| *** Haomeng|2 has joined #openstack-ironic | 00:00 | |
| *** harlowja is now known as harlowja_away | 00:01 | |
| *** Haomeng has quit IRC | 00:01 | |
| *** praneshp has joined #openstack-ironic | 00:03 | |
| *** ccit has joined #openstack-ironic | 00:04 | |
| *** ccit has left #openstack-ironic | 00:05 | |
| NobodyCam | have a good night ya'll see ya in the morning. | 00:10 |
|---|---|---|
| JayF | night | 00:10 |
| NobodyCam | :) | 00:10 |
| *** harlowja_away is now known as harlowja | 00:11 | |
| rloo_afk | dlaube: fwiw, we're in the middle of updating the install guide. https://review.openstack.org/#/c/118614/ | 00:12 |
| takadayuiko | Good morning, Ironic | 00:13 |
| rloo_afk | dlaube: there's some stuff there about glance but I don't recall now how accurate it is. but I think the stuff from that baremetal wiki is/was/should basically be the same/similar. | 00:13 |
| rloo_afk | morning takadayuiko | 00:13 |
| *** praneshp has quit IRC | 00:14 | |
| takadayuiko | rloo_afk, Hi :) | 00:14 |
| *** openstackstatus has quit IRC | 00:23 | |
| *** openstackstatus has joined #openstack-ironic | 00:23 | |
| *** ChanServ sets mode: +v openstackstatus | 00:24 | |
| *** praneshp has joined #openstack-ironic | 00:25 | |
| dlaube | cool, thank you rloo_afk | 00:25 |
| *** praneshp has quit IRC | 00:30 | |
| * devananda steps away for a few hours | 00:34 | |
| *** smoriya has joined #openstack-ironic | 00:35 | |
| takadayuiko | rloo_afk, https://review.openstack.org/#/c/118614/ will be a very nice install guide, but there is no user guide doint baremetal provisioning with REAL-PHYSICAL-SERVER using "ironic node-create", then "nova boot". | 00:48 |
| takadayuiko | there is only http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#deploying-ironic-with-devstack using FAKE driver, isn't it? | 00:48 |
| jroll | takadayuiko: that guide is for real physical hardware | 00:51 |
| jroll | it just doesn't have the nova boot command I guess? | 00:51 |
| jroll | because it's for setting up ironic, not for using nova to boot an ironic server (which is easy, normal nova boot, different flavor) | 00:52 |
| jroll | /b 76 | 00:52 |
| jroll | oops | 00:52 |
| takadayuiko | jroll, yes, for real physical hardware. And doesn't have commands ironic and nova, right? just preparation, not baremetal provisioniong. | 00:54 |
| jroll | right, it doesn't cover adding a node to ironic or booting a node through nova | 00:54 |
| jroll | but that is the same regardless of if hardware is real or not | 00:54 |
| jroll | the driver may change, which may change node-create, but that should be in the docs for that driver | 00:55 |
| takadayuiko | ironic developers feel "it's necessary to do ironic node-create then nova boot ordinarily, it's easy", but beginners don't feel so. | 00:55 |
| jroll | I'm not saying it's easy, I'm saying it's documented elsewhere | 00:56 |
| takadayuiko | so, perhaps, for example by just writting "after this, please see the docment xxx..." will be better | 00:56 |
| jroll | sure | 00:57 |
| jroll | you should put that on the review :) | 00:57 |
| takadayuiko | hmm, I'll do so. thanks :) | 00:57 |
| jroll | you're welcome :) | 00:58 |
| rloo_afk | jroll, takadayuiko: yes, we're missing that part of the documentation. I guess it is more of an operator's guide? or maybe we can throw it into the same install guide. will have to think about it. | 00:58 |
| jroll | I presume if you deploy ironic, you'll want to verify that it works by provisioning a server :) | 00:59 |
| takadayuiko | jroll, +1. rloo_afk, I think so, too | 00:59 |
| rloo_afk | jroll: yeah. I suspect it'd be easier for us to throw all that info in the install guide, than to start a new page etc. | 00:59 |
| jroll | indeed | 01:00 |
| takadayuiko | agree | 01:00 |
| rloo_afk | takadayuiko: you could add a comment to that patch or open a bug about it. and you're welcome to write that up too ;) | 01:00 |
| takadayuiko | rloo_afk, yes, I'll write comments to that patch. | 01:01 |
| *** Haomeng has joined #openstack-ironic | 01:05 | |
| *** Haomeng|2 has quit IRC | 01:06 | |
| *** marcoemorais has quit IRC | 01:10 | |
| *** Kui has quit IRC | 01:13 | |
| *** Kui has joined #openstack-ironic | 01:14 | |
| *** rloo_afk has quit IRC | 01:21 | |
| *** todd_dsm has joined #openstack-ironic | 01:27 | |
| *** Haomeng|2 has joined #openstack-ironic | 01:27 | |
| *** lazy_prince has quit IRC | 01:27 | |
| *** achanda has joined #openstack-ironic | 01:29 | |
| *** Haomeng has quit IRC | 01:30 | |
| *** killer_prince has joined #openstack-ironic | 01:49 | |
| *** achanda_ has joined #openstack-ironic | 01:49 | |
| *** killer_prince is now known as lazy_prince | 01:49 | |
| *** achanda has quit IRC | 01:52 | |
| *** todd_dsm has quit IRC | 01:56 | |
| *** achanda_ has quit IRC | 02:00 | |
| *** achanda has joined #openstack-ironic | 02:02 | |
| *** lazy_prince has quit IRC | 02:05 | |
| *** lazy_prince has joined #openstack-ironic | 02:05 | |
| *** nosnos has joined #openstack-ironic | 02:06 | |
| *** chenglch has joined #openstack-ironic | 02:34 | |
| *** openstackgerrit has quit IRC | 02:45 | |
| *** rainya has joined #openstack-ironic | 02:55 | |
| *** coolsvap|afk is now known as coolsvap | 02:55 | |
| *** pcrews has quit IRC | 03:03 | |
| *** ramineni has joined #openstack-ironic | 03:05 | |
| *** dlaube has quit IRC | 03:09 | |
| *** rushiagr_away is now known as rushiagr | 03:12 | |
| *** lazy_prince is now known as killer_prince | 03:13 | |
| *** achanda has quit IRC | 03:19 | |
| *** achanda has joined #openstack-ironic | 03:24 | |
| *** penick has joined #openstack-ironic | 03:26 | |
| *** chuckC_ has joined #openstack-ironic | 03:27 | |
| *** penick_ has joined #openstack-ironic | 03:27 | |
| *** killer_prince has quit IRC | 03:28 | |
| *** harlowja is now known as harlowja_away | 03:29 | |
| *** lazy_prince has joined #openstack-ironic | 03:29 | |
| *** penick has quit IRC | 03:30 | |
| *** penick_ is now known as penick | 03:30 | |
| *** harlowja_away is now known as harlowja | 03:31 | |
| *** loki184 has joined #openstack-ironic | 03:33 | |
| *** ChuckC__ has joined #openstack-ironic | 03:34 | |
| *** nosnos has quit IRC | 03:35 | |
| *** coolsvap is now known as coolsvap|afk | 03:37 | |
| *** ChuckC__ has quit IRC | 03:37 | |
| *** achanda_ has joined #openstack-ironic | 03:39 | |
| *** achanda has quit IRC | 03:42 | |
| *** Haomeng|2 has quit IRC | 03:45 | |
| *** coolsvap|afk is now known as coolsvap | 03:45 | |
| *** spandhe has quit IRC | 03:49 | |
| *** ryanpetrello has quit IRC | 03:50 | |
| *** ryanpetrello has joined #openstack-ironic | 03:50 | |
| *** penick has quit IRC | 03:53 | |
| *** rushiagr is now known as rushiagr_away | 04:08 | |
| *** todd_dsm has joined #openstack-ironic | 04:18 | |
| *** spandhe has joined #openstack-ironic | 04:19 | |
| *** todd_dsm has quit IRC | 04:21 | |
| *** marcoemorais has joined #openstack-ironic | 04:26 | |
| *** vinbs has joined #openstack-ironic | 04:28 | |
| *** marcoemorais1 has joined #openstack-ironic | 04:30 | |
| *** marcoemorais has quit IRC | 04:33 | |
| *** nosnos has joined #openstack-ironic | 04:35 | |
| *** rushiagr_away is now known as rushiagr | 04:37 | |
