thoughts on sean comments at https://review.openstack.org/#/c/221848/ ? | 19:53 |
| lucasagomes | jroll, pin the nova driver to a specific API version in Ironic | 19:53 |
| jroll | lucasagomes: I'll comment, I think we need the config | 19:54 |
| jroll | lucasagomes: which will allow people to configure it to work with their version of ironic | 19:54 |
| jroll | and the code to adapt to different versions | 19:54 |
| lucasagomes | jroll, yeah, I like that approach better | 19:55 |
| lucasagomes | it's more flexible | 19:55 |
| lucasagomes | and I will call it a day | 19:56 |
| lucasagomes | have a good evening (ugt) everyone! | 19:56 |
| *** lucasagomes is now known as lucas-dinner | 19:57 | |
| NobodyCam | night lucas-dinner | 19:58 |
| devananda | jroll: I'm inclined to agree with sean. a given version of the nova code could (and should) be pinned to an ironic API version | 19:59 |
| devananda | jroll: the supported upgrade path is: upgrade ironic, then upgrade nova. so pinning the ironic api version w/in the nova virt driver should be fine, and it doesn't need to be configurable | 20:00 |
| NobodyCam | morning devananda :) | 20:00 |
| devananda | in fact, allowing a single version of hte nova virt driver to interact at different ironic API versions will significantly complicate the nova virt driver code | 20:00 |
| jroll | devananda: part of the reason for microversions is to allow applications to figure out the capabilities of the API, and do different things based on that | 20:00 |
| devananda | jroll: uhh | 20:01 |
| jroll | he might be right in that it doesn't need to be a config | 20:01 |
| jroll | but it should be able to negotiate | 20:01 |
| jroll | (which I realized while beginning to reply on that patch | 20:01 |
| devananda | jroll: yes, and also to decouple the upgrading of a server from the upgrading of a client (or in this case another service that is a client) | 20:01 |
| jroll | devananda: right | 20:01 |
| jroll | so I actually now agree, not a config | 20:01 |
| devananda | the nova operator derives no benefit from configuring it | 20:01 |
| jroll | but rather negotiation | 20:01 |
| jroll | but not pinning. | 20:02 |
| devananda | but the virt driver does need to negotiate it | 20:02 |
| devananda | well, pinning seems fine to me, actually | 20:02 |
| jroll | you can't negotiate if it's pinned | 20:02 |
| devananda | eg, pin the liberty release of hte nova virt driver to X.XX, pin the Mitaka release to X.YY, | 20:02 |
| jroll | that means I can't run mitaka nova with liberty ironic, for example | 20:02 |
| jroll | which is pretty lame IMO | 20:03 |
| devananda | where X.XX == the liberty release of ironic's API, and X.YY == the mitaka release of ironic's API | 20:03 |
| devananda | jroll: well, yes | 20:03 |
| devananda | like i said, the supported upgrade path is to upgrade ironic first | 20:03 |
| jroll | so why artificially limit that, if we can negotiate it | 20:03 |
| jroll | I also think that's lame and want to nuke that assertion. | 20:03 |
| devananda | newer nova + older ironic => unsupported, at least previously | 20:03 |
| devananda | ahh | 20:03 |
| devananda | well then, that should be called out in docs, testing, etc | 20:03 |
| devananda | it was not my assumption | 20:04 |
| jroll | and when we can discover the version of ironic, we can get rid of that requirement | 20:04 |
| *** upma has joined #openstack-ironic | 20:07 | |
| *** toure is now known as toure|biab | 20:08 | |
| *** aginwala has quit IRC | 20:09 | |
| piet | The OpenStack UX team is conducting a series of interviews focused on hardware provisioning with the intent of using customer feedback to help drive the development of the OpenStack Ironic project. Individuals using MaaS, Cobbler and Foreman are also welcome to participate in the study. | 20:09 |
| piet | If interested, please select a date/time from the following link http://doodle.com/poll/8a7ftkqe4d2bacts | 20:10 |
| JayF | For what purposes? Horizon? | 20:10 |
