| openstackgerrit | A change was merged to openstack/ironic: Add set_spawn_error_hook to TaskManager https://review.openstack.org/100957 | 00:01 |
|---|---|---|
| *** Haomeng|2 has joined #openstack-ironic | 00:03 | |
| *** Haomeng has quit IRC | 00:04 | |
| openstackgerrit | A change was merged to openstack/ironic: Add more INFO logging to ironic/conductor https://review.openstack.org/99744 | 00:05 |
| rloo | devananda: want to make sure you know that this change to the nova.virt.ironic.driver was approved/merged (since you didn't approve it): https://review.openstack.org/#/c/105050/ | 00:14 |
| JayF | rloo: am I missing something or was that also approved with only 1 +2? | 00:19 |
| rloo | JayF: yeah, I saw that too. | 00:19 |
| rloo | JayF: money changing hands there ;) | 00:20 |
| *** hemna is now known as hemna_ | 00:40 | |
| *** dkehnx1 has quit IRC | 00:41 | |
| *** dwalleck has joined #openstack-ironic | 00:44 | |
| *** dwalleck_ has joined #openstack-ironic | 00:47 | |
| *** dwalleck has quit IRC | 00:47 | |
| *** Penick has joined #openstack-ironic | 00:52 | |
| *** jerryz has joined #openstack-ironic | 01:18 | |
| *** max_lobur has quit IRC | 01:19 | |
| *** nosnos has joined #openstack-ironic | 01:42 | |
| *** killer_prince is now known as lazy_prince | 02:03 | |
| *** eghobo has quit IRC | 02:15 | |
| *** lazy_prince is now known as killer_prince | 02:25 | |
| *** rloo has quit IRC | 02:32 | |
| *** ramineni has joined #openstack-ironic | 02:55 | |
| *** nosnos has quit IRC | 03:22 | |
| *** Penick has quit IRC | 03:27 | |
| *** zer0c00l has joined #openstack-ironic | 03:33 | |
| *** rameshg87_afk is now known as rameshg87 | 03:44 | |
| *** rameshg87 has left #openstack-ironic | 03:47 | |
| *** rameshg87 has quit IRC | 03:47 | |
| *** rameshg87 has joined #openstack-ironic | 03:50 | |
| *** coolsvap|afk is now known as coolsvap | 03:56 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Updated from global requirements https://review.openstack.org/105167 | 04:00 |
| openstackgerrit | Ramakrishnan G proposed a change to openstack/ironic-specs: iLO Virtual Media Deploy Driver https://review.openstack.org/97744 | 04:08 |
| *** pcrews has quit IRC | 04:10 | |
| *** killer_prince is now known as lazy_prince | 04:12 | |
| *** pradipta_away is now known as pradipta | 04:17 | |
| *** aswadr has joined #openstack-ironic | 04:20 | |
| *** Nisha has joined #openstack-ironic | 04:20 | |
| *** geekyogi has quit IRC | 04:24 | |
| *** sseago has joined #openstack-ironic | 04:25 | |
| *** ajc_ has joined #openstack-ironic | 04:25 | |
| *** nosnos has joined #openstack-ironic | 04:25 | |
| *** sabah has joined #openstack-ironic | 04:27 | |
| *** eghobo has joined #openstack-ironic | 04:27 | |
| *** sabah has quit IRC | 04:30 | |
| *** sseago has quit IRC | 04:30 | |
| *** sabah has joined #openstack-ironic | 04:31 | |
| *** dwalleck_ has quit IRC | 04:37 | |
| *** rakesh_hs has joined #openstack-ironic | 04:45 | |
| *** dwalleck has joined #openstack-ironic | 04:47 | |
| *** k4n0 has joined #openstack-ironic | 04:50 | |
| *** amitpp has joined #openstack-ironic | 04:50 | |
| *** sseago has joined #openstack-ironic | 04:51 | |
| *** lazy_prince has quit IRC | 04:53 | |
| *** dwalleck has quit IRC | 04:55 | |
| *** dwalleck has joined #openstack-ironic | 04:59 | |
| *** lazy_prince has joined #openstack-ironic | 05:01 | |
| *** max_lobur has joined #openstack-ironic | 05:03 | |
| *** pradipta is now known as pradipta_away | 05:10 | |
| *** geekyogi has joined #openstack-ironic | 05:10 | |
| *** sabah has quit IRC | 05:13 | |
| *** sabah has joined #openstack-ironic | 05:13 | |
| *** subah has joined #openstack-ironic | 05:14 | |
| *** sabah has quit IRC | 05:14 | |
| *** dwalleck has quit IRC | 05:15 | |
| *** dwalleck has joined #openstack-ironic | 05:15 | |
| *** bvivek has joined #openstack-ironic | 05:19 | |
| *** killer_prince has joined #openstack-ironic | 05:19 | |
| *** vinbs has joined #openstack-ironic | 05:22 | |
| *** lazy_prince has quit IRC | 05:26 | |
| *** killer_prince is now known as lazy_prince | 05:26 | |
| *** dwalleck_ has joined #openstack-ironic | 05:28 | |
| vinbs | Morning Ironic! | 05:29 |
| *** dwalleck has quit IRC | 05:29 | |
| *** sseago has quit IRC | 05:34 | |
| rameshg87 | good morning vinbs | 05:38 |
| *** dwalleck_ has quit IRC | 05:41 | |
| *** harlowja is now known as harlowja_away | 05:42 | |
| *** sseago has joined #openstack-ironic | 05:46 | |
| *** eghobo has quit IRC | 05:50 | |
| *** eghobo has joined #openstack-ironic | 05:51 | |
| *** Ng_ has joined #openstack-ironic | 05:56 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/105359 | 06:02 |
| *** max_lobur has quit IRC | 06:08 | |
| Nisha | jroll: NobodyCam hi | 06:20 |
| *** zigo has quit IRC | 06:27 | |
| *** zigo has joined #openstack-ironic | 06:31 | |
| *** lazy_prince has quit IRC | 06:32 | |
| *** lazy_prince has joined #openstack-ironic | 06:50 | |
| *** jcoufal has joined #openstack-ironic | 07:01 | |
| *** eguz has joined #openstack-ironic | 07:30 | |
| *** eghobo has quit IRC | 07:32 | |
| *** jistr has joined #openstack-ironic | 07:56 | |
| *** pradipta_away is now known as pradipta | 07:58 | |
| *** killer_prince has joined #openstack-ironic | 08:04 | |
| *** derekh_ has joined #openstack-ironic | 08:16 | |
| *** ndipanov has joined #openstack-ironic | 08:19 | |
| *** wanyen has quit IRC | 08:19 | |
| *** lucasagomes has joined #openstack-ironic | 08:21 | |
| *** lazy_prince has quit IRC | 08:26 | |
| *** killer_prince is now known as lazy_prince | 08:26 | |
| *** overlayer has joined #openstack-ironic | 08:26 | |
| *** athomas has quit IRC | 08:27 | |
| *** martyntaylor has joined #openstack-ironic | 08:28 | |
| *** athomas has joined #openstack-ironic | 08:30 | |
| *** overlayer has quit IRC | 08:40 | |
| *** igordcard has joined #openstack-ironic | 08:40 | |
| *** pradipta is now known as pradipta_away | 09:01 | |
| *** Ng has quit IRC | 09:02 | |
| *** Ng_ is now known as Ng | 09:02 | |
| *** pelix has joined #openstack-ironic | 09:03 | |
| *** romcheg has joined #openstack-ironic | 09:19 | |
| *** ajc_ has quit IRC | 09:20 | |
| openstackgerrit | lawrancejing proposed a change to openstack/python-ironicclient: Add CONTRIBUTING.rst https://review.openstack.org/103784 | 09:21 |
| *** lazy_prince has quit IRC | 09:30 | |
| *** lazy_prince has joined #openstack-ironic | 09:36 | |
| *** coolsvap has quit IRC | 09:37 | |
| *** coolsvap has joined #openstack-ironic | 09:41 | |
| *** amitpp has quit IRC | 09:41 | |
| *** foexle has joined #openstack-ironic | 09:41 | |
| *** Haomeng has joined #openstack-ironic | 09:41 | |
| *** Haomeng|2 has quit IRC | 09:42 | |
| *** Nisha has quit IRC | 09:42 | |
| *** jcoufal has quit IRC | 09:48 | |
| openstackgerrit | A change was merged to openstack/ironic: Updated from global requirements https://review.openstack.org/105167 | 09:51 |
| *** romcheg1 has joined #openstack-ironic | 10:02 | |
| *** romcheg has quit IRC | 10:03 | |
| *** romcheg has joined #openstack-ironic | 10:05 | |
| *** romcheg1 has quit IRC | 10:07 | |
| openstackgerrit | Ramakrishnan G proposed a change to openstack/ironic: Add support for creating vfat disk images https://review.openstack.org/105413 | 10:22 |
| *** vinbs_ has joined #openstack-ironic | 10:45 | |
| *** vinbs has quit IRC | 10:47 | |
| *** vinbs_ is now known as vinbs | 10:48 | |
| *** jcoufal has joined #openstack-ironic | 10:56 | |
| *** vinbs has quit IRC | 10:58 | |
| *** ramineni has quit IRC | 11:02 | |
| *** geekyogi has quit IRC | 11:06 | |
| *** geekyogi has joined #openstack-ironic | 11:06 | |
| *** mkerrin has quit IRC | 11:07 | |
| *** subah has quit IRC | 11:10 | |
| *** steveh2 has joined #openstack-ironic | 11:11 | |
| *** steveh1 has quit IRC | 11:12 | |
| *** rameshg87 is now known as rameshg87_afk | 11:20 | |
| *** nosnos has quit IRC | 11:23 | |
| *** amitpp has joined #openstack-ironic | 11:25 | |
| *** coolsvap is now known as coolsvap|afk | 11:28 | |
| *** lucasagomes is now known as lucas-hungry | 11:41 | |
| *** bvivek has quit IRC | 11:44 | |
| *** pelix has quit IRC | 12:00 | |
| *** jdob has joined #openstack-ironic | 12:01 | |
| *** steveh2 has quit IRC | 12:14 | |
| *** steveh1 has joined #openstack-ironic | 12:16 | |
| *** foexle has quit IRC | 12:22 | |
| *** lucas-hungry is now known as lucasagomes | 12:41 | |
| *** jgrimm has quit IRC | 12:42 | |
| *** k4n0 has quit IRC | 12:44 | |
| *** foexle has joined #openstack-ironic | 12:46 | |
| *** ramineni has joined #openstack-ironic | 12:48 | |
| *** pelix has joined #openstack-ironic | 12:49 | |
| openstackgerrit | Jim Rollenhagen proposed a change to openstack/ironic-specs: Add deploy driver for ironic-python-agent https://review.openstack.org/98506 | 12:53 |
| *** Poornima has joined #openstack-ironic | 12:55 | |
| *** sseago has quit IRC | 12:59 | |
| *** sseago_ has joined #openstack-ironic | 12:59 | |
| *** killer_prince has joined #openstack-ironic | 13:03 | |