| *** Poornima has joined #openstack-ironic | 04:38 | |
| *** rainya has quit IRC | 04:40 | |
| *** loki184 has quit IRC | 04:42 | |
| *** loki184 has joined #openstack-ironic | 04:47 | |
| *** lazy_prince has quit IRC | 04:50 | |
| *** loki184 has quit IRC | 04:53 | |
| *** lazy_prince has joined #openstack-ironic | 04:57 | |
| *** achanda has joined #openstack-ironic | 05:09 | |
| *** achanda_ has quit IRC | 05:12 | |
| *** rainya has joined #openstack-ironic | 05:16 | |
| *** achanda_ has joined #openstack-ironic | 05:16 | |
| *** rainya has quit IRC | 05:18 | |
| *** achanda has quit IRC | 05:20 | |
| *** jcoufal has joined #openstack-ironic | 05:22 | |
| *** achanda has joined #openstack-ironic | 05:24 | |
| *** achanda_ has quit IRC | 05:27 | |
| *** achanda_ has joined #openstack-ironic | 05:29 | |
| *** achanda has quit IRC | 05:32 | |
| *** harlowja is now known as harlowja_away | 05:35 | |
| *** chenglch has quit IRC | 05:51 | |
| *** rakesh_hs has joined #openstack-ironic | 05:51 | |
| *** chenglch has joined #openstack-ironic | 05:51 | |
| *** achanda_ has quit IRC | 05:53 | |
| *** loki184 has joined #openstack-ironic | 05:59 | |
| *** coolsvap is now known as coolsvap|afk | 06:00 | |
| *** loki184 has quit IRC | 06:09 | |
| *** loki184 has joined #openstack-ironic | 06:09 | |
| *** coolsvap|afk is now known as coolsvap | 06:12 | |
| *** achanda has joined #openstack-ironic | 06:16 | |
| *** achanda has quit IRC | 06:20 | |
| *** pradipta_away is now known as pradipta | 06:27 | |
| *** achanda has joined #openstack-ironic | 06:29 | |
| *** pensu has joined #openstack-ironic | 06:31 | |
| *** achanda has quit IRC | 06:34 | |
| *** pcrews has joined #openstack-ironic | 06:39 | |
| *** pcrews has quit IRC | 06:47 | |
| *** pradipta is now known as pradipta_away | 06:49 | |
| *** pradipta_away is now known as pradipta | 06:51 | |
| *** spandhe has quit IRC | 06:53 | |
| GheRivero | morning all | 06:56 |
| *** k4n0 has joined #openstack-ironic | 06:56 | |
| *** achanda has joined #openstack-ironic | 06:59 | |
| *** k4n0 has quit IRC | 07:02 | |
| takadayuiko | GheRivero, Hi | 07:05 |
| *** andreykurilin_ has joined #openstack-ironic | 07:06 | |
| *** andreykurilin_ has quit IRC | 07:13 | |
| *** k4n0 has joined #openstack-ironic | 07:15 | |
| *** ifarkas has joined #openstack-ironic | 07:19 | |
| *** jistr has joined #openstack-ironic | 07:23 | |
| *** achanda has quit IRC | 07:25 | |
| *** romcheg has joined #openstack-ironic | 07:31 | |
| *** killer_prince has joined #openstack-ironic | 07:31 | |
| *** andreykurilin_ has joined #openstack-ironic | 07:35 | |
| *** k4n0 has quit IRC | 07:42 | |
| *** andreykurilin_ has quit IRC | 07:44 | |
| *** Haomeng has joined #openstack-ironic | 07:45 | |
| *** achanda has joined #openstack-ironic | 07:50 | |
| *** dtantsur|afk is now known as dtantsur | 07:50 | |
| dtantsur | Morning Ironic, morning, GheRivero, takadayuiko :) | 07:50 |
| takadayuiko | dtantsur, Hi! | 07:51 |
| Haomeng | morning Ironic:) | 07:51 |
| takadayuiko | Haomeng, Hi :) | 07:52 |
| Haomeng | takadayuiko: :) | 07:53 |
| *** k4n0 has joined #openstack-ironic | 07:55 | |
| dtantsur | Haomeng, morning | 07:56 |
| Haomeng | dtantsur: morning :) | 07:57 |
| *** lazy_prince has quit IRC | 07:58 | |
| *** killer_prince is now known as lazy_prince | 07:58 | |
| yuriyz | morning Ironic | 07:59 |
| Haomeng | I have a question, we can see some advanced feature in IPA(https://wiki.openstack.org/wiki/Ironic-python-agent) such as configure RAID, but can not find the code in our current IPA | 08:00 |
| Haomeng | maybe these advanced agent feature are not contributed back from rackspace, not sure:) | 08:00 |
| Haomeng | any comments? :) | 08:01 |
| *** achanda has quit IRC | 08:01 | |
| GheRivero | I think it is planned and waiting to be coded | 08:01 |
| Haomeng | GheRivero: ok, got, thanks:) | 08:03 |
| *** athomas has joined #openstack-ironic | 08:03 | |
| Haomeng | GheRivero: but still has concern, our ironic will not focus on such hardware management functions, raid configuration, firmware updating etc | 08:03 |
| GheRivero | I don't know how people form Rackspace want to do it, but the agent could be intelligent enough to handle all of it without requiring any data from ironic (server) | 08:06 |
| Haomeng | GheRivero: yes, agree with you | 08:06 |
| Haomeng | agent should work with itself, and our ironic conductor just take as controller role I think | 08:07 |
| GheRivero | we don't want ironic to be a cmdb | 08:08 |
| dtantsur | "not a cmdb" relates more to discovery, it has barely anything to do with raid configuration, firmware updating etc | 08:10 |
| dtantsur | it's more about "ready-state" talks and ability to securely implement metal-to-tenant | 08:10 |
| *** jistr has quit IRC | 08:12 | |
| GheRivero | the raid config, firmware versions... should be stored some place to be accessed by IPA | 08:13 |
| devananda | GheRivero: there are drivers other than IPA which may need that info, and may require ironic to pass it directly to the BMC | 08:15 |
| devananda | so a model where IPA polls an external-to-ironic service to gather that info is not generalizable to other drivers | 08:16 |
| dtantsur | devananda, g'evening. are you going to sleep today? ;) | 08:18 |
| dtantsur | and yes, good point about generalization | 08:18 |
| devananda | dtantsur: maybe? | 08:18 |
| devananda | I should ... | 08:18 |
| * devananda is hacking on ansible | 08:19 | |
| Haomeng | devananda: good night:) | 08:19 |
| devananda | Haomeng: hi! | 08:19 |
| Haomeng | devananda: :) | 08:19 |
| devananda | dtantsur: i think ready-state, as Rob uses the term, has some cognitive mismatch to how Ironic treats hardware as cattle | 08:20 |
| devananda | dtantsur: namely, what ever operations need to be performed between "machine is racked and I can power it on" and "there's a user on it" | 08:20 |
| devananda | which is what ready-state is about | 08:20 |
| dtantsur | I'm not aware of details, so I'm using it as more general concept, like "we probably need something..." :) | 08:21 |
| devananda | need to be performed *every* time a new tenant is put on the box | 08:21 |
| dtantsur | heh I envision an interesting discussion on the summit | 08:21 |
| devananda | words are challenging because they often mean different things to different people :) | 08:21 |
| *** marcoemorais1 has quit IRC | 08:22 | |
| devananda | who needs words? let's just use ... um ... paintings. and coffee. | 08:22 |
| devananda | wow, i should not drink coffee at 1am | 08:22 |
| devananda | dtantsur: I think it'll be a fascinating conversation. I'd actually love it if we can get Rob to join | 08:23 |
| dtantsur | yeah | 08:23 |
| devananda | he and I talked a tiny bit about crowbar in Atlanta | 08:24 |
| *** takadayuiko has quit IRC | 08:25 | |
| *** lucasagomes has joined #openstack-ironic | 08:25 | |
| *** takadayuiko has joined #openstack-ironic | 08:26 | |
| *** derekh has joined #openstack-ironic | 08:30 | |
| *** pradipta is now known as pradipta_away | 08:30 | |
| *** jistr has joined #openstack-ironic | 08:31 | |
| *** chenglch has quit IRC | 08:31 | |
| *** chenglch has joined #openstack-ironic | 08:33 | |