| upma | Hi I have created a baremetal vm from nova and state is still spawning from last 40 minutes. | 20:10 |
| upma | How long does it take to boot? | 20:10 |
| piet | jayf: Horizon plugin and a standalone | 20:11 |
| piet | Primary focus is on the standalone | 20:12 |
| upma | NobodyCam: Hi | 20:12 |
| TheJulia | piet: infracloud and gozer folks would be good involve | 20:12 |
| piet | Its a good opportunity for operators to have a significant impact on the interface | 20:12 |
| jroll | TheJulia: infracloud folks use GUIs? | 20:13 |
| TheJulia | jroll: no, but they might have opinions | 20:13 |
| TheJulia | :) | 20:13 |
| jroll | :P | 20:13 |
| piet | heh | 20:13 |
| upma | jroll: Hi | 20:13 |
| piet | thejulia; let me find the channels and I'll invite them | 20:13 |
| jroll | hi upma | 20:14 |
| upma | jroll: I have created a baremetal vm from nova and state is still spawning from last 40 minutes. | 20:14 |
| upma | How long does it take in usual? | 20:14 |
| TheJulia | piet: gozer is the internal infra team at HPE, I can put you in touch with them | 20:14 |
| jroll | upma: it probably isn't pxe booting correctly, check serial console and such | 20:14 |
| jroll | upma: it shouldn't take that long | 20:14 |
| piet | thejulia: that would be awesone | 20:15 |
| piet | thejulia; having trouble finding folks | 20:15 |
| upma | jroll: Ok I am checking | 20:16 |
| BadCub | piet: are you on HipChat? | 20:17 |
| TheJulia | BadCub: piet is no longer with HPE | 20:17 |
| BadCub | Oh.. ugh... I am so out of touch | 20:17 |
| rloo | hey jroll, are/have you thought about splitting the claims/filter API spec into two, and allowing for claims w/o filteriing? | 20:17 |
| * BadCub crawls back under his rock | 20:17 | |
| jroll | rloo: like, claim by node uuid? | 20:17 |
| rloo | jroll: right | 20:17 |
| TheJulia | BadCub: yeah, you were out then... | 20:17 |
| jroll | rloo: not really, but I'd be open to making {uuid: <bla>} a valid claims filter | 20:18 |
| rloo | jroll: you haven't responded for awhile on that spec and curious minds were wondering | 20:18 |
| BadCub | ahh, that 'splains it | 20:18 |
| jroll | rloo: we've been in email trying to tell people why that spec isn't terrible | 20:18 |
| TheJulia | piet: sent you names in /msg | 20:18 |
| jroll | rloo: I'm not going to work on it if people are goin to sit here blocking it | 20:18 |
| rloo | jroll: I know. | 20:19 |
| rloo | jroll: you want me to tell them? :) | 20:19 |
| jroll | rloo: sure, I'm curious what the use case is as you can deploy to a node without a claim | 20:19 |
| rloo | jroll: my understanding is to change the scheduler to claim a node, and then pass the claim to compute so it doesn't block compute part | 20:20 |
| *** sinh has joined #openstack-ironic | 20:20 | |
| NobodyCam | hey hey upma | 20:20 |
| jroll | rloo: oh, so hack the scheduler instead of doing this thing? | 20:21 |
| piet | thejulia: looking now | 20:21 |
| rloo | jroll: if that's what you call it. In the interim anyway. | 20:21 |
| jroll | rloo: gotcha | 20:21 |
| upma | Hi NobodyCam | 20:21 |
| rloo | jroll: it should be faster to get the claims API working, before the filtering works etc. | 20:22 |
| NobodyCam | :) | 20:22 |
| upma | jroll answered my question, thanks for coming up. | 20:22 |
| jroll | rloo: I'd prefer if people just worked with us to solve their use case in the real world :/ | 20:22 |
| jroll | rloo: s/real/new/ | 20:22 |
| upma | I will get back checking serial console NobodyCam | 20:22 |
| *** Marga_ has joined #openstack-ironic | 20:22 | |
| rloo | jroll: i think they do want to work but it isn't clear how the new world will solve their problems in the short or near term. | 20:22 |
| jroll | rloo: which is what long deprecation periods are for :) | 20:23 |
| jroll | but yeah | 20:23 |
| jroll | it's fine | 20:23 |
| rloo | jroll: so do you want me to get some of them to unblock the spec, and/or would you be good with splitting that spec into two? | 20:24 |
| rloo | jroll: and why is it called 'claim' instead of 'reservation'? (my curiosity on that) | 20:24 |