| *** jistr has quit IRC | 13:04 | |
| *** jistr has joined #openstack-ironic | 13:06 | |
| *** jbjohnso has joined #openstack-ironic | 13:06 | |
| *** rloo has joined #openstack-ironic | 13:12 | |
| *** amitpp has quit IRC | 13:15 | |
| *** steveh1 has quit IRC | 13:18 | |
| openstackgerrit | Jim Rollenhagen proposed a change to openstack/ironic-specs: Expose configdrive to instances https://review.openstack.org/99235 | 13:19 |
| openstackgerrit | Andrey Kurilin proposed a change to openstack/python-ironicclient: Remove aliases `arg` and `env` from utils https://review.openstack.org/104553 | 13:19 |
| *** matty_dubs|gone is now known as matty_dubs | 13:19 | |
| *** sseago_ has quit IRC | 13:19 | |
| *** sseago__ has joined #openstack-ironic | 13:19 | |
| *** steveh1 has joined #openstack-ironic | 13:20 | |
| jroll | mornin y'all | 13:20 |
| *** lazy_prince has quit IRC | 13:21 | |
| *** killer_prince is now known as lazy_prince | 13:21 | |
| jroll | Shrews, adam_g, when y'all are around can we chat about deva's last comment here? https://review.openstack.org/#/c/98930/4/specs/juno/use-configdrive-with-ironic.rst | 13:21 |
| *** amitpp has joined #openstack-ironic | 13:23 | |
| *** amitpp has quit IRC | 13:25 | |
| NobodyCam | good morning Ironic | 13:28 |
| jroll | heya NobodyCam | 13:28 |
| NobodyCam | morning jroll :) | 13:29 |
| matty_dubs | Hola amigos | 13:29 |
| NobodyCam | hola matty_dubs | 13:29 |
| jroll | hey matty | 13:29 |
| romcheg | ¡Hola Ironic! | 13:29 |
| NobodyCam | morning romcheg | 13:29 |
| jroll | hablamos español ahora? | 13:30 |
| romcheg | jroll: ¿Por qué no? | 13:30 |
| jroll | hablo español un poquito | 13:31 |
| jroll | ¿como se dice ironic? nova? etc etc :P | 13:32 |
| matty_dubs | irónico apparently | 13:32 |
| lucasagomes | morning jroll NobodyCam matty_dubs romcheg | 13:32 |
| romcheg | (: ja-ja : | 13:32 |
| jroll | lol | 13:33 |
| jroll | morning lucasagomes | 13:33 |
| lucasagomes | matty_dubs, irónico is right sir! | 13:33 |
| NobodyCam | orning lucasagomes | 13:34 |
| romcheg | Irónico conductor para Nueva | 13:34 |
| lucasagomes | haha | 13:35 |
| lucasagomes | condutor | 13:35 |
| lucasagomes | oh it's conductor! holy * | 13:35 |
| lucasagomes | in ptgz we don't have that 'c' there | 13:36 |
| matty_dubs | ceilómetro | 13:37 |
| matty_dubs | There's got to be some word that's totally different! | 13:37 |
| matty_dubs | There we go! cinder == ceniza | 13:37 |
| lucasagomes | glance :) | 13:37 |
| lucasagomes | mirada? | 13:38 |
| *** jgrimm has joined #openstack-ironic | 13:38 | |
| romcheg | Y integración de irónico con ceniza, ¿como lo esta? | 13:38 |
| romcheg | ** always forget about accents | 13:38 |
| lucasagomes | todo va bien | 13:38 |
| matty_dubs | Heh, don't forget the accents! That's how I wished someone a "feliz ano nueve" | 13:39 |
| matty_dubs | Which apparently means something _very_ different from año | 13:39 |
| romcheg | ja ja :) | 13:39 |
| lucasagomes | LOL | 13:40 |
| lucasagomes | yeah, it's diff indeed :P | 13:40 |
| romcheg | It's like г and ґ in Ukrainian. It sounds almost the same but if confused "Hey guys!" turns to "Gay guys!" | 13:42 |
| jroll | devananda: about your tempest comment here... https://review.openstack.org/#/c/98930/ | 13:43 |
| jroll | devananda: tempest has almost no instance of 'config.*drive'... only in the schemas. | 13:44 |
| jroll | devananda: so, it seems its not tested in tempest at all | 13:44 |
| jroll | people just pray | 13:44 |
| jroll | or something | 13:44 |
| devananda | jroll: well - it's not exposed to the API | 13:45 |
| NobodyCam | good morning devananda | 13:45 |
| romcheg | Morning devananda! | 13:45 |
| devananda | jroll: tempest is black box testing. that's an internal switch | 13:45 |
| devananda | morning! | 13:45 |
| jroll | devananda: right, but they don't force any test to use configdrive or anything, best I can tell | 13:45 |
| jroll | devananda: where would I look, then, for nova's current configdrive testing? | 13:45 |
| devananda | jroll: check in -qa or -nova? | 13:46 |
| jroll | right, will do | 13:46 |
| rloo | hello ironickers :) | 13:48 |
| NobodyCam | morning rloo | 13:48 |
| rloo | lucasagomes: let me know when you have a few minutes to discuss https://review.openstack.org/#/c/102914/ | 13:48 |
| jroll | heya rloo | 13:48 |
| lucasagomes | rloo, ack I;m in a hangout right now, I ping u in some minutes when it's finished | 13:48 |
| jroll | rloo: I updated https://review.openstack.org/#/c/98506/ for you :) | 13:48 |
| rloo | no worries lucasagomes. in the meantime, guess i'll look at jroll's stuff :-) | 13:49 |
| jroll | \o/ | 13:49 |
| lucasagomes | :D | 13:49 |
| *** foexle has quit IRC | 13:50 | |
| *** zul has joined #openstack-ironic | 14:08 | |
| *** romcheg1 has joined #openstack-ironic | 14:09 | |
| *** romcheg2 has joined #openstack-ironic | 14:09 | |
| *** romcheg1 has quit IRC | 14:09 | |
| *** romcheg has quit IRC | 14:10 | |
| *** dkehnx1 has joined #openstack-ironic | 14:10 | |
| *** steveh1 has quit IRC | 14:11 | |
| NobodyCam | brb | 14:12 |
| *** pcrews has joined #openstack-ironic | 14:13 | |
| *** steveh1 has joined #openstack-ironic | 14:14 | |
| rloo | jroll: thx for updating. I had a few more comments, but i think that'll be it for me. | 14:16 |
| jroll | rloo: cool, will check it out | 14:17 |
| lucasagomes | rloo, done | 14:17 |
| rloo | lucasagomes: my turn :-) | 14:17 |
| lucasagomes | :) | 14:17 |
| rloo | so. I have no idea if humans and/or computers will be using this new API. | 14:18 |
| rloo | lucasagomes: deva made a comment that he thinks computers might; I was only thinking humans. but once you have something, you never know what will consume it. | 14:18 |
| lucasagomes | right, I just say that because usually I would think that as it's in the api other services to be the ones consuming that | 14:19 |
| rloo | lucasagomes: the reason for making the one-of-a-set-of-properties optional, is that this is more in line with the results from validate(). | 14:19 |
| *** steveh1 has quit IRC | 14:20 | |
| *** steveh1 has joined #openstack-ironic | 14:21 | |
| rloo | lucasagomes: the initial bug report was descibing how people didn't know what properties to specify, but yeah. i can add something to the spec about consumption by humans and/or computers. | 14:21 |
| lucasagomes | yeah that would be good to clarify, because I was thinking about the nova driver consuming it at one point | 14:22 |
| lucasagomes | rloo, do you think that, it's possible and worth investing time to make those options to be able to be consumed programatically? | 14:22 |
| rloo | lucasagomes: i saw your comment there wrt nova driver impact. so the api changes itself don't impact the driver. it seems to me that if the driver wants to consume it, that is another spec/bug? | 14:23 |
| lucasagomes | rloo, yeah, it would be separate | 14:23 |
| lucasagomes | but, on that format it's hard to be consumed by it | 14:23 |
| lucasagomes | cause programatically it's difficult to know whether options are required or not | 14:23 |
| lucasagomes | CLIs have a nice way to express those conditions | 14:24 |
| rloo | lucasagomes: even programmatically (as with your example with driver_opts.add_property(), you still can't indicate clearly 1-of-a-set-of-is-required. | 14:24 |
| lucasagomes | rloo, yeah not on that example, that would need to more advanced parser | 14:25 |
| lucasagomes | rloo, like CLIs would do somehting like: ./program do (--opt1 | --opt2 | --opt3) | 14:26 |
| rloo | lucasagomes: sorry, by 'on that format it's hard to be consumed by it', i'm not sure what you mean. the eg nova ironic driver could issue the request and get a dictionary back iwht the prop name, required/optiona. | 14:26 |
| lucasagomes | do is the command the one of the opts needs to be present ( '( )' means required) | 14:26 |
| *** stendulker has joined #openstack-ironic | 14:26 | |
| lucasagomes | rloo, yeah it gets the dictornary | 14:26 |
| lucasagomes | but for the ssh_options case | 14:26 |
| lucasagomes | they are all marked as optional | 14:27 |
| lucasagomes | but one of them are required | 14:27 |
| lucasagomes | how would the driver know that? | 14:27 |
| lucasagomes | for humans we can read the help string | 14:27 |
| lucasagomes | which says that one of those are required, but programatically that's not easy to see that | 14:27 |
| rloo | lucasagomes: sorry, i don't understand your CLI example and how that addresses the problem. | 14:28 |
| lucasagomes | rloo, ah, maybe forget the CLI, what I'm trying to say is that CLIs parser have to deal with those complex parameters conditional parsing | 14:29 |
| rloo | lucasagomes: if the '(...)' indicates one needs to be present, what if we have a case where 2-of needs to be present? | 14:29 |
| lucasagomes | and they kinda have a format language to deal with it | 14:29 |
| lucasagomes | rloo, the 2-of makes my head explode heh I don't know how to express that really | 14:30 |