| *** loki184 has quit IRC | 08:37 | |
| *** chenglch|2 has joined #openstack-ironic | 08:38 | |
| takadayuiko | lucasagomes, thank you for having teach me about driver_info day before yesterday :) | 08:39 |
| takadayuiko | I've read nova/virt/ironic/patcher.py and (maybe) understand current status. sorry for taking long time to understand. | 08:39 |
| lucasagomes | takadayuiko, g'evening :) oh np at all! ur very welcome | 08:39 |
| lucasagomes | takadayuiko, it's all good :) | 08:39 |
| *** chenglch has quit IRC | 08:40 | |
| *** vdrok_ has joined #openstack-ironic | 08:40 | |
| takadayuiko | lucasagomes, haha, g'evening or g'morning :) | 08:40 |
| takadayuiko | as you said, ironic node should have pxe_deploy_kernel and pxe_deploy_ramdisk information as driver_info, right? | 08:41 |
| lucasagomes | takadayuiko, :D here's good morning | 08:41 |
| lucasagomes | yeah... before it was in the flavor, but we want to get rid of the behavior of adding it to the flavor | 08:41 |
| lucasagomes | because otherwise we can't use the same flavor for diff drivers | 08:41 |
| takadayuiko | in Kilo, isn't it? | 08:41 |
| lucasagomes | yes | 08:42 |
| takadayuiko | then, pxe_deploy_kernel and pxe_deploy_ramdisk should be set when ironic node creation? like "ironic node-create ... -i pxe_deploy_kernel=xxx pxe_deploy_ramdisk=xxx ..." | 08:42 |
| lucasagomes | takadayuiko, yeah, or included in the node later via node-update | 08:45 |
| takadayuiko | OK, I understand | 08:45 |
| takadayuiko | thank you so much! | 08:46 |
| lucasagomes | yvw :) | 08:50 |
| takadayuiko | ;) | 08:53 |
| *** takadayuiko is now known as takadayuiko_away | 08:58 | |
| *** loki184 has joined #openstack-ironic | 09:01 | |
| ramineni | Haomeng : hi | 09:12 |
| *** yuriyz has quit IRC | 09:13 | |
| *** yuriyz has joined #openstack-ironic | 09:17 | |
| *** romcheg1 has joined #openstack-ironic | 09:28 | |
| *** romcheg has quit IRC | 09:28 | |
| *** vdrok_ has quit IRC | 09:28 | |
| *** vdrok_ has joined #openstack-ironic | 09:30 | |
| *** chenglch|2 has quit IRC | 09:31 | |
| lucasagomes | NobodyCam, devananda saw ur email about ipmi-listener | 09:36 |
| lucasagomes | haha I was doing something pretty similar to that! I was calling it BiMiC (BMC mimic, name given by derekh :P ) | 09:37 |
| lucasagomes | NobodyCam, devananda I have some code here, that does the auth part of the IPMI (like the client first send this discover package to find out what are the authentication method etc) | 09:38 |
| lucasagomes | I'm returning like no auth etc so it starts sending the ipmi commands package now | 09:38 |
| *** athomas has quit IRC | 09:42 | |
| dtantsur | brb | 09:43 |
| *** pelix has joined #openstack-ironic | 09:44 | |
| *** athomas has joined #openstack-ironic | 09:49 | |
| *** coolsvap is now known as coolsvap|afk | 09:56 | |
| vdrok_ | morning ironic! | 10:10 |
| vdrok_ | lucasagomes, around? | 10:11 |
| *** mkerrin has quit IRC | 10:29 | |
| lucasagomes | vdrok_, hi there | 10:29 |
| vdrok_ | lucasagomes, morning | 10:30 |
| lucasagomes | vdrok_, morning | 10:30 |
| vdrok_ | lucasagomes, i'm looking at api/controllers/v1/port.py | 10:30 |
| vdrok_ | lucasagomes, and sample port object there doesn't seem to contain links attribute | 10:31 |
| vdrok_ | lucasagomes, convert_with_links is not called | 10:31 |
| vdrok_ | lucasagomes, should it be so? | 10:31 |
| lucasagomes | vdrok_, hmm /me thinking | 10:34 |
| lucasagomes | I'm not sure the other resources are calling it? if so yes it should | 10:34 |
| lucasagomes | if none are, maybe worth investigating why | 10:34 |
| lucasagomes | I'm don't see why it's not being called | 10:34 |
| vdrok_ | chassis and node are calling it | 10:34 |
| vdrok_ | driver does not | 10:34 |
| lucasagomes | right, yeah seems it should then | 10:35 |
| vdrok_ | also there are 2 of them for chassis and node | 10:35 |
| vdrok_ | one with underscore in the beginning and one without | 10:35 |
| romcheg1 | Hi folks! | 10:36 |
| *** romcheg1 is now known as romcheg | 10:36 | |
| vdrok_ | hi romcheg ! | 10:36 |
| lucasagomes | vdrok_, the one prefixed with a _ seems to be a helper fucntion | 10:36 |
| lucasagomes | that the one without _ will call | 10:36 |
| vdrok_ | aha, so maybe make the same way in port? | 10:37 |
| *** mkerrin has joined #openstack-ironic | 10:37 | |
| lucasagomes | vdrok_, I see it looks like they have that helper function because it's also used for the sample | 10:39 |
| lucasagomes | so yeah it makes sense to make the same for ports | 10:39 |
| vdrok_ | lucasagomes, yup | 10:39 |
| vdrok_ | lucasagomes, thanks! | 10:39 |
| lucasagomes | np :) | 10:39 |
| *** Haomeng|2 has joined #openstack-ironic | 10:46 | |
| *** Haomeng has quit IRC | 10:47 | |
| lucasagomes | dtantsur, ping re ironic-discoverd | 10:52 |
| lucasagomes | dtantsur, r you thinking about making that service part of the ironic umbrella ? like pyghmi for e.g | 10:52 |
| lucasagomes | having it on stackforge, we may want to do that no? | 10:53 |
| *** igordcard has joined #openstack-ironic | 10:53 | |
| *** lazy_prince is now known as killer_prince | 10:56 | |
| dtantsur | lucasagomes, well, it should be somewhere in OpenStack. The question is whether it will be under TripleO umbrella (because it's made for OOO use case) or Ironic one (because it complements Ironic). | 11:00 |
| *** vigneshvar has joined #openstack-ironic | 11:00 | |
| dtantsur | it's something to discuss with devananda and others during the summit | 11:00 |
| dtantsur | I don't know if baremetal project wants something like that | 11:01 |
| vigneshvar | can anyone help on this https://ask.openstack.org/en/question/50080/ironicdriversmodulesagent-node-command-status-errored-error-downloading-image/ | 11:01 |
| vigneshvar | have issues while downloading images from swift tempurl | 11:03 |
| dtantsur | I guess jroll could, he's one to know agent the best | 11:04 |
| *** ramineni has quit IRC | 11:06 | |
| dtantsur | lucasagomes, judging by http://ci.openstack.org/stackforge.html it's not that hard to get the project to stackforge, but I'm not sure it's worth bothering, while it has exactly one contributor and one company using it :D | 11:11 |
| *** loki184 has quit IRC | 11:14 | |
| lucasagomes | dtantsur, right yeah... | 11:19 |
| lucasagomes | dtantsur, just throwing the idea, but def worth looking | 11:19 |
| *** killer_prince has quit IRC | 11:23 | |
| *** lazy_prince has joined #openstack-ironic | 11:24 | |
| *** coolsvap|afk is now known as coolsvap | 11:46 | |
| *** pensu has quit IRC | 11:46 | |
| *** mgagne has quit IRC | 11:48 | |
| *** bigjools has quit IRC | 11:50 | |
| *** mgagne has joined #openstack-ironic | 11:51 | |
| *** mgagne is now known as Guest92054 | 11:51 | |
| *** bigjools has joined #openstack-ironic | 11:51 | |
| *** bigjools has joined #openstack-ironic | 11:51 | |
| *** rakesh_hs has quit IRC | 11:58 | |
| *** bradjones has quit IRC | 11:59 | |
| *** coolsvap is now known as coolsvap|afk | 12:01 | |
| *** lazy_prince is now known as killer_prince | 12:08 | |
| *** rushiagr is now known as rushiagr_away | 12:20 | |
| *** openstackgerrit has joined #openstack-ironic | 12:24 | |