| rloo | jroll: not that I can 'get' them to do anything but I can try :) | 20:24 |
| jroll | rloo: because we already have node.reservation | 20:25 |
| *** rajathagasthya has joined #openstack-ironic | 20:26 | |
| rloo | jroll: Oh, right! thx. | 20:26 |
| jroll | rloo: I'll respond to comments and update it to make us able to land the claims api with uuid first | 20:26 |
| jroll | and then we can keep adding filters | 20:26 |
| jroll | does that work? | 20:26 |
| rloo | thx jroll! that'd be great! | 20:26 |
| jroll | np :) | 20:27 |
| jroll | need to get this done soon | 20:27 |
| *** e0ne has joined #openstack-ironic | 20:27 | |
| *** aginwala has joined #openstack-ironic | 20:28 | |
| *** aginwala has quit IRC | 20:29 | |
| *** alexpilotti has joined #openstack-ironic | 20:34 | |
| *** alexpilotti has quit IRC | 20:38 | |
| upma | NobodyCam: I am not sure why I am not able to create vm now. I am getting "no valid host". | 20:47 |
| upma | NobodyCam: I found no error on ironic-api or ironic-conductoe | 20:47 |
| *** Pespa has quit IRC | 20:48 | |
| NobodyCam | can you paste (paste.openstack.org) a copy of the command you are running and the error? | 20:48 |
| upma | Yes sure | 20:48 |
| upma | NobodyCam: http://paste.openstack.org/show/482227/ | 20:50 |
| * NobodyCam clicks | 20:50 | |
| NobodyCam | upma: how about a ironic node-list and nova hypervisor-stats | 20:52 |
| *** piet has quit IRC | 20:53 | |
| upma | NobodyCam: http://paste.openstack.org/show/482228/ | 20:53 |
| NobodyCam | upma: okay I see why you got the error | 20:55 |
| upma | NobodyCam: ironic node-show http://paste.openstack.org/show/482229/ | 20:55 |
| NobodyCam | you have two node in ironics inventory and neither is avaidable | 20:55 |
| upma | Yes node-0 is clean failed | 20:55 |
| upma | These nodes are vm or what? | 20:56 |
| NobodyCam | that I'm not sure. | 20:56 |
| upma | np :) NobodyCam | 20:56 |
| NobodyCam | what does ironic node-show node-0 show | 20:56 |
| *** daemontool has quit IRC | 20:57 | |
| *** piet has joined #openstack-ironic | 20:58 | |
| *** daemontool has joined #openstack-ironic | 20:58 | |
| upma | NobodyCam: http://paste.openstack.org/show/482229/ | 20:58 |
| NobodyCam | upma: did you see lines 21 & 22? | 20:59 |
| upma | Yes | 21:00 |
| *** aginwala has joined #openstack-ironic | 21:01 | |
| *** raildo is now known as raildo-afk | 21:01 | |
| upma | NobodyCam: This was because I created a vm from node-0 and that timed out | 21:02 |
| *** rbudden has joined #openstack-ironic | 21:04 | |
| upma | NobodyCam: there? | 21:06 |
| *** mbound_ has joined #openstack-ironic | 21:07 | |
| *** mbound_ has quit IRC | 21:08 | |
| *** rbudden has quit IRC | 21:09 | |
| NobodyCam | upma: sorry several windows going at once | 21:12 |
| upma | NobodyCam: No problem. I will be here only. Please ping me when you get time | 21:12 |
| NobodyCam | I am back | 21:12 |
| NobodyCam | so these nodes are real HW or VM acting as HW? | 21:13 |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Streamline pip installs https://review.openstack.org/259138 | 21:14 |
| NobodyCam | ie I dont understand | 21:14 |
| NobodyCam | "I created a vm from node-0" | 21:14 |
| NobodyCam | does that mean you nova booted node-0? | 21:14 |
| *** mbound_ has joined #openstack-ironic | 21:16 | |
| upma | Yes | 21:16 |
| upma | I created a node in ironic: node-0 | 21:16 |
| NobodyCam | okay :) | 21:16 |
| upma | And then booted a vm from nova using baremetal flavor | 21:17 |
| NobodyCam | which I assume was node-1 | 21:17 |
| upma | No a kvm instance named new | 21:18 |
| upma | Sorry I am confused | 21:18 |
| NobodyCam | can you paste a nova list ? | 21:19 |
| upma | Yes | 21:19 |
| NobodyCam | you manually started a kvm guest named new? | 21:20 |
| upma | http://paste.openstack.org/show/482237/ | 21:20 |
| upma | Yes using nova boot | 21:20 |
| *** rajathagasthya has quit IRC | 21:21 | |
| NobodyCam | is nova configured to use ironic? Ie http://docs.openstack.org/developer/ironic/deploy/install-guide.html#configure-compute-to-use-the-bare-metal-service | 21:23 |