| rloo | lucasagomes: my concern is that if we jump over a hoop to handle the 1-of case, and the 2-of or some other case comes up, then what? | 14:30 |
| stendulker | lucasgomes: Hi | 14:30 |
| lucasagomes | rloo, yeah that's the problem... so it comes down to | 14:31 |
| lucasagomes | if it's human consumable | 14:31 |
| lucasagomes | should it be in the API? | 14:31 |
| lucasagomes | stendulker, hey | 14:31 |
| rloo | lucasagomes: you mean, if it is *only* human consumable | 14:31 |
| lucasagomes | rloo, yes, if it's only human consumable | 14:31 |
| *** max_lobur has joined #openstack-ironic | 14:31 | |
| rloo | lucasagomes: cuz almost all API calls are/can be human consumable :-) | 14:31 |
| stendulker | Did you get chance to review firmware settings related design spec https://review.openstack.org/#/c/101122 | 14:32 |
| rloo | lucasagomes: if it was only human consumable then I wouldn't have even had to worry about performance ;) | 14:32 |
| NobodyCam | lucasagomes: is there a spec for then "use the new ManagementInterface" patches? | 14:33 |
| lucasagomes | stendulker, sorry not yet :( I was travelling last week, and this week I'm still catching up with the things we are doing. I reviewed some specs today but didn't look at the firmware one yet :( I will do soon | 14:33 |
| lucasagomes | NobodyCam, yes I'm working on that | 14:33 |
| NobodyCam | :) | 14:33 |
| lucasagomes | rloo, true, the cache thing | 14:33 |
| stendulker | lucasgomes: ok. Thank you. | 14:33 |
| lucasagomes | rloo, sorry, I don't want to block that spec... I'm just a bit concerned about it | 14:34 |
| devananda | not to be overly literal first thing in the morning, but API calls are, by definition, meant to be consumed by applications | 14:34 |
| lucasagomes | yeah that's what I'm thinking about ^ | 14:34 |
| stendulker | JayF: Hi | 14:34 |
| rloo | lucasagomes: I think it is good to voice your concern. | 14:35 |
| lucasagomes | in the perfect scenario, I think that all the conditions we have for the parameters should be somehow able to be consumed programatically | 14:35 |
| lucasagomes | rloo, thanks | 14:35 |
| rloo | lucasagomes: I agree about the perfect scenario. | 14:35 |
| devananda | lucasagomes: IMO that should be a requirement, not a goal | 14:35 |
| devananda | s/goal/perfect scenario/ | 14:36 |
| JayF | stendulker: hello | 14:36 |
| devananda | we're not there today, clearly | 14:36 |
| lucasagomes | yeah | 14:36 |
| rloo | lucasagomes: the problem I see is that we don't know what all the conditions might be 'at the end of the day', but we do know what the conditions are now. | 14:36 |
| devananda | but it makes using our APIs more difficult | 14:36 |
| JayF | stendulker: I'll add that spec to the queue of things to review today | 14:36 |
| lucasagomes | rloo, right, so do you think we should put some effort and try to create a format to express the conditions we have for the parameters we have _today_? | 14:36 |
| lucasagomes | that wouldn't include the 2-of or Y-of-X scenario | 14:37 |
| stendulker | JayF: Wanted to discuss your comment on firmware settings related comment https://review.openstack.org/#/c/101122/ | 14:37 |
| lucasagomes | but have to include 1-of | 14:37 |
| stendulker | JayF: ok | 14:37 |
| rloo | lucasagomes: so deva mentioned that there's some other spec that has some x-of-a-set but i don't know if it is 1-of-a set or something else. | 14:37 |
| lucasagomes | devananda, true, yeah | 14:37 |
| JayF | stendulker: I still have pretty significant concerns with exposing individual BIOS settings via public API | 14:38 |
| JayF | stendulker: Primarily because they're going to be incredibly difficult, if not impossible, to keep abstracted. | 14:38 |
| devananda | JayF: that ^ sounds like a non-starter to me, but I haven't read the spec | 14:38 |
| devananda | for just that reason | 14:38 |
| JayF | stendulker: which is why I suggested some specific settings (like VT) or sets of settings (like 'virt-friendly' or 'power management') which /can/ be abstracted by drivers | 14:39 |
| JayF | stendulker: without a change to allow it to truly be abstracted, I'm going to remain -1 on the idea :x | 14:39 |
| lucasagomes | rloo, which spec is that? | 14:39 |
| devananda | JayF: ++ | 14:39 |
| rloo | lucasagomes: have you looked at the IPMI double-bridging spec? I haven't but look at deva's comments here, line 121: https://review.openstack.org/#/c/102914/1/specs/juno/get-required-driver-info.rst | 14:39 |
| lucasagomes | rloo, ack | 14:39 |
| devananda | rloo: it appears to be n-of-a-set | 14:40 |
| devananda | rloo: set of 5 options. under certain cases, these 2 are needed, at other times, others are needed. it's not clear to me how the driver even determines this | 14:41 |
| lucasagomes | yeah that's quite complex | 14:42 |
| devananda | rloo: I dont want to block the get-required-driver-info spec on the ipmi-double-bridging -- merely pointing out the extensibility (or lack thereof) of the API you proposed so we can all consider it | 14:42 |
| JayF | stendulker: I reflected those comments back on the review so they'll be there for others to see: https://review.openstack.org/#/c/101122/ | 14:42 |
| rloo | devananda, lucasagomes: now we get into smart drivers ;) but if the driver can determine it, it can programmatically return the properties with the right required/optional. as long as that is static. if it changes based on the weather or something, then ugh. | 14:42 |
| devananda | rloo: right. afaik, for ipmi double bridging, it'll be static per device, but could vary between devices (eg, by different mfg, or different hardware rev, or something) | 14:43 |
| *** rakesh_hs has quit IRC | 14:43 | |
| devananda | honestly, i had not considered such a complicated thing before | 14:44 |
| devananda | and feel like either we accept the limitation and document it, or we push back on the ipmi-double-bridging proposal to make them simplify that API | 14:44 |
| stendulker | JayF: Regarding you comment, can we discuss a bit. I still did not understand them fully. | 14:45 |
| JayF | stendulker: I'm actually not going to be around much today, but if you can ask for clarification in gerrit I'll look at it :). I also like using gerrit for this so the conversations persist for everyone to see (not just everyone reading IRC at 745am) | 14:46 |
| rloo | devananda: so there wouldn't be a problem if there was no api. and we want an api so that it can be consumed by applications. or do we? | 14:46 |
| rloo | devananda: if the problem is getting the info to humans, then for now, maybe we just document. | 14:47 |
| devananda | rloo: we've created an application (ironic) which requires certain inputs (driver info) which vary by some condition that we expose (node.driver) but we never indicate what the structure of that info is | 14:48 |
| stendulker | JayF: With IRC, it helps to get the context quicker. I can post it on the gerrit as well, if that is preferred. | 14:49 |
| devananda | rloo: i meant document the wacky conditions around "n of a set" stuff | 14:50 |
| devananda | rloo: and expose the set as optional | 14:50 |
| rloo | devananda: but as lucasagomes sez, exposing the entire set as optional isn't accurate enough. | 14:51 |
| devananda | i agree that it's not perfect | 14:52 |
| devananda | but is it good enough? | 14:52 |
| rloo | devananda: no, it isn't good enough according to lucasagomes. | 14:53 |
| lucasagomes | that's hard to answer | 14:53 |
| rloo | devananda: and it won't be good enough for some one out there, i'm sure. | 14:53 |
| lucasagomes | I'm thinking as if I had to parse that info somehow for my app consume it | 14:53 |
| rloo | devananda, lucasagomes: devananda, lucasagomes: let me think about it. I think we can do 'exactly n-of-a-set', but need to think about other weird things like 'n-or-more-of-a-set' and other weird stuff. | 14:53 |
| rloo | devananda, lucasagomes: it'll make the output more complicated but applications don't care about that. | 14:54 |
| lucasagomes | rloo, right I will also try to google it a bit see if I can come up with some ideas for how to represent it | 14:55 |
| devananda | rloo: (username and (password or key)) is a good example, and close to the SSH driver. ((if foo then also bar) and (if baz then also kazoo)) is not, however this is closer to ipmi double bridge spec. | 14:56 |
| stendulker | JayF: I have posted my query in gerrit. | 14:56 |
| rloo | devananda: thx for another case :-) | 14:56 |
| devananda | lucasagomes: so if I were writing an app to consume this, I'd have local data, and some mapping from that to the application's data structure | 14:57 |
| devananda | lucasagomes: and some validation that the transformation I expect to work, actually works | 14:57 |
| stendulker | NobodyCam: Hi | 14:58 |
| devananda | lucasagomes: if the app changes its data structure, that validation would (presumably) fail until I change the mapping | 14:58 |