| *** bradjones has joined #openstack-ironic | 12:26 | |
| *** nosnos has quit IRC | 12:35 | |
| *** nosnos has joined #openstack-ironic | 12:36 | |
| openstackgerrit | Vinay B S proposed a change to openstack/ironic: Adds instructions for deploying instances on real hardware https://review.openstack.org/118614 | 12:36 |
| *** smoriya has quit IRC | 12:37 | |
| *** nosnos has quit IRC | 12:40 | |
| *** vinbs has quit IRC | 12:42 | |
| vdrok_ | it seems that at the moment pep8 doesn't complain at too long lines | 12:44 |
| vdrok_ | e.g. in this change https://review.openstack.org/#/c/126500/4/ironic/common/keystone.py | 12:44 |
| vdrok_ | line 98 is 80 characters long | 12:44 |
| vdrok_ | but no complaints from the gate | 12:44 |
| vdrok_ | locally it was failing, but after doing tox -r -epep8 it passes | 12:45 |
| vdrok_ | and it doesn't fail for any line length | 12:45 |
| yuriyz | the same, 150 symbols line passed pep8 test | 12:46 |
| yuriyz | maybe something broken | 12:46 |
| lucasagomes | :/ | 12:47 |
| *** rushiagr_away is now known as rushiagr | 12:49 | |
| *** pensu has joined #openstack-ironic | 12:56 | |
| lucasagomes | the pxe template contains a lot of deploy informations :( | 13:02 |
| *** killer_prince is now known as lazy_prince | 13:09 | |
| *** dprince has joined #openstack-ironic | 13:09 | |
| *** jjohnson2 has joined #openstack-ironic | 13:10 | |
| *** zz_jgrimm is now known as jgrimm | 13:13 | |
| *** vigneshvar has quit IRC | 13:19 | |
| *** rloo has joined #openstack-ironic | 13:32 | |
| NobodyCam | Good Morning Ironic | 13:41 |
| dtantsur | NobodyCam, morning | 13:42 |
| NobodyCam | morning dtantsur | 13:42 |
| rloo | morning NobodyCam. aft dtantsur | 13:48 |
| rloo | and hello to everyone else :) | 13:48 |
| NobodyCam | morning rloo, :) | 13:48 |
| yuriyz | morning NobodyCam rloo | 13:50 |
| rloo | hi yuriyz! | 13:51 |
| NobodyCam | morning yuriyz :) | 13:51 |
| *** Kui has quit IRC | 13:51 | |
| Shrews | anyone understand the purpose behind the sample() methods in our api classes? in reference to https://review.openstack.org/127210 | 13:53 |
| dtantsur | rloo, hi | 13:54 |
| rloo | Shrews, those samples are used in the generated docs. | 13:54 |
| lucasagomes | NobodyCam, rloo Shrews yuriyz morning | 13:54 |
| lucasagomes | NobodyCam, you're coding the ipmi-listener thing? | 13:54 |
| Shrews | rloo: ah ha. that's totally unobvious to me :) | 13:54 |
| rloo | Shrews: yeah. I know. | 13:55 |
| rloo | http://docs.openstack.org/developer/ironic/webapi/v1.html | 13:55 |
| rloo | Shrews: took me a few secs to even remember where. | 13:55 |
| lucasagomes | Shrews, yeah the pecan sphinx use it to generate the samples for the api docs as rloo said | 13:55 |
| lucasagomes | pecan sphinx extension | 13:55 |
| rloo | Shrews: maybe worth adding a comment then? (in another patch?) | 13:55 |
| Shrews | rloo: yeah, another patch for that i'd say | 13:56 |
| *** lazy_prince is now known as killer_prince | 13:57 | |
| rloo | Shrews: I think I recall thinking, way back when, how to NOT forget to update the samples. Obviously, we forgot... | 13:57 |
| NobodyCam | morning lucasagomes, Shrews | 13:58 |
| Shrews | morning <Insert Name Here>! | 13:59 |
| *** jgrimm is now known as zz_jgrimm | 13:59 | |
| NobodyCam | oh ... brb | 14:05 |
| *** pensu has quit IRC | 14:12 | |
| Shrews | rloo: you beat me to the punch on the bad call | 14:12 |
| rloo | Shrews: sorry. Next time I'll wait (maybe) :-) | 14:12 |
| Shrews | nah, i'm slow. waiting for me gets you nowhere fast | 14:13 |
| rloo | Shrews: well, some things we're not in a hurry for. I think we ought to give a chance to the slower folks. After all, we want to welcome everyone ;) | 14:13 |
| rloo | Shrews: I do wonder how it passed everything. I suppose we don't have any test for it. Am not sure it is worth adding though. | 14:15 |
| Shrews | rloo: it just happens to work because 'sample' also has a 'uuid' attribute. | 14:16 |
| Shrews | and 'url' becomes either True or False (the 'expand' value) | 14:17 |
| rloo | Shrews: ahh. Yeah, take a look at the generated doc: http://docs-draft.openstack.org/10/127210/1/check/gate-ironic-docs/4499e57/doc/build/html/webapi/v1.html#ports | 14:17 |
| Shrews | yep | 14:17 |
| Shrews | lucasagomes: ^^^ that's the reason there was no error in docs generation | 14:19 |
| * lucasagomes clicks | 14:21 | |
| lucasagomes | ohh | 14:21 |
| lucasagomes | damn that was tricky | 14:21 |
| Shrews | yeah, had me scratching my head for a bit | 14:22 |
| *** ryanpetrello has quit IRC | 14:23 | |
| lucasagomes | "href": "False/v1/ports/27e3153e-d5bf-4b7e-b517-fb518e17f34c", | 14:23 |
| lucasagomes | heh | 14:23 |
| lucasagomes | False/v | 14:23 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Port PXE drivers to use the new boot interface https://review.openstack.org/127213 | 14:24 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Add the base BootInterface class https://review.openstack.org/127212 | 14:24 |
| openstackgerrit | Vladyslav Drok proposed a change to openstack/ironic: Add missing attributes to sample API objects https://review.openstack.org/127210 | 14:24 |
| lucasagomes | idk if I should split the second patch in multiple ones :/ | 14:24 |
| lucasagomes | it's def doing more than one thing | 14:24 |
| vdrok_ | Shrews, rloo, lucasagomes sorry, didn't notice :) fixed now | 14:24 |
| * NobodyCam is back | 14:25 | |
| vdrok_ | and morning Shrews rloo NobodyCam ! | 14:25 |
| NobodyCam | morning vdrok_ :) | 14:25 |
| lucasagomes | NobodyCam, r u working on the ipmi listener thing? | 14:25 |
| rloo | hi vdrok_ | 14:25 |
| rloo | vdrok_: no worries. that's what the reviewers are for ;) | 14:26 |
| NobodyCam | I chatted with jjohnson2 about it | 14:26 |
| *** pensu has joined #openstack-ironic | 14:26 | |
| lucasagomes | NobodyCam, right, because me and derekh were thinking about it | 14:28 |
| lucasagomes | NobodyCam, I have some code, super messy | 14:28 |
| lucasagomes | and doesnt work yet but was near start receiving some of parameters | 14:28 |
| lucasagomes | NobodyCam, instead of call it ipmi-listener we called it BiMiC (BMC mimic) | 14:29 |
| NobodyCam | lucasagomes: I have a port listener that listens on 623 and recieves the commands | 14:29 |
| lucasagomes | NobodyCam, right but r u answering it? cause the way it works is | 14:29 |
| NobodyCam | its the decoding of the hex values that is getting me | 14:29 |
| lucasagomes | there's a "discover" package | 14:29 |
| lucasagomes | right I think the code I have does some of the authentication part | 14:29 |
| lucasagomes | because the first packages r for authentication | 14:29 |
| lucasagomes | the first one actually is a discover, it asks the BMC what are the auth method etc... | 14:30 |
| lucasagomes | afair my patch was answering always "no auth" so the client started sending the ipmi commands | 14:30 |
| lucasagomes | oh no, after the auth one there's one to initialize a session too | 14:30 |
| NobodyCam | ya | 14:30 |
| * lucasagomes doesn't remember all the order of commands it does | 14:30 | |
| lucasagomes | NobodyCam, we can work together on that if u want | 14:31 |
| NobodyCam | yea!!!! | 14:31 |