| *** rajathagasthya has joined #openstack-ironic | 21:23 | |
| *** cemason has joined #openstack-ironic | 21:24 | |
| *** aginwala has quit IRC | 21:24 | |
| upma | NobodyCam: http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#deploying-ironic-with-devstack I followed this doc to setup ironic | 21:24 |
| upma | I will check my nova.conf | 21:24 |
| *** krtaylor has quit IRC | 21:24 | |
| NobodyCam | okay so your running devstack | 21:25 |
| *** rajathagasthya has quit IRC | 21:26 | |
| * NobodyCam notes his new alienware was just delivered | 21:27 | |
| upma | NobodyCam: Yes | 21:27 |
| upma | NobodyCam: I see last 3 conf parameters are not defined | 21:27 |
| NobodyCam | upma: I would say dbl check the entire setup / config and retry | 21:28 |
| upma | Ok | 21:28 |
| upma | I will give it a try | 21:28 |
| *** aginwala has joined #openstack-ironic | 21:28 | |
| upma | NobodyCam: brb | 21:28 |
| *** cemason has quit IRC | 21:28 | |
| *** mkovacik has quit IRC | 21:32 | |
| *** piet has quit IRC | 21:32 | |
| *** krtaylor has joined #openstack-ironic | 21:37 | |
| *** rajathagasthya has joined #openstack-ironic | 21:41 | |
| mrda | Morning Ironic | 21:47 |
| BadCub | hiya mrda | 21:50 |
| *** mannidi has joined #openstack-ironic | 21:52 | |
| mrda | BadCub: o/ | 21:52 |
| NobodyCam | morning mrda | 21:52 |
| BadCub | how are things on teh wrong side of the earth? | 21:52 |
| mrda | NobodyCam: \o | 21:52 |
| mrda | BadCub: It's definitely summer here. 95C over night, and today will be out 3rd day in two weeks over 110F during the day. | 21:53 |
| BadCub | EEEK | 21:53 |
| BadCub | that is the same as our summers in Palm Springs | 21:54 |
| BadCub | that we are never here for lol | 21:54 |
| mrda | BadCub: So us Aussies are hitting the beach | 21:54 |
| BadCub | sweet | 21:54 |
| mrda | One day I should come visit Palm Springs, but not in the summer :) | 21:55 |
| *** upma has quit IRC | 21:55 | |
| BadCub | no summer here sucks | 21:56 |
| BadCub | can be 115 and raining | 21:56 |
| mrda | Eeew | 21:56 |
| BadCub | yeah. monsoon season can be real fun. Another reason for us not being here then | 21:57 |
| *** piet has joined #openstack-ironic | 21:59 | |
| *** mannidi has quit IRC | 22:02 | |
| *** aginwala_ has joined #openstack-ironic | 22:04 | |
| *** aginwala has quit IRC | 22:07 | |
| *** rajathagasthya has quit IRC | 22:10 | |
| *** trown is now known as trown|outttypeww | 22:10 | |
| *** rajathagasthya has joined #openstack-ironic | 22:13 | |
| *** Pespa has joined #openstack-ironic | 22:14 | |
| *** rajathagasthya has quit IRC | 22:14 | |
| *** rajathagasthya has joined #openstack-ironic | 22:15 | |
| *** aginwala_ has quit IRC | 22:21 | |
| *** aginwala has joined #openstack-ironic | 22:22 | |
| *** aginwala has quit IRC | 22:24 | |
| *** aginwala has joined #openstack-ironic | 22:25 | |
| *** aginwala has quit IRC | 22:27 | |
| *** aginwala has joined #openstack-ironic | 22:27 | |
| *** priteau has quit IRC | 22:27 | |
| *** aginwala has quit IRC | 22:32 | |
| *** aginwala has joined #openstack-ironic | 22:33 | |
| *** absubram has quit IRC | 22:36 | |
| *** Pespa has quit IRC | 22:45 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Fix python3 compatability for CSV reader support https://review.openstack.org/259207 | 22:46 |
| *** e0ne has quit IRC | 22:47 | |
| *** piet has quit IRC | 22:51 | |
| *** rajathagasthya has quit IRC | 22:54 | |
| *** jaybeale has joined #openstack-ironic | 22:59 | |
| openstackgerrit | Ruby Loo proposed openstack/python-ironicclient: Add 'node-set-provision-state <node> clean' https://review.openstack.org/258694 | 23:06 |
| *** yuanying has joined #openstack-ironic | 23:12 | |
| *** aginwala has quit IRC | 23:21 | |
| *** aginwala has joined #openstack-ironic | 23:29 | |
| *** dprince has quit IRC | 23:39 | |
| *** jaypipes has quit IRC | 23:47 | |
| *** mbound_ has quit IRC | 23:51 | |
| *** naohirot has joined #openstack-ironic | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!