| lucasagomes | devananda, right, my main objection is about that "required" field being bogus | 15:00 |
| NobodyCam | stendulker: good morning | 15:00 |
| lucasagomes | if we require a mapping etc, maybe we should not even have that required field there | 15:00 |
| devananda | lucasagomes: did you see my suggestion to make it optional? | 15:00 |
| lucasagomes | devananda, yea just read it | 15:01 |
| devananda | lucasagomes: or you mean just remove the whole field | 15:01 |
| devananda | and simply expose the list of parameters this driver understands | 15:01 |
| stendulker | NobodyCam: Good morning. Its evening (8:30PM) on my side :) | 15:01 |
| lucasagomes | devananda, I mean remove the whole required field | 15:01 |
| devananda | lucasagomes: works for me | 15:01 |
| lucasagomes | so the help string can say whether it's required or not | 15:02 |
| NobodyCam | stendulker: :) | 15:02 |
| *** max_lobur has quit IRC | 15:02 | |
| lucasagomes | or if one-of-X is required | 15:02 |
| lucasagomes | etc | 15:02 |
| lucasagomes | rloo, ^ | 15:02 |
| rloo | lucasagomes: works for me too. what is an application going to do with it though? | 15:02 |
| lucasagomes | rloo, having an internal map | 15:02 |
| devananda | lucasagomes: yep. let a human understand that, but let an app retrieve the list to generate its ETL | 15:02 |
| stendulker | NobodyCam: Did you get chance to look at comment resolution for firmware settings spec review https://review.openstack.org/#/c/101122/ | 15:02 |
| rloo | lucasagomes, devananda: yay, much easier for me to do then! | 15:03 |
| NobodyCam | stendulker: not yet.. Will take a look now | 15:03 |
| lucasagomes | devananda, yeah I think that's good | 15:03 |
| lucasagomes | rloo, :) right | 15:03 |
| stendulker | NobodyCam: Ok. Thanks. | 15:03 |
| rloo | thx lucasagomes for simplifying it! | 15:03 |
| *** JayF has quit IRC | 15:03 | |
| *** JayF has joined #openstack-ironic | 15:04 | |
| *** JayF has quit IRC | 15:04 | |
| *** JayF has joined #openstack-ironic | 15:04 | |
| lucasagomes | rloo, devananda thanks | 15:04 |
| devananda | cheers :) | 15:05 |
| lucasagomes | rloo, last thing, about generating that json automatically? | 15:08 |
| lucasagomes | rloo, so we keep those options and help strings updated | 15:09 |
| lucasagomes | rloo, do you think it's workable as part of that spec? or an improvement we can do later? | 15:09 |
| *** Poornima has quit IRC | 15:10 | |
| *** coolsvap|afk is now known as coolsvap | 15:10 | |
| *** romcheg2 has quit IRC | 15:10 | |
| devananda | also, i really need to spend a few hours doing paperwork and booking travel today... going to be afk a lot | 15:11 |
| *** romcheg has joined #openstack-ironic | 15:12 | |
| *** jistr has quit IRC | 15:12 | |
| *** jistr has joined #openstack-ironic | 15:13 | |
| NobodyCam | stendulker: I see your reply, My concern is, how a failed update will effect a node. while a node may be usable after a out of band failure, should it be. | 15:13 |
| NobodyCam | So I would like other to weigh in on that comment | 15:14 |
| *** steveh1 has quit IRC | 15:14 | |
| *** steveh1 has joined #openstack-ironic | 15:17 | |
| *** max_lobur has joined #openstack-ironic | 15:17 | |
| rloo | lucasagomes: definitely an improvement we can do later. let me look/think about it as part of this spec though. | 15:19 |
| lucasagomes | ack | 15:19 |
| lucasagomes | yeah was thinking bout just create a small wrapper around the optparse lib to do that | 15:20 |
| lucasagomes | but yeah can be done later | 15:20 |
| rloo | lucasagomes: so I think the opts stuff might be overkill and i'm not sure it will work. My current impl doesn't explicitly convert anything to json. All developer has to do now is add the property name/help string, which can just be done in a dictionary. | 15:28 |
| devananda | lucasagomes: client paginatin patch - i keep forgetting about that. how's it going? | 15:28 |
| lucasagomes | devananda, oh that's done | 15:28 |
| lucasagomes | devananda, https://review.openstack.org/#/c/91585/ | 15:28 |
| devananda | ya, reviewing now | 15:29 |
| lucasagomes | devananda, limiting and marker added for all the listing functions etc | 15:29 |
| devananda | awesome | 15:29 |
| lucasagomes | rloo, right yeah that's true :) | 15:29 |
| devananda | that reminds me that i should probably release another client rev soon :) | 15:29 |
| lucasagomes | devananda, +1 | 15:29 |
| rloo | lucasagomes: sorry, maybe the opts stuff will work, just might be overkill. | 15:29 |
| lucasagomes | yeah indeed it might be overkill | 15:29 |
| lucasagomes | now that we don't need the required stuff | 15:29 |
| * NobodyCam thinks devananda will never finish his paper work | 15:29 | |
| lucasagomes | it's very achieveable with the dictornary | 15:30 |
| rloo | lucasagomes: yeah. probably best to stick to something simple. it is internal so can be changed later if need be. | 15:30 |
| lucasagomes | rloo, +10000 | 15:30 |
| rloo | lucasagomes: :D | 15:30 |
| devananda | lucasagomes: what's the relationship between _list_pagination's "limit" param and the REST API ?limit=X ? | 15:31 |
| jroll | rloo: thinking about this "deploy agent" again, I think that text is fine. IPA is a deploy agent, and IMO a deploy agent should also be able to assist with tear downs, or "undeploy" if you will | 15:31 |
| devananda | lucasagomes: looks like the same value is being used for both, but it's not immediately clear why | 15:31 |
| lucasagomes | devananda, yeah, well if not specified the api will get the max number from a config option | 15:32 |
| rloo | jroll: ok, if IPA is a deploy agent. I thought maybe the agent would be doing other tasks besides deploy, that's all. | 15:32 |
| lucasagomes | devananda, I can remove that from the CLI, it's because I made the library to returning everything by default | 15:32 |
| lucasagomes | and added a limit on the CLI | 15:32 |
| * lucasagomes looks at the patch, don't remember the details | 15:33 | |
| rloo | jroll: 'deploy' == deploy/teardown etc, whatever the DeployInterface does ;) | 15:33 |
| JayF | rloo: it all hinges on what you consider "deploy". I don't think it's crazy to think that 'deployment' can include cleaning up after (or before) the deployment in order to give an Ironic user a consistent ux | 15:33 |
| jroll | rloo: it will do other things besides DeployDriver.deploy(), yes, but I tend to think of a 'deploy agent', as more than that | 15:33 |
| devananda | lucasagomes: ah. so if CLI doesn't include --limit, it'll loop (based on the API service's config'd limit) until it has fetched everything ? | 15:33 |
| jroll | rloo: right, decommissioning should be part of DeployInterface :) | 15:33 |
| jroll | etc | 15:33 |
| lucasagomes | devananda, the client library will | 15:34 |
| devananda | lucasagomes: cool | 15:34 |
| lucasagomes | devananda, for the CLI I added 100 as a default limit | 15:34 |
| lucasagomes | devananda, https://review.openstack.org/#/c/91585/3/ironicclient/v1/chassis_shell.py | 15:34 |
| lucasagomes | default=100, | 15:34 |
| devananda | yea | 15:34 |
| rloo | jroll: fine then. | 15:34 |
| jroll | ok :) | 15:35 |
| rloo | jroll: so ah, for the first pass, is it *just* deploying? | 15:35 |
| lucasagomes | devananda, a easy way to test it, is set a very low limit on the Ironic API (like 1) and use the CLI to fetch the values | 15:36 |
| stendulker | NobodyCam: Than you. I will wait for others comment on the same. | 15:36 |
| lucasagomes | 0 means no limit, so it should fetch everything | 15:37 |
| openstackgerrit | Jim Rollenhagen proposed a change to openstack/ironic-specs: Add deploy driver for ironic-python-agent https://review.openstack.org/98506 | 15:37 |
| jroll | rloo: yes, initial implementation is to (almost) match pxe driver | 15:37 |
| jroll | e.g. patch 101020 | 15:37 |
| jroll | and then from there we will add features | 15:37 |
| rloo | jroll: ok thx. | 15:37 |
| jroll | no problem, thanks for your comments :) | 15:37 |
| jroll | put one more patchset up, specs cores :) https://review.openstack.org/98506 | 15:38 |
| NobodyCam | brb | 15:40 |
| devananda | lucasagomes: bug -- it's making two requests even if the first response is complete | 15:40 |
| lucasagomes | devananda, ew :( lemme take a look | 15:41 |
| lucasagomes | devananda, the first request returned a 'next' link? | 15:41 |
| devananda | eg, 127.0.0.1 - - [08/Jul/2014 08:40:10] "GET /v1/nodes/?limit=1 HTTP/1.1" 200 485 | 15:42 |
| devananda | 127.0.0.1 - - [08/Jul/2014 08:40:10] "GET /v1/nodes?sort_key=id&sort_dir=asc&limit=1&marker=8f68da58-3853-4a8e-a7d9-3abe96cfa607 HTTP/1.1" 200 485 | 15:42 |
| devananda | even though limit=1 :) | 15:42 |
| *** igordcard has quit IRC | 15:42 | |
| *** Nisha has joined #openstack-ironic | 15:42 | |
| lucasagomes | devananda, oh, right will take a look into it | 15:43 |
| lucasagomes | devananda, I probably have to check if we already reached the limit before getting the 'next' link and requesting again | 15:43 |