| lucasagomes | I can try organize my code a bit and put it somewhere | 14:31 |
| NobodyCam | the code I have is really just the post listener with some hard coded replys at this point | 14:31 |
| lucasagomes | right | 14:32 |
| lucasagomes | and is it for ipmi 2.0 or 1.5? | 14:32 |
| *** pensu has quit IRC | 14:32 | |
| NobodyCam | i'm happy to put is in a paste if you want to take a look but there nothing special there | 14:32 |
| NobodyCam | 1.5 I believe | 14:33 |
| lucasagomes | right... aight no worries I will try to make the code a bit more tidy and upload it somewhere | 14:33 |
| lucasagomes | cause I didn't progress much too, was kinda a few hours experiment | 14:33 |
| *** rloo has quit IRC | 14:33 | |
| NobodyCam | ya | 14:34 |
| *** rloo has joined #openstack-ironic | 14:34 | |
| lucasagomes | awesome! that would solve a couple of problems | 14:34 |
| lucasagomes | first we can test our ipmi* drivers in gate | 14:34 |
| lucasagomes | then we can dump the ssh driver and consequently that problem of having tripleo to rebuild their ci system when we change some ssh commands goes away too :) | 14:35 |
| NobodyCam | yea! and would allow us to get rid of the ssh driver all toghether | 14:35 |
| lucasagomes | yush | 14:35 |
| NobodyCam | lucasagomes: ++ | 14:36 |
| lucasagomes | NobodyCam, you may find it interesting as reference as well http://www.spinics.net/lists/kvm/msg75154.html | 14:36 |
| lucasagomes | there's a series of 9 patches adding ipmi support to qemu there | 14:37 |
| *** rloo_ has joined #openstack-ironic | 14:37 | |
| lucasagomes | tho it doesn't do it via rmcp as we are doing | 14:37 |
| lucasagomes | I mean rmcp+ | 14:38 |
| *** pensu has joined #openstack-ironic | 14:39 | |
| *** rloo has quit IRC | 14:40 | |
| NobodyCam | what I chatted with jjohnson2 about was something like adding a module to pyghmi ('pyghmi.ipmi.bmc') with the ability to set handlers for 'power_request' and so on and so forth | 14:40 |
| *** rainya has joined #openstack-ironic | 14:41 | |
| NobodyCam | that would listen on 623 and then act on the command recieved | 14:41 |
| *** k4n0 has quit IRC | 14:42 | |
| lucasagomes | hmm | 14:42 |
| lucasagomes | but pyghmi is a client lib :/ if we start adding server code it would be a bit odd no? | 14:42 |
| NobodyCam | I think the pyghmi already has most of whats needed to do it. take a look at https://github.com/stackforge/pyghmi/blob/master/pyghmi/ipmi/private/session.py | 14:44 |
| NobodyCam | but that said I am also happy to work on a parallel solution | 14:46 |
| *** zz_jgrimm is now known as jgrimm | 14:46 | |
| *** Poornima has quit IRC | 14:49 | |
| lucasagomes | right, yeah def there's some stuff we can use there | 14:49 |
| *** pcrews has joined #openstack-ironic | 14:52 | |
| dtantsur | "Release notes are available here: https://wiki.openstack.org/wiki/Ironic/ReleaseNotes/Icehouse" <-- that's in our docs introduction :) | 14:56 |
| devananda | dtantsur: Juno final release isn't out yet :) | 14:57 |
| NobodyCam | morning devananda :) | 14:58 |
| dtantsur | well right, are we going to patch released docs then? | 14:58 |
| dtantsur | devananda, and morning | 14:58 |
| devananda | lucasagomes: !! a few years back, someone tried to add ipmi to KVM. don't have the link now, but it was not accepted by kernel community ... | 14:58 |
| NobodyCam | devananda: we're cutting RC2 in an hour (or so)?? | 14:58 |
| devananda | NobodyCam: yes. unless someone points out another critical bug | 14:59 |
| lucasagomes | devananda, yeah I think that's the link I passed | 14:59 |
| lucasagomes | that's from the KVM list, tho the patch was to qemu really | 14:59 |
| devananda | NobodyCam: pyghmi talks TO a bmc... | 14:59 |
| *** killer_prince has quit IRC | 15:00 | |
| jjohnson2 | hello | 15:00 |
| NobodyCam | devananda: yes. | 15:00 |
| NobodyCam | hey jjohnson2 :) | 15:00 |
| devananda | NobodyCam: that session class is the client side | 15:00 |
| NobodyCam | devananda: yes that has some that could be useful (or the inverse of) | 15:01 |
| jjohnson2 | lucasagomes, pyghmi I would consider to be a library that is not particularly about client or server in the long haul | 15:01 |
| jroll | morning ironic :) | 15:02 |
| NobodyCam | jjohnson2: I was doing a poor job of explaining the things we chatted about the other day | 15:02 |
| jjohnson2 | the session class has a lot of common code that would be relevant to both client and server | 15:02 |
| NobodyCam | morning jroll | 15:02 |
| devananda | NobodyCam: if you are interested in implementing an ipmi-listening proxy, you should chat with jang / Jan Grant. at the tripleo midcycle, he had expressed an interest in working on it then | 15:02 |
| jjohnson2 | so for example it knows how to generate an rakp1, validate an rakp2, etc, doing the converse makes sense | 15:02 |
| lucasagomes | jjohnson2, right hmm... I think we then should split it into a pyghmi-common thing | 15:03 |
| lucasagomes | for functions that can be used on client and server | 15:03 |
| jjohnson2 | I really don't have much of a horse in the race of how it is packaged I suppose... I don't know if there is a significant enough volume of server specific code in that scenario to warrant a distinct package.. | 15:08 |
| *** jistr has quit IRC | 15:08 | |
| lucasagomes | jjohnson2, right, you want pyghmi to be something like www.coreipm.com ? | 15:11 |
| lucasagomes | but in python? | 15:11 |
| devananda | jjohnson2: there's an interest by some folks (who I'm essentially proxying for right now) to have a proxy service which looks like a BMC and understands some basic IPMI calls but emits other calls out the other side | 15:11 |
| lucasagomes | NobodyCam, btw for reference too https://code.google.com/p/coreipm/ | 15:11 |
| NobodyCam | oh nice | 15:12 |
| jjohnson2 | devananda, fwiw, I have also been approached by a couple of people about using the stack | 15:12 |
| jjohnson2 | in a real bmc | 15:12 |
| devananda | jjohnson2: why does that not surprise me :) | 15:12 |
| jjohnson2 | devananda, the real surprise is that it wasn't my own company... | 15:12 |
| jjohnson2 | anyway... | 15:13 |
| lucasagomes | hah | 15:13 |
| devananda | jjohnson2: also from the realm of crazy, someone asked me if they could just run ironic on the BMC ... | 15:13 |
| jroll | I like it, one conductor per node :D | 15:13 |
| jjohnson2 | but anyway, a base class | 15:14 |
| *** killer_prince has joined #openstack-ironic | 15:14 | |
| *** killer_prince is now known as lazy_prince | 15:14 | |
| jjohnson2 | that would have functions for things like 'do_power_request' | 15:14 |
| jjohnson2 | or whatever people think would make sense | 15:14 |
| jjohnson2 | and then if someone wanted ipmi-to-executable or ipmi-to-libvirt or ipmi-to-other, then they could have a fairly straightforward time | 15:15 |
| lucasagomes | +1 | 15:15 |
| jjohnson2 | then you could have the full authenticated fun stuff | 15:16 |
| lucasagomes | yeah cause for now, at least the code I was playing with | 15:16 |
| lucasagomes | I was just retunring no auth :P | 15:16 |
| jjohnson2 | the authentication gets a bit weird of course | 15:16 |