| devananda | lucasagomes: i see the problem. base.py L75 | 15:43 |
| devananda | lucasagomes: limit_reached=True is set only if limit > 1 | 15:44 |
| devananda | or rather, it can not be set on the first pass through the loop | 15:44 |
| lucasagomes | devananda, yeah I think that while look break should check the object_count | 15:46 |
| *** amitpp has joined #openstack-ironic | 15:46 | |
| lucasagomes | and not the limit_reached flag (maybe that flag is not even needed) | 15:46 |
| lucasagomes | lemme re work it | 15:46 |
| lucasagomes | I had to redo that loop cause I was using recursion before, and python has a limit (of 1000 by default) for recursion | 15:47 |
| devananda | right. this is much better | 15:47 |
| devananda | lucasagomes: suggestion added in comments | 15:50 |
| lucasagomes | devananda, ta much I will address that and resubmit the patch soon | 15:50 |
| devananda | ty | 15:50 |
| Nisha | jroll: NobodyCam Hi | 15:54 |
| rameshg87_afk | rloo: request you to change vote for https://review.openstack.org/#/c/89500/ | 15:56 |
| rameshg87_afk | rloo: its spec got submitted yesterday, https://review.openstack.org/#/c/97455/ | 15:57 |
| rameshg87_afk | rloo: request you to review the code as well :-) | 15:57 |
| *** rameshg87_afk is now known as rameshg87 | 15:57 | |
| rloo | rameshg87: I changed the vote. One of these days I may review the code too, but not now ;) | 15:57 |
| devananda | fwiw, looks like there are 3 other client patches that could use some reviews | 15:58 |
| *** vinbs has joined #openstack-ironic | 15:58 | |
| *** vinbs_ has joined #openstack-ironic | 16:01 | |
| rameshg87 | thanks rloo | 16:02 |
| *** vinbs has quit IRC | 16:03 | |
| *** vinbs_ is now known as vinbs | 16:03 | |
| *** dwalleck has joined #openstack-ironic | 16:04 | |
| *** chuckC has quit IRC | 16:04 | |
| *** eghobo has joined #openstack-ironic | 16:04 | |
| *** eghobo has quit IRC | 16:05 | |
| *** eghobo has joined #openstack-ironic | 16:06 | |
| *** igordcard has joined #openstack-ironic | 16:06 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/python-ironicclient: Add pagination support to {node, port, chassis}-list https://review.openstack.org/91585 | 16:07 |
| lucasagomes | devananda, ^ had to add a dependency on #102312 (ur patch updating the docstrings) cause I just approved that and it would conflict | 16:08 |
| devananda | ack | 16:08 |
| *** chuckC has joined #openstack-ironic | 16:09 | |
| *** martyntaylor has left #openstack-ironic | 16:10 | |
| *** vinbs has quit IRC | 16:11 | |
| *** steveh1 has quit IRC | 16:11 | |
| *** steveh1 has joined #openstack-ironic | 16:14 | |
| *** martyntaylor1 has joined #openstack-ironic | 16:15 | |
| *** lazy_prince is now known as killer_prince | 16:19 | |
| *** martyntaylor1 has quit IRC | 16:20 | |
| *** matty_dubs is now known as matty_dubs|lunch | 16:21 | |
| *** chuckC has quit IRC | 16:23 | |
| *** chuckC has joined #openstack-ironic | 16:23 | |
| *** martyntaylor has joined #openstack-ironic | 16:27 | |
| *** aweeks_ is now known as aweeks | 16:29 | |
| openstackgerrit | lawrancejing proposed a change to openstack/python-ironicclient: Use suitable assert https://review.openstack.org/105512 | 16:30 |
| rameshg87 | devananda: request you to take a look at virtual media deploy spec: https://review.openstack.org/#/c/97744/ | 16:33 |
| *** dwalleck has quit IRC | 16:33 | |
| *** jistr has quit IRC | 16:40 | |
| *** ramineni has quit IRC | 16:41 | |
| *** aswadr has quit IRC | 16:42 | |
| rameshg87 | lucasagomes: request you to have a look at ilo power driver code you had reviewed once early https://review.openstack.org/#/c/89500/ | 16:43 |
| *** jcoufal has quit IRC | 16:43 | |
| *** hemna has joined #openstack-ironic | 16:44 | |
| openstackgerrit | Ruby Loo proposed a change to openstack/ironic-specs: API to get driver_info properties https://review.openstack.org/102914 | 16:49 |
| openstackgerrit | A change was merged to openstack/python-ironicclient: Make a few minor updates to node shell help strings https://review.openstack.org/102312 | 16:55 |
| *** dwalleck has joined #openstack-ironic | 16:55 | |
| *** derekh_ has quit IRC | 16:58 | |
| *** steveh1 has quit IRC | 17:01 | |
| openstackgerrit | A change was merged to openstack/python-ironicclient: Add CONTRIBUTING.rst https://review.openstack.org/103784 | 17:02 |
| *** Nisha has quit IRC | 17:02 | |
| *** matty_dubs|lunch is now known as matty_dubs | 17:02 | |
| *** steveh has joined #openstack-ironic | 17:03 | |
| *** zdiN0bot has joined #openstack-ironic | 17:06 | |
| *** stendulker has quit IRC | 17:07 | |
| *** lucasagomes is now known as lucas-dinner | 17:07 | |
| lucas-dinner | aight will call it a day, have a good night everybody | 17:09 |
| lucas-dinner | rameshg87, I will try to do it soon, I didn't have much time (was traveling last week) | 17:09 |
| lucas-dinner | sorry for the delay to re-review it | 17:09 |
| rloo | good night lucas-dinner | 17:10 |
| *** rameshg87 is now known as rameshg87_afk | 17:11 | |
| *** pelix has quit IRC | 17:12 | |
| NobodyCam | night lucas-dinner | 17:20 |
| *** Nisha has joined #openstack-ironic | 17:20 | |
| openstackgerrit | A change was merged to openstack/ironic: Specify rootfstype=ramfs deploy kernel parameter https://review.openstack.org/104728 | 17:23 |
| *** harlowja_away is now known as harlowja | 17:29 | |
| JoshNang | devananda: for swift temp urls, if we add python-swiftclient to requirements.txt and need to pin it to > the version that adds them, do we have to update the global requirements file to that version too? or can we have slightly more strict requirements for ironic? | 17:38 |
| *** igordcard has quit IRC | 17:39 | |
| Nisha | jroll: there? | 17:41 |
| jroll | Nisha: hi | 17:41 |
| Nisha | jroll: hi | 17:42 |
| Nisha | jroll: i was just going through yesterday's weekly meeting logs | 17:42 |
| *** steveh has quit IRC | 17:42 | |
| Nisha | jroll: NobodyCam has discussed about https://review.openstack.org/#/c/100951/ in the meeting | 17:43 |
| *** steveh has joined #openstack-ironic | 17:43 | |
| *** overlayer has joined #openstack-ironic | 17:43 | |
| Nisha | jroll: I received the comment from NobodyCam saying that inventory field will be added by the other generic hw specand hence my spec has dependency on the other one | 17:44 |
| jroll | Nisha: go on | 17:44 |
| jroll | yes | 17:45 |
| Nisha | jroll: but the field 'inventory' can be added by any spec, correct? | 17:45 |
| jroll | it's not just about that field | 17:45 |
| Nisha | it is the new field for either of the spec | 17:45 |
| jroll | look | 17:46 |
| jroll | the two specs overlap a lot | 17:46 |
| Nisha | No, | 17:46 |
| jroll | the generic hw discovery spec, specifies how to discover hw properties | 17:46 |
| Nisha | the spec proposed by Dmitry do a autodicovery of the node | 17:47 |
| jroll | your spec, specifies that discovery should happen on node-create and node-update | 17:47 |
| jroll | sure | 17:47 |
| Nisha | it doesnt discover hardware properties at node-create and node-update | 17:47 |
| jroll | it also specifies how to discover hardware | 17:47 |
| jroll | right | 17:47 |
| jroll | so your spec is an extra feature on top of dtantsur's spec, IMO | 17:47 |
| Nisha | No, both are different......the properties discovery done at node-create and node-update will never do autodiscovery | 17:48 |
| *** sseago__ has quit IRC | 17:49 | |
| Nisha | Dmitry's spec has more of IPA content which may not apply to all the OOB methods, | 17:49 |
| jroll | so comment on Dmitry's spec with that concern | 17:50 |
| Nisha | I had discussion with Dmitry also when he had proposed the spec. He told our spec's do not overlap... | 17:50 |
| jroll | look | 17:50 |
| jroll | specs cores want you to make that dependency | 17:50 |
| jroll | I am not a spec core | 17:50 |
| jroll | but I believe they are correct | 17:50 |
| jroll | I can deal with that dependency not happening, if you all come to some agreement | 17:50 |
| jroll | because that's not my decision to make | 17:50 |
| Nisha | my point of main discussion is here that just because the new field is common, i need to add a dependency on the other spec..... | 17:51 |
| jroll | there is a lot of common code that will come out of these two | 17:52 |
| jroll | it's not just that field | 17:52 |
| jroll | Nisha: also | 17:52 |
| jroll | It will add inventory as a new field to the node object as proposed by 211 | 17:52 |
| jroll | https://review.openstack.org/#/c/102565/. | 17:52 |
| jroll | perhaps that is the confusion | 17:53 |
| Nisha | Yes, that is where the dependency has come from | 17:53 |
| jroll | you directly reference a spec, but are not dependent on it? | 17:53 |
| Nisha | Yes | 17:53 |
| jroll | perhaps you should reword that | 17:53 |
| Nisha | I referred it because Dmitry gave the comment that the inventory field shall be same as he has proposed in other spec | 17:53 |