| jjohnson2 | I think I'd provide a simple model for now that does not replicate the ability to have multiple users with the same username | 15:17 |
| jjohnson2 | and if really requested, then a way to describe the true rabbit hole that is ipmi user account management... | 15:17 |
| *** achanda has joined #openstack-ironic | 15:17 | |
| jjohnson2 | because 99.9% of the people never ever ever want to do the stupid things, and accomodating the stupid things requires some more complex input that is really unnatural | 15:18 |
| jjohnson2 | and besides, then it has some nice self contained framework for integration testing | 15:19 |
| lucasagomes | jjohnson2, I think for what we are trying to do now, if the RCMP ping just reploy authentication type = None is great already | 15:19 |
| lucasagomes | reply* | 15:19 |
| jjohnson2 | so either way, it might be a few days before I can make significant progress, but I can probably take some time today to start a draft review | 15:21 |
| jjohnson2 | with some hints about some of the utility functions that would be warranted | 15:22 |
| lucasagomes | that would be awesome | 15:23 |
| lucasagomes | all that discovery and authentication part is a bit hairy :/ | 15:24 |
| lucasagomes | getting session, activating session, setting priv level ... | 15:24 |
| jjohnson2 | I'll probably start with a static channel auth capability set, not allowing customization | 15:24 |
| *** achanda has quit IRC | 15:24 | |
| lucasagomes | I remember I was looking into it and it's quite messy | 15:24 |
| jjohnson2 | but it is fun | 15:24 |
| NobodyCam | yes | 15:24 |
| jjohnson2 | or at least for me.. maybe I'm perverse | 15:24 |
| lucasagomes | hah | 15:24 |
| NobodyCam | well not sure I would use the word fun :-p | 15:24 |
| lucasagomes | well it was fun to start understanding how it works | 15:26 |
| jjohnson2 | but yeah, once you get through the mess of framing the packets and authenticating, it gets more straightforward | 15:26 |
| lucasagomes | but at the same time u see that the protocol is like ew | 15:26 |
| lucasagomes | yeah | 15:26 |
| jjohnson2 | I'm a bit torn on that count | 15:26 |
| jjohnson2 | on the one hand, it is a bit peculiar | 15:26 |
| jjohnson2 | on the other hand, I can hit up thousands of endpoints with very little resource utilization | 15:26 |
| jjohnson2 | and it is exceedingly specific... 0,2,0 means 'turn the hell off' | 15:27 |
| jjohnson2 | oh well | 15:27 |
| jjohnson2 | I think SNMP and IPMI are underrated in their circumstances | 15:28 |
| lucasagomes | hah well I dunno if I got what u mean on the 0,2,0 bit but yeah hah | 15:28 |
| jjohnson2 | though IPMI could have taken a nice page from SNMPv3 in how to do key localization | 15:28 |
| jjohnson2 | the one underappreciated facet of the protocols is use of shared secrets to cause mutual authentication to become automatic with good password management | 15:29 |
| lucasagomes | right, I dunno much SNMP in general, even IPMI... I have very little hands on with it | 15:29 |
| lucasagomes | when I started with Ironic I didn't even know what IPMI was :P | 15:29 |
| jjohnson2 | TLS-SRP I think would be a better fit for usually-internal stuff along the lines of IPMI and SNMPv3 | 15:30 |
| *** corvus is now known as jeblair | 15:30 | |
| jjohnson2 | (based on a career of seeing that nealy every sysadmin accepts the default cert and just always clicks 'ok' to 'what the hell is this certificate?' prompts | 15:31 |
| *** lnxnut has joined #openstack-ironic | 15:31 | |
| jjohnson2 | anyway, when I get my equipment back up, I'm going to RC my console server | 15:34 |
| jjohnson2 | if people are interested | 15:34 |
| jjohnson2 | web-enabled and cli-enabled consoles, console logging with plaintext data and binary metadata files | 15:34 |
| jjohnson2 | low memory, single process, low file descriptor usage for many consoles | 15:35 |
| openstackgerrit | A change was merged to openstack/ironic: Fix markup-related issues in documentation https://review.openstack.org/127034 | 15:35 |
| jjohnson2 | console sharing, automating buffer replay and timestamp display | 15:36 |
| *** hemna has joined #openstack-ironic | 15:37 | |
| *** dtantsur is now known as dtantsur|afk | 15:37 | |
| lucasagomes | jjohnson2, is that the confluent console server? | 15:38 |
| arif-ali | jjohnson2, do you have an updated demo at all? | 15:45 |
| openstackgerrit | A change was merged to openstack/ironic: Add missing attributes to sample API objects https://review.openstack.org/127210 | 15:55 |
| *** dlaube has joined #openstack-ironic | 16:03 | |
| *** Guest92054 is now known as mgagne | 16:03 | |
| *** mgagne is now known as mgagne_ | 16:03 | |
| *** mgagne_ is now known as Guest31178 | 16:04 | |
| *** derekh has quit IRC | 16:05 | |
| *** Guest31178 has quit IRC | 16:05 | |
| *** Guest31178 has joined #openstack-ironic | 16:05 | |
| *** todd_dsm has joined #openstack-ironic | 16:06 | |
| *** Guest31178 is now known as mgagne | 16:07 | |
| openstackgerrit | Ruby Loo proposed a change to openstack/ironic-specs: Set up for kilo specs https://review.openstack.org/127047 | 16:12 |
| *** igordcard has quit IRC | 16:16 | |
| *** ChuckC__ has joined #openstack-ironic | 16:18 | |
| JayF | Haomeng|2: GheRivero: IPA has no support for configuring raid, and in-band raid configurations have been said to be out of scope of Ironic by deva | 16:22 |
| JayF | Haomeng|2: GheRivero: I'll look at that wiki and see why it says wrong things, we certainly didn't put them there | 16:22 |
| JayF | yeah that IPA wiki page is ... very wrong it seems | 16:24 |
| *** athomas has quit IRC | 16:26 | |
| *** andreykurilin_ has joined #openstack-ironic | 16:28 | |
| *** ChuckC__ has quit IRC | 16:29 | |
| rloo_ | devananda: I'm not sure I like that you removed the content from https://wiki.openstack.org/wiki/Baremetal. Besides (for me) copy/pasting from that for ironic docn, Juno hasn't been released yet. (I know, I can use history to see what had been there.) | 16:30 |
| *** ifarkas has quit IRC | 16:31 | |
| *** ChuckC___ has joined #openstack-ironic | 16:33 | |
| *** ChuckC___ has quit IRC | 16:33 | |
| *** ChuckC___ has joined #openstack-ironic | 16:33 | |
| *** ChuckC___ has quit IRC | 16:34 | |
| *** ChuckC__ has joined #openstack-ironic | 16:35 | |
| *** ChuckC__ has quit IRC | 16:40 | |
| *** ChuckC__ has joined #openstack-ironic | 16:41 | |
| *** spandhe has joined #openstack-ironic | 16:42 | |
| *** igordcard has joined #openstack-ironic | 16:45 | |
| *** bradjones has quit IRC | 16:46 | |
| lucasagomes | aight folks I'm heading home | 16:54 |
| lucasagomes | have a good night! | 16:54 |
| *** lucasagomes has quit IRC | 16:55 | |
| *** marcoemorais has joined #openstack-ironic | 16:56 | |
| *** ChuckC__ has quit IRC | 16:57 | |
| *** praneshp has joined #openstack-ironic | 16:57 | |
| *** ChuckC__ has joined #openstack-ironic | 16:57 | |
| NobodyCam | night lucas | 17:00 |
| *** praneshp has quit IRC | 17:00 | |
| *** hyakuhei has quit IRC | 17:01 | |
| *** marcoemorais has quit IRC | 17:01 | |
| *** marcoemorais has joined #openstack-ironic | 17:02 | |
| *** marcoemorais has quit IRC | 17:02 | |
| *** marcoemorais has joined #openstack-ironic | 17:02 | |
| *** praneshp has joined #openstack-ironic | 17:05 | |