| jroll | sure | 17:54 |
| jroll | so, NobodyCam gave you a -1 | 17:54 |
| Nisha | so that the data discovered by two methods are stored at one place and in same manner | 17:54 |
| jroll | I did not | 17:54 |
| jroll | you responded to his comment on gerrit | 17:54 |
| jroll | you should either wait for a comment from him there, or chat with him here | 17:54 |
| jroll | talking to me isn't going to change any of this | 17:54 |
| Nisha | Yes, i saw your comments in weekly meeting so just wanted to know why you also think so | 17:54 |
| jroll | there is potential overlap | 17:55 |
| jroll | possibly a *lot* of overlap | 17:55 |
| jroll | and I don't see a good reason not to depend on the other spec | 17:55 |
| Nisha | I think i sort of convinced him yesterday about this but he wanted to get others opinions, after that the meeting happened | 17:55 |
| jroll | I think you should leverage work done for autodiscovery | 17:56 |
| jroll | or vice versa | 17:56 |
| jroll | as in, I would be fine with dmitry's spec depending on your spec | 17:56 |
| jroll | or your spec depending on dmitry's | 17:56 |
| Nisha | hmmm | 17:56 |
| jroll | but that's just my opinion :) | 17:56 |
| Nisha | NobodyCam: there? | 17:57 |
| Nisha | jroll: thanks, just wanted to understand why core team feels that the spec should have dependency | 17:57 |
| NobodyCam | maybe.. | 17:57 |
| jroll | Nisha: I'm not on the core team, but glad you value my opinion :) | 17:58 |
| Nisha | NobodyCam: hi, saw yesterdays weekly meeting discussion ....jroll tend to agree there that the dependency shall be there. Here he is fine the reverse way too.... | 17:58 |
| Nisha | jroll: :) Thanks | 17:59 |
| jroll | np | 17:59 |
| *** martyntaylor has left #openstack-ironic | 18:00 | |
| Nisha | NobodyCam: do you still see that the spec 100951 shall be dependent on 102565? | 18:00 |
| *** geekyogi has quit IRC | 18:01 | |
| Nisha | The spec 102565 is autodiscovery which is not the scope of 100951 | 18:01 |
| *** sseago__ has joined #openstack-ironic | 18:02 | |
| NobodyCam | Nisha: what I don't what is for two (or more spec's) to implament the same functions in different ways. | 18:02 |
| Nisha | NobodyCam: yes i understand that. so what if the 102565 puts dependency on 100951? | 18:04 |
| jroll | Nisha: you do realize that if you depend on 102565, you have less work to do, right? :) | 18:05 |
| jroll | you would just need to implement the ilo discovery bits, and the node-create/update hooks | 18:05 |
| NobodyCam | sorry in office converstion going on atm | 18:06 |
| Nisha | jroll: :) same applies there too :) | 18:06 |
| Nisha | jroll: ilo discovery bits are simple and are not covered in 100951 | 18:07 |
| jroll | right, I'm aware | 18:07 |
| Nisha | jroll: Lets see what NobodyCam says.... since you can have dependency either way, so let me see what's his opinion now | 18:09 |
| Nisha | jroll: actually i think 102565 will take time to take shape as it is designed solely on IPA implementation as of now which may not be applicable to other OOB mechanisms | 18:11 |
| jroll | it is *not* dependent on IPA | 18:11 |
| jroll | (as far as I know) | 18:11 |
| jroll | and if it is, it shouldn't be. oob should fit there too | 18:11 |
| *** jcoufal has joined #openstack-ironic | 18:12 | |
| Nisha | the spec itself says that as of now it is designed based on IPA implementation | 18:12 |
| jroll | "designed based on" does not mean "requires" | 18:12 |
| jroll | I'm not sure how well it applies to OOB mechanisms | 18:12 |
| jroll | but the plan for non-IPA is to make a separate discovery ramdisk | 18:13 |
| adam_g | jroll, hey, ive got a few mins to chat about https://review.openstack.org/#/c/98930/4/specs/juno/use-configdrive-with-ironic.rst if you do | 18:13 |
| Nisha | Ok | 18:13 |
| jroll | adam_g: talking with sdague in -qa right now... but in general looking for a good test plan for that | 18:16 |
| jroll | adam_g: seems nova doesn't test it explicitly right now | 18:16 |
| NobodyCam | sorry guys am currently distracted. was there a patch that allowed changing of glance id's when rebuilding... I cann't find it .. did it land? | 18:18 |
| jroll | NobodyCam: I seem to remember that... not sure if it landed | 18:19 |
| *** steveh has quit IRC | 18:21 | |
| *** steveh has joined #openstack-ironic | 18:23 | |
| NobodyCam | ahh: merged: https://review.openstack.org/#/c/103227/ | 18:23 |
| *** overlayer has quit IRC | 18:26 | |
| *** shardy is now known as shardy_afk | 18:34 | |
| *** amitpp has quit IRC | 18:34 | |
| Nisha | NobodyCam: | 18:40 |
| *** kevinbenton has quit IRC | 18:40 | |
| NobodyCam | hi | 18:45 |
| Nisha | NobodyCam: hi | 18:45 |
| NobodyCam | sorry was working with someone here in the office | 18:46 |
| Nisha | Yes, understand that ....the office meeting over? | 18:46 |
| Nisha | or still going on? | 18:46 |
| NobodyCam | Nisha: I am ok if 102565 depends on 100951 or vis versa, however Dmitry is out on pto for another week I believe | 18:48 |
| NobodyCam | I just dont want two specs implamenting the functions | 18:49 |
| Nisha | NobodyCam: Thanks, Yes he is on vacation | 18:49 |
| NobodyCam | spereatily | 18:49 |
| Nisha | I understand that. | 18:49 |
| NobodyCam | :) | 18:49 |
| Nisha | Given above would you like to change your vote? | 18:50 |
| Nisha | ;) | 18:50 |
| *** steveh has quit IRC | 18:50 | |
| *** steveh has joined #openstack-ironic | 18:50 | |
| NobodyCam | I can but it would still be -1 as 100951 would need to not refference 102565 | 18:51 |
| NobodyCam | so we can change 100951 to describe creating the inventory field and then have 102565 depend on it | 18:52 |
| NobodyCam | does that make sense | 18:53 |
| NobodyCam | ? | 18:53 |
| Nisha | Ok, fine. i will just do the proposed changes and post it back :) | 18:53 |
| *** kevinbenton has joined #openstack-ironic | 18:53 | |
| Nisha | NobodyCam: Yes, thanks | 18:53 |
| Nisha | for considering the request | 18:53 |
| Nisha | :) | 18:53 |
| NobodyCam | I'm not thrying to make this hard. | 18:53 |
| NobodyCam | I just want to make sure we don't have conflicting specs | 18:54 |
| *** lucas-dinner has quit IRC | 18:54 | |
| NobodyCam | as your spec refferenced the other I made the link between the two | 18:54 |
| NobodyCam | And thank you Nisha for all of your work | 18:55 |
| *** Penick has joined #openstack-ironic | 18:56 | |
| Nisha | NobodyCam: yes thats true we shouldnt have conflicting specs | 18:57 |
| NobodyCam | another option would be to create a add inventory spec and have both 100951 and 102565 depend on that :) | 18:58 |
| NobodyCam | Nisha: I find single item changes tend to have a eaiser time | 19:00 |
| Nisha | NobodyCam: i didnt get ^^^^^^ | 19:01 |
| Shrews | adam_g: looks like the only actual tempest cinder test to fail with ironic is test_attach_volume.AttachVolumeTestXML.test_attach_detach_volume. I don't think we need to disable all of cinder in d-g, just skip that one test. | 19:03 |
| Shrews | adam_g: unless you saw other failures that i'm not aware of | 19:03 |
| NobodyCam | I was just thinking that might be easier if we split out creating of the inventory field and have both of the current specs depend on it | 19:04 |
| NobodyCam | that way its very clear as to which spec is adding what | 19:05 |
| NobodyCam | that is just my option thou | 19:05 |
| Nisha | Yes thats also an option, agreed. | 19:05 |
| NobodyCam | devananda: are you around? I have a spec-review question | 19:08 |
| NobodyCam | and you are the only spec core member that may be on line | 19:08 |
| Shrews | NobodyCam: they're all online, they're just all ignoring you :-P | 19:11 |
| NobodyCam | lol.. | 19:12 |
| matty_dubs | If I may ask a dumb question... is driver-list meant to list only active drivers, or all available? | 19:13 |
| NobodyCam | actually several are away, dmitry is on pto, lucas is (or should be) in bed sleeping, russell on PTO | 19:13 |
| matty_dubs | And if the former, can I make it do the latter? | 19:13 |
| Nisha | matty_dubs: it gives only the active drivers | 19:14 |
| matty_dubs | Nisha: Thanks. That's what it seemed to be doing, but not what I was expecting. I swear there was some way I got it to spit out all drivers it knew about previously. | 19:15 |
| *** ryanpetrello has joined #openstack-ironic | 19:15 | |
| NobodyCam | matty_dubs: this setting? https://github.com/openstack/ironic/blob/master/etc/ironic/ironic.conf.sample#L229 | 19:15 |
| Nisha | matty_dubs: there is a change in activating the drivers now. You have to add the driver name in ironic.conf | 19:16 |
| Nisha | in enabled_drivers | 19:16 |
| NobodyCam | matty_dubs: or https://github.com/openstack/ironic/blob/master/setup.cfg#L32 | 19:16 |
| matty_dubs | Ah, maybe that change is what threw me. What I was looking for, and it's admittedly somewhat academic now, is sort of a list for possible values of enabled_drivers= | 19:16 |
| matty_dubs | NobodyCam: Ah! That second one is it it. | 19:17 |
| matty_dubs | So I'm not crazy! | 19:17 |
| NobodyCam | matty_dubs: that is ALL drivers :) | 19:17 |
| matty_dubs | (At least on this subject.) | 19:17 |
| *** zdiN0bot has quit IRC | 19:17 | |
| *** Penick has quit IRC | 19:19 | |
| *** steveh has quit IRC | 19:20 | |
| *** steveh has joined #openstack-ironic | 19:22 | |
| *** Penick has joined #openstack-ironic | 19:28 | |
| *** overlayer has joined #openstack-ironic | 19:30 | |
| *** Penick has quit IRC | 19:32 | |
| *** overlayer has quit IRC | 19:50 | |
| *** coolsvap is now known as coolsvap|afk | 19:52 | |
| ryanpetrello | hey folks, look like the stable Ironic tests are broken again: http://logs.openstack.org/47/105547/4/check/gate-pecan-tox-ironic-stable/2639f1a/ :\ | 19:55 |
| ryanpetrello | this look familiar to anybody? | 19:55 |
| *** coolsvap|afk is now known as coolsvap | 19:55 | |
| Shrews | ryanpetrello: due to an updated glanceclient. latest ironic HEAD has the fix | 19:55 |
| ryanpetrello | can these sorts of things be backported? | 19:56 |
| ryanpetrello | or should I just not rely on stable branch tests to be in a passing state? | 19:56 |
| ryanpetrello | if that’s the case, I’m fine w/ it | 19:56 |
| ryanpetrello | it’s just that I gate pecan against stable ironic to ensure I don’t break anything | 19:56 |
| ryanpetrello | but most of the time, it’s the ironic branch that’s holding *me* up with breakage | 19:56 |
| *** romcheg has quit IRC | 19:57 | |
| Shrews | you wouldn't have run across this in a stable branch unless you (or something) upgraded python-glanceclient | 19:57 |
| *** dwalleck has quit IRC | 19:57 | |
| Shrews | as far as backporting, ask devananda | 19:58 |
| devananda | righ | 19:58 |
| devananda | right | 19:58 |
| *** coolsvap1 has joined #openstack-ironic | 19:58 | |
| *** coolsvap is now known as coolsvap|afk | 19:58 | |
| ryanpetrello | Shrews: well, you would if you cloned stable/icehouse and ran the tests | 19:58 |
| devananda | so if we're going to commit to a stable branch (which i think we already have) | 19:58 |
| devananda | then we need a jenkins job testing it | 19:59 |
| ryanpetrello | +1 | 19:59 |
| devananda | ryanpetrello: when / how often do other projects which you're gating with pecan test their stable branches? | 19:59 |
| NobodyCam | devananda: oh that sounds like a good idea | 19:59 |
| devananda | nightly? on every commit? in their gate or just check queue? | 20:00 |
| *** coolsvap1 has quit IRC | 20:00 | |
| devananda | also - the glanceclient failure itself means either glanceclient made a non-backwards-compatible change (which they shouldn't) or we're using it in a way they didn't expect | 20:01 |
| ryanpetrello | devananda: not sure | 20:01 |
| ryanpetrello | they may not be at all, either | 20:01 |
| ryanpetrello | but they don’t have as much history and touching points w/ other projects as Ironic does | 20:01 |
| ryanpetrello | I suspect something like nightly would probably be fine, though | 20:02 |
| ryanpetrello | just some periodic way to know “Hey, a current checkout of stable is broken.” | 20:02 |
| *** zdiN0bot has joined #openstack-ironic | 20:02 | |
| devananda | Shrews: backport ++. IIRC it was a small patch | 20:03 |
| devananda | anyone want to propose a stable/icehouse job to -infra? :) | 20:03 |
| * devananda lurks in the TC meeting while booking more travel | 20:03 | |
| ryanpetrello | Shrews: I can backport it if you like | 20:05 |
| ryanpetrello | https://review.openstack.org/#/c/103670/6 , is looks like ? | 20:05 |
| ryanpetrello | *it looks like | 20:05 |
| *** shardy_afk is now known as shardy | 20:05 | |
| *** ellenh_ has joined #openstack-ironic | 20:05 | |
| Shrews | ryanpetrello: yep | 20:06 |
| ryanpetrello | k, will do | 20:06 |
| *** romcheg has joined #openstack-ironic | 20:07 | |
| adam_g | Shrews, IIRC theres a number of scenario tests that deal with voluem attach, create volume from image, etc | 20:19 |
| *** Nisha has quit IRC | 20:19 | |
| adam_g | Shrews, may be able to get away adding a feature flag for volume attachment? i dunno | 20:20 |
| Shrews | adam_g: ah | 20:20 |
| adam_g | Shrews, it was suggested that cinder be disabled, since ironic itself provides no support for volumes and AFAIK cinder doesn't really have any use cases without compute support | 20:22 |
| devananda | adam_g: there is a case for cinder with ironic. two, in fact | 20:23 |
| adam_g | devananda, do tell | 20:24 |
| devananda | 1) create cinder volume, pass attach data via cloud init | 20:24 |
| devananda | 2) create bootable cinder volume, use (OOB channel || iPXE chainload) to boot from that volume | 20:24 |
| adam_g | devananda, does 1 depend on agent? | 20:25 |
| devananda | neither one are working today AFAIK | 20:25 |
| devananda | no | 20:25 |
| devananda | it probably requires some tweaks to the in-image init scripts (ie, dib element) | 20:26 |
| devananda | but other than that, should work, i think | 20:26 |
| soren | I feel like I'm really, really close to having this all working. Right now, my problem seems to be that dnsmasq isn't receiving dhcp requests from my nodes. Is anyone willing to share the output of their "ovs-vsctl show" with me? | 20:26 |
| devananda | the iPXE chainload stuff obviously depends on ipxe (lucas' spec is close to beign approved) and then a little scripting to get that right | 20:26 |
| adam_g | devananda, the guest is expected to connect to volume somehow? | 20:27 |
| devananda | adam_g: yes | 20:27 |
| devananda | it's not exactly the same as virt .... heh | 20:27 |
| devananda | short story - i dont think we should be blocking tests on these today | 20:27 |
| devananda | but feature wise, it'd be nice to get there in K | 20:27 |
| adam_g | devananda, seems like the guest would need lots of knowledge of every supported cinder backends | 20:27 |
| devananda | adam_g: let's assume iSCSI ... :) | 20:28 |
| devananda | soren: i can paste that from my devstack, but not sure how helpful that'll be | 20:28 |
| adam_g | actually the cinder attach data would likely contain lots of sensitive infoz you'd not want to expose to guests, meant for hypervisors-eyes-only | 20:28 |
| soren | This is what I'm looking at right now: http://paste.openstack.org/show/85708/ | 20:28 |
| soren | devananda: I'm not sure either. | 20:29 |
| devananda | adam_g: trusted guests, then. like tripleo undercloud | 20:29 |
| soren | Ok, so in that paste, I see this problem: | 20:29 |
| soren | dnsmasq has --interface=tapdf373ea3-7b | 20:29 |
| soren | Yet that interface is connected to br-int, not br-ctlplane. | 20:30 |
| soren | So of course the DHCP requests coming in on eth2 aren't ever getting to dnsmasq. | 20:30 |
| soren | I just can't work out what I'm doing wrong. | 20:30 |
| soren | Any ideas about where to look? | 20:33 |
| soren | I strace'd the dnsmasq process as the other node was attempting to PXE. Nothing. | 20:33 |
| *** overlayer has joined #openstack-ironic | 20:35 | |
| soren | Does anything look wrong to you here? http://paste.openstack.org/show/85710/ (added more info since last paste) | 20:39 |
| *** overlayer has quit IRC | 20:40 | |
| soren | Can someone with a running, physical Ironic setup check how their br-ctlplane is connected and maybe help me understand how dhcp requests are meant to find their way to dnsmasq? | 20:44 |
| *** tatyana has joined #openstack-ironic | 20:48 | |
| adam_g | soren, this isn't physical hardware but http://paste.openstack.org/show/85711/ | 20:49 |
| soren | adam_g: I wonder where int-br-ctlplane came from? | 20:49 |
| adam_g | soren, also, this is a tripleo devtest setup | 20:50 |
| soren | adam_g: Hm, hang on, I may have something here. | 20:50 |
| NobodyCam | soren: this is from my seed vm: http://paste.openstack.org/show/pC9NBnRn5x5pgr5D1Tmj/ | 20:51 |
| * soren begins fuming | 20:53 | |
| NobodyCam | ??? | 20:54 |
| soren | Sorry, not you. | 20:55 |
| soren | I just found this in my upstart job definition: | 20:55 |
| openstackgerrit | Jim Rollenhagen proposed a change to openstack/ironic: Allow Ironic URL from config file https://review.openstack.org/105590 | 20:56 |
| jroll | devananda: this touches nova driver ^ | 20:56 |
| *** romcheg has quit IRC | 20:56 | |
| jroll | devananda: but is a quick fix and unbreaks a thing :) | 20:56 |
| soren | "--config-file=/etc/neutron/plugins/ml2/ml2_conf.ini" | 20:57 |
| soren | So it's been ignoring my ovs config. | 20:57 |
| devananda | jroll: ack | 20:57 |
| *** jdob has quit IRC | 21:01 | |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Add Node decommission in conductor https://review.openstack.org/102984 | 21:13 |
| *** tatyana has quit IRC | 21:16 | |
| NobodyCam | back in a bit | 21:17 |
| *** zdiN0bot has quit IRC | 21:18 | |
| openstackgerrit | Ellen Hui proposed a change to openstack/ironic: Add more INFO logging to ironic/common https://review.openstack.org/99495 | 21:18 |
| *** jbjohnso has quit IRC | 21:20 | |