| *** marcoemorais has quit IRC | 17:06 | |
| *** marcoemorais has joined #openstack-ironic | 17:06 | |
| *** ryanpetrello has joined #openstack-ironic | 17:11 | |
| *** rloo_ has quit IRC | 17:12 | |
| *** harlowja_away is now known as harlowja | 17:19 | |
| *** ChuckC__ has quit IRC | 17:19 | |
| jjohnson2 | arif-ali, I really need to, but right now we have gobs and gobs and gobs of stuff to unpack.... | 17:22 |
| arif-ali | jjohnson2, no worries, let me know once it is out, I will like to to see how far it is gone from when I tested it a few months back | 17:24 |
| NobodyCam | is there a down side to leaving an "example.rst" file in the cycles spec folder? | 17:25 |
| NobodyCam | Re: https://review.openstack.org/#/c/127047 | 17:26 |
| *** praneshp has quit IRC | 17:32 | |
| *** achanda has joined #openstack-ironic | 17:32 | |
| *** todd_dsm has quit IRC | 17:32 | |
| *** marcoemorais has quit IRC | 17:32 | |
| *** marcoemorais has joined #openstack-ironic | 17:33 | |
| *** marcoemorais has quit IRC | 17:33 | |
| *** marcoemorais has joined #openstack-ironic | 17:33 | |
| *** marcoemorais has quit IRC | 17:33 | |
| *** praneshp has joined #openstack-ironic | 17:34 | |
| *** marcoemorais has joined #openstack-ironic | 17:35 | |
| *** jcoufal has quit IRC | 17:35 | |
| *** achanda has quit IRC | 17:37 | |
| NobodyCam | comments posted ... feel free to disagree | 17:41 |
| * NobodyCam makes more coffee | 17:43 | |
| *** marcoemorais has quit IRC | 17:45 | |
| *** marcoemorais has joined #openstack-ironic | 17:45 | |
| dlaube | hey guys, are there any public ubuntu repos/packages out there built from the latest RC for Juno? | 17:47 |
| dlaube | was just peeking at this -> https://launchpad.net/~openstack-ubuntu-testing/+archive/ubuntu/juno | 17:48 |
| dlaube | nvm, I see that ironic was updated 9 hours ago | 17:48 |
| NobodyCam | dlaube: I think we cut RC2 this morning around 9:00 am (pst) | 17:49 |
| JayF | dlaube: and you really want to run RC2 :) | 17:49 |
| JayF | dlaube: instead of RC1 :) | 17:49 |
| dlaube | :D gotcha | 17:49 |
| dlaube | any idea when the bot will run next? | 17:50 |
| NobodyCam | devananda: was a there new client release too, I don't think so | 17:50 |
| *** marcoemorais1 has joined #openstack-ironic | 17:50 | |
| *** marcoemorais has quit IRC | 17:52 | |
| *** vdrok_ has quit IRC | 17:56 | |
| JayF | NobodyCam: you think it's cool if I go ahead and land 127047? | 17:59 |
| JayF | NobodyCam: the fix for tests that ruby has up | 17:59 |
| NobodyCam | JayF: yea I +2'd it.. see my comment, but ya :) | 17:59 |
| *** lnxnut has quit IRC | 18:00 | |
| JayF | Yeah I mean | 18:00 |
| JayF | If anything, I'd say the example should go at the bottom | 18:00 |
| JayF | given you should have a real life example | 18:00 |
| JayF | as soon as any other specs are made | 18:00 |
| NobodyCam | JayF: I'm ok with top or bottom, or deleteing it, I was saying keeping an "example" in each folder is not a supper bad thing, and I would be ok with it | 18:01 |
| *** marcoemorais1 has quit IRC | 18:02 | |
| *** marcoemorais has joined #openstack-ironic | 18:03 | |
| *** andreykurilin_ has quit IRC | 18:04 | |
| *** marcoemorais has quit IRC | 18:04 | |
| *** marcoemorais has joined #openstack-ironic | 18:04 | |
| *** rainya has quit IRC | 18:05 | |
| openstackgerrit | A change was merged to openstack/ironic-specs: Set up for kilo specs https://review.openstack.org/127047 | 18:06 |
| *** derekh has joined #openstack-ironic | 18:08 | |
| *** rainya has joined #openstack-ironic | 18:08 | |
| NobodyCam | didn't we have four slots or this that just my brain off in never never land (again) | 18:14 |
| NobodyCam | I see three slots on: https://openstacksummitnovember2014paris.sched.org | 18:14 |
| *** pelix has quit IRC | 18:17 | |
| *** achanda has joined #openstack-ironic | 18:20 | |
| *** rloo has joined #openstack-ironic | 18:20 | |
| *** achanda has quit IRC | 18:20 | |
| *** marcoemorais has quit IRC | 18:22 | |
| *** achanda has joined #openstack-ironic | 18:23 | |
| *** achanda has quit IRC | 18:25 | |
| *** hemna has quit IRC | 18:34 | |
| *** achanda has joined #openstack-ironic | 18:36 | |
| *** spandhe has quit IRC | 18:38 | |
| openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Fix pep8 test https://review.openstack.org/127311 | 18:38 |
| *** spandhe has joined #openstack-ironic | 18:38 | |
| *** spandhe has quit IRC | 18:38 | |
| yuriyz | NobodyCam, devananda there is high priority bug with fix IMO https://review.openstack.org/#/c/127311/ | 18:40 |
| *** achanda has quit IRC | 18:41 | |
| NobodyCam | oh wow | 18:43 |
| *** derekh has quit IRC | 18:44 | |
| NobodyCam | yuriyz: how did you come up with that exclude list? just so I under stand | 18:44 |
| *** achanda has joined #openstack-ironic | 18:44 | |
| yuriyz | this list is current issues with hacking 0.9 | 18:45 |
| *** achanda has quit IRC | 18:45 | |
| yuriyz | after tests | 18:45 |
| yuriyz | please look at Glance for example | 18:46 |
| yuriyz | and you can try create pep8 errors with current tox.ini | 18:46 |
| yuriyz | good day I going to home :) | 18:47 |
| NobodyCam | have a good night yuriyz TY.. I may tag that bug with juno-rc-potential | 18:47 |
| *** praneshp has quit IRC | 18:48 | |
| *** marcoemorais has joined #openstack-ironic | 18:50 | |
| jroll | NobodyCam: for why, that doesn't actually affect the release | 18:50 |
| jroll | it's just pep8 | 18:50 |
| NobodyCam | I was thinking the fix to ironic/common/keystone.py in that patch may be required so we don't break other folks tests with our code | 18:52 |
| *** dprince has quit IRC | 18:52 | |
| jroll | again, it's just pep8, I assume other projects aren't running pep8 against our code | 18:52 |
| NobodyCam | :) and the line is ==80 chr which should pass | 18:53 |
| jroll | nah, 79 is the limit iirc | 18:54 |
| jroll | otherwise no reason to fix :) | 18:54 |
| *** datajerk has quit IRC | 18:56 | |
| *** praneshp has joined #openstack-ironic | 18:58 | |
| *** rushiagr is now known as rushiagr_away | 19:00 | |
| *** achanda has joined #openstack-ironic | 19:00 | |
| *** ChuckC__ has joined #openstack-ironic | 19:00 | |
| *** achanda has quit IRC | 19:02 | |
| openstackgerrit | Adam Gandelman proposed a change to openstack/ironic: Testing sideways migration (Do not merge) https://review.openstack.org/124474 | 19:03 |
| *** foexle has joined #openstack-ironic | 19:03 | |
| openstackgerrit | Adam Gandelman proposed a change to openstack/ironic: Testing sideways migration (Do not merge) https://review.openstack.org/124474 | 19:04 |
| *** datajerk has joined #openstack-ironic | 19:04 | |
| rloo | NobodyCam: I don't understand why 127311 is high priority? | 19:05 |
| jroll | rloo: because pep8 always passes right now | 19:06 |
| jroll | don't need it backported though | 19:06 |
| rloo | it's *only* pep8. | 19:06 |
| jroll | right, but more pep8 bugs will come | 19:07 |
| jroll | the longer we wait, the bigger the "fix pep8" patch will become | 19:07 |
| rloo | jroll: right, so we should fix it. sooner rather than later. maybe that's the def of high priority :-) | 19:08 |
| NobodyCam | so we need to land it but not back-port it | 19:08 |
| jroll | yes :) | 19:08 |
| rloo | yeah, no reason to backport that I can see. | 19:08 |
| jroll | I guess it's low priority, high urgency | 19:08 |
| devananda | yes. no need for a backport here | 19:08 |
| NobodyCam | :) i did say "MAY" add the tag :-p lol | 19:08 |
| *** hemna has joined #openstack-ironic | 19:09 | |
| *** jcoufal has joined #openstack-ironic | 19:10 | |
| *** ChuckC__ has quit IRC | 19:24 | |
| *** foexle has quit IRC | 19:27 | |
| *** marcoemorais has quit IRC | 19:33 | |
| *** marcoemorais1 has joined #openstack-ironic | 19:33 | |
| NobodyCam | oh really | 19:34 |
| NobodyCam | check-tempest-dsvm-ironic-pxe_ssh-postgres failed | 19:34 |
| *** ChuckC__ has joined #openstack-ironic | 19:35 | |
| rloo | so NobodyCam, were you going to approve 127311? | 19:40 |
| *** romcheg1 has joined #openstack-ironic | 19:40 | |
| NobodyCam | go distracted | 19:42 |
| NobodyCam | go = got | 19:42 |
| NobodyCam | rloo: I also need to dbl check the exclude codes... /me has not done that | 19:43 |
| rloo | NobodyCam: so I think what Yuriy did was just to exclude everything that generated warnings, except for the long line one. | 19:43 |
| rloo | NobodyCam: and I think/guess his plan is to go through them all one by one, and address them. | 19:44 |
| rloo | NobodyCam: but I'm not totally sure. I want to make sure that we get rid of all the ignores that weren't there initially. or that we decide we want to ignore. | 19:45 |
| rloo | NobodyCam: so I only +1'd cuz I wasn't sure I liked that it closes the bug. | 19:45 |
| rloo | NobodyCam: OTOH, maybe we should just get it merged, so people can start getting rid of the warnings. | 19:46 |
| rloo | NobodyCam: people == that crowd just waiting to work on it... | 19:47 |
| *** ChuckC__ has quit IRC | 19:48 | |
| devananda | rloo: lol | 19:48 |
| devananda | so the bug description seems a bit misleading | 19:48 |
| devananda | this bug only started with the upgrade from hacking 0.8.x | 19:49 |
| devananda | now, we should definitely fix it | 19:49 |
| devananda | and yuriy's patch is the first step | 19:49 |
| rloo | devananda: i was thinking of changing his commit to a partial-bug and +2'ing it | 19:50 |
| *** hemna has quit IRC | 19:51 | |
| devananda | rloo: well, that patch fixes the bug | 19:52 |
| devananda | there are other pep errors which will now fail tests validly | 19:52 |
| rloo | devananda: I know. That's my dilemma. | 19:52 |
| devananda | which ones we choose to ignore is up to us | 19:52 |
| adam_g | NobodyCam, hmph http://logs.openstack.org/11/127311/1/check/check-tempest-dsvm-ironic-pxe_ssh-postgres/5433909/logs/ironic-bm-logs/baremetalbrbm_1_console.txt.gz | 19:53 |
| devananda | NobodyCam: are you digging into the postgres job failure? | 19:53 |
| NobodyCam | would it best to land that patch and open another bug for redusing the exclude list | 19:53 |
| rloo | but if we just merge it, the commit isn't complete. it doesn't say that we are ignoring a bunch that we didn't before (if it had been passing) | 19:53 |
| devananda | oh | 19:53 |
| devananda | adam_g: that's neat | 19:53 |
| NobodyCam | devananda: time out waiting for node to become active | 19:54 |
| devananda | [ 422.148000] BUG: unable to handle kernel NULL pointer dereference at 0000000000000100 | 19:54 |
| adam_g | devananda, ya, ha | 19:54 |
| devananda | NobodyCam: time out because the VM kernel panicked | 19:54 |
| NobodyCam | hadent gotten that far yet | 19:54 |
| NobodyCam | :-p | 19:54 |
| adam_g | makes sense theres a kernel panic, i mean look at the change https://review.openstack.org/#/c/127311/1/ironic/common/keystone.py | 19:55 |
| *** andreykurilin_ has joined #openstack-ironic | 19:57 | |
| * devananda wonders why there is a case change to that string | 19:57 | |
| * NobodyCam thought that too, but then thought it is a proper name | 19:58 | |
| * rloo wonders why that change is the cause of the panic attack | 19:59 | |
| * NobodyCam thinks we're talking to much in the thrid person | 19:59 | |
| devananda | oh. consistency. | 20:01 |
| devananda | "Keystone" is in several other translated strings. "keystone" is on only one | 20:02 |
| *** dvorak has quit IRC | 20:02 | |
| rloo | the doc folks have some rule wrt when to uppercase and when not. i can never keep it straight. | 20:02 |
| SpamapS | is there an Ironic story for things like raid configuration before deployment? | 20:04 |
| *** Kui has joined #openstack-ironic | 20:05 | |
| pensu | NobodyCam: Hi...:) | 20:07 |
| NobodyCam | may be an "epic" on a cocktail napkin | 20:07 |
| NobodyCam | hi pensu | 20:08 |
| NobodyCam | :) | 20:08 |
| pensu | NobodyCam: quick question regarding devstack setup, I gave my VM public IP, so it's accessible to public network, changed the scheduler manager, still when I added a new node, my nova-compute is not able to collect its resources... | 20:09 |
| pensu | NobodyCam: What else am I missing? | 20:09 |
| NobodyCam | pensu: how long did you give it for the node to show.. | 20:10 |
| NobodyCam | and how did you add the node? | 20:10 |
| pensu | NobodyCam: Followed the example showed in the end of this: http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#exercising-the-services-locally | 20:11 |
| pensu | NobodyCam: I can see the node, I can switch on and off the IPMI using ironic, that part works fine, but somehow nova-compute still can't get the resources from the node... | 20:12 |
| *** jcoufal has quit IRC | 20:13 | |
| devananda | SpamapS: that's a hot topic which we're going to visit at the summit | 20:13 |
| SpamapS | devananda: is there an email thread or spec of something with seeded details? | 20:13 |
| devananda | SpamapS: several! but nothing condensed summarizing it yet | 20:13 |
| NobodyCam | pensu: sounds like a schulder issue | 20:14 |
| NobodyCam | pensu: what do you get from nova hypervison-stats | 20:14 |
| NobodyCam | *nova hypervisor-stats | 20:15 |
| devananda | SpamapS: I am unfortunately unlikely to make time today to write up such a summary | 20:15 |
| SpamapS | devananda: I'm sitting here manually configuring RAID on 12 blades and wanting to choke the person who decided the "press F8" timeout sould be 4 seconds on smartarrays... | 20:15 |
| devananda | ... | 20:15 |
| devananda | yea :( | 20:15 |
| * NobodyCam recall that exact same feeling | 20:16 | |
| NobodyCam | *recalls | 20:16 |
| NobodyCam | SpamapS: ilo web interface too? | 20:16 |
| pensu | NobodyCam: It shows me 4 nodes, scheduler works fine, it starts with 4 nodes, but compute can't collect resources, so scheduler rejects the option.... | 20:16 |
| devananda | SpamapS: short answer, there is not a way to do taht today via Ironic. I think there should be, but we went around in circles at the last summit and the midcycle about how to do it | 20:17 |
| jroll | pensu: compute talks to ironic to collect the resources... if it shows you nodes they should have resources | 20:17 |
| NobodyCam | yea | 20:17 |
| *** dvorak has joined #openstack-ironic | 20:17 | |
| jroll | pensu: unless, you didn't fill in node.properties when you created the node | 20:18 |
| devananda | SpamapS: the Fuel team wanted software RAID, which I