| soren | NobodyCam, adam_g: It totally works now. It was the wrong plugin config all along. | 21:21 |
| adam_g | soren, doh | 21:22 |
| zer0c00l | Just a quick question, i have been following the utility ramdisk blueprint. Is it meant to be a replacement for all kind of operating system installers? | 21:23 |
| zer0c00l | Does ironic support UEFI/secure boots? | 21:23 |
| jroll | zer0c00l: which blueprint? (link)? | 21:23 |
| jroll | 2) not yet, there's a spec up | 21:23 |
| zer0c00l | https://blueprints.launchpad.net/ironic/+spec/utility-ramdisk | 21:23 |
| zer0c00l | jroll: ^^ | 21:24 |
| jroll | ok yeah, so that is kind of out of date a bit | 21:24 |
| *** matty_dubs is now known as matty_dubs|gone | 21:24 | |
| jroll | and has turned into https://github.com/openstack/ironic-python-agent + this blueprint https://review.openstack.org/#/c/98506/ | 21:24 |
| jroll | zer0c00l: the idea is, small python app that runs in a ramdisk and has a rest api | 21:25 |
| jroll | that rest api has endpoints to do things like: write image, reboot, erase disks, update firmware, write bios settings, etc | 21:26 |
| zer0c00l | i see | 21:26 |
| zer0c00l | Does it support things like partitioning | 21:27 |
| zer0c00l | Also how bootloader installation is handled ,we have grub2 grub1 etc for various distros | 21:27 |
| jroll | it's... getting there | 21:27 |
| NobodyCam | soren: awesone | 21:28 |
| jroll | right now it only supports whole disk images | 21:28 |
| zer0c00l | i see | 21:28 |
| zer0c00l | thanks for the info jroll | 21:28 |
| Shrews | adam_g, devananda: https://review.openstack.org/105599 | 21:28 |
| jroll | zer0c00l: but I think we landed some partitioning tools at some point | 21:29 |
| jroll | and plan to support partitioning etc | 21:29 |
| jroll | zer0c00l: I'm one of the cores on the agent project, feel free to ping if you want to chat more about it whenever :) | 21:30 |
| zer0c00l | Thanks, i will :) | 21:30 |
| devananda | Shrews: i suspect that'll be contentious, and i'm not convinced yet myself that it should be disabled | 21:35 |
| Shrews | devananda: i'm open to other options | 21:37 |
| Shrews | devananda: feel free to -1 the review if you think it needs more discussion | 21:38 |
| adam_g | devananda, the alternative would be to add a 'volume_attach' compute feature flag to tempest and skip anything related (in addtion to existing skips that skip based on cinder being available) | 21:38 |
| adam_g | volume attach is a feature on https://wiki.openstack.org/wiki/HypervisorSupportMatrix, so it might be a welcomed flag | 21:38 |
| *** athomas has quit IRC | 21:38 | |
| *** overlayer has joined #openstack-ironic | 21:39 | |
| *** rloo has quit IRC | 21:44 | |
| ryanpetrello | so an attempted backport of mine https://review.openstack.org/#/c/105583/ failed with http://logs.openstack.org/83/105583/1/check/check-tempest-dsvm-virtual-ironic/848e3b7/logs/devstacklog.txt.gz#_2014-07-08_20_32_42_828 :\ | 21:47 |
| ryanpetrello | this look familiar to anybody? | 21:47 |
| ryanpetrello | or have any thoughts as to what else I should look for? | 21:47 |
| Shrews | adam_g: devananda: clarkb suggested a newer, fancier way to disable that. i won't push up that change until we decide which route to take | 21:48 |
| NobodyCam | jroll: I feel like I missed something this morning . where we going to chat with Lucas about "SOMETHING"? | 21:48 |
| Shrews | calling it a night, everyone. later | 21:49 |
| NobodyCam | night Shrews :) | 21:50 |
| NobodyCam | jroll: ahh yes... the TFTP patches | 21:50 |
| mrda | Good morning Ironic! | 21:51 |
| NobodyCam | good mornig mrda | 21:51 |
| mrda | \o | 21:51 |
| devananda | night, Shrews ! | 21:55 |
| devananda | o/ mrda :) | 21:55 |
| mrda | hey deva | 21:56 |
| *** ellenh_ has quit IRC | 21:56 | |
| *** ellenh_ has joined #openstack-ironic | 21:57 | |
| *** jcoufal has quit IRC | 22:03 | |
| jroll | NobodyCam: ohhh right | 22:06 |
| jroll | NobodyCam: I might look again and decide I'll just base it off of his patch | 22:06 |
| NobodyCam | :) | 22:06 |
| jroll | but... we'll see | 22:06 |
| NobodyCam | lol | 22:07 |
| NobodyCam | if not its on my list to chat with lucas tomorrow | 22:07 |
| NobodyCam | :) | 22:07 |
| jroll | cool | 22:08 |
| NobodyCam | anyone not on Ubuntu.. can you check your users home directory for a .cache directory? | 22:19 |
| *** ellenh_ has quit IRC | 22:20 | |
| jroll | NobodyCam: exists on debian | 22:20 |
| NobodyCam | :) | 22:21 |
| NobodyCam | anyone have fedora? | 22:21 |
| NobodyCam | handy? | 22:21 |
| mgagne | NobodyCam: fresh install with nothing? I have centos 6 | 22:21 |
| NobodyCam | mgagne: :) so you do not have .cache? | 22:22 |
| mgagne | nothing in there | 22:22 |
| NobodyCam | :) | 22:22 |
| NobodyCam | TY | 22:22 |
| *** ellenh_ has joined #openstack-ironic | 22:22 | |
| jroll | NobodyCam: out of curiousity, why are you asking? | 22:22 |
| mgagne | not sure who is supposed to create it... | 22:22 |
| NobodyCam | it from a review... (COMPLETION_CACHE=~/.ironicclient/*/*-cache) is the line. I'm going to suggest that it get moved to .cache/python-ironicclient/ | 22:24 |
| jroll | ah | 22:24 |
| mgagne | NobodyCam: http://askubuntu.com/questions/22361/what-is-the-relationship-between-gconf-gnome2-cache-local-and-other-dot | 22:24 |
| NobodyCam | gah and noow I have to change my corp password :-p | 22:24 |
| mgagne | NobodyCam: 2011 but still | 22:24 |
| *** derekh_ has joined #openstack-ironic | 22:24 | |
| mrda | there's enough crap in $HOME/.cache that I think it's a safe place to store things | 22:26 |
| mgagne | NobodyCam: See $XDG_CACHE_HOME on http://standards.freedesktop.org/basedir-spec/basedir-spec-latest.html#variables | 22:26 |
| NobodyCam | mgagne: with out a gui I dont have XDG_CACHE_HOME set in my env. but good post to point out :) | 22:29 |
| *** zdiN0bot has joined #openstack-ironic | 22:30 | |
| *** jgrimm has quit IRC | 22:33 | |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Nova fast deletes while decommissioning https://review.openstack.org/105614 | 22:41 |
| *** dwalleck has joined #openstack-ironic | 22:44 | |
| *** dwalleck has quit IRC | 22:45 | |
| *** dwalleck has joined #openstack-ironic | 22:45 | |
| *** overlayer has quit IRC | 22:46 | |
| *** ellenh_ has quit IRC | 22:55 | |
| *** Penick has joined #openstack-ironic | 22:55 | |
| NobodyCam | anyone know Andrey Kurilin's irc handle? | 22:56 |
| *** Penick_ has joined #openstack-ironic | 22:56 | |
| *** Penick has quit IRC | 22:59 | |
| *** Penick_ is now known as Penick | 22:59 | |
| *** ellenh_ has joined #openstack-ironic | 23:00 | |
| *** blamar_ has joined #openstack-ironic | 23:02 | |
| *** blamar has quit IRC | 23:04 | |
| *** blamar_ is now known as blamar | 23:04 | |
| *** dwalleck has quit IRC | 23:06 | |
| mgagne | NobodyCam: LP says its andreykurilin | 23:06 |
| mgagne | NobodyCam: marked as away on #openstack-dev | 23:07 |
| NobodyCam | ya | 23:12 |
| NobodyCam | mgagne: Ty :) | 23:13 |
| *** hemna has quit IRC | 23:14 | |
| *** zdiN0bot has quit IRC | 23:15 | |
| devananda | harlowja: hiya! just saw your email to the dev list about mad rechecks -- somehow gmail thinks it is spam | 23:18 |
| harlowja | devananda gmail not like yahoo-inc? | 23:18 |
| harlowja | haha | 23:18 |
| harlowja | those google folks | 23:18 |
| devananda | harlowja: thought you'd like to know that :) | 23:18 |
| devananda | harlowja: also, you're not alone. that sort of came up in the project meeting today, too | 23:19 |
| harlowja | devananda which part came up, the part about a 'stability cycle' ? | 23:19 |
| devananda | even if no one replied ;) | 23:19 |
| harlowja | all my emails go directly to peoples brains, no replies needed | 23:19 |
| devananda | harlowja: no. the problem with testing too many unstable things together | 23:19 |
| devananda | harlowja: to which that might be a possible remedy | 23:20 |
| harlowja | gotcha, interting | 23:20 |
| harlowja | #openstack-meeting? | 23:20 |
| devananda | harlowja: my patch to create an empty __init__.py file in Nova failed 5/6 times, IIRC | 23:20 |
| devananda | ya | 23:20 |
| devananda | look in the "project" meeting logs | 23:20 |
| harlowja | woah, are u really sure it was empty devananda | 23:20 |
| harlowja | did it have a comment without a period :-P | 23:21 |
| devananda | harlowja: https://review.openstack.org/#/c/103164/ | 23:21 |
| harlowja | or with a period (i can never remember) | 23:21 |
| harlowja | ya, its an interesting dillema, reading project logs | 23:21 |
| devananda | 5 fails 2 pass | 23:21 |
| harlowja | in vegas that would not be a happy rate, lol | 23:22 |
| harlowja | all reviews should have a vegas meter | 23:22 |
| *** ellenh_ has quit IRC | 23:22 | |
| harlowja | *i guess the casino would be happy | 23:22 |
| * devananda afk's for a while | 23:26 | |
| *** derekh_ has quit IRC | 23:41 | |
| *** chuckC has quit IRC | 23:44 | |
| *** chuckC has joined #openstack-ironic | 23:45 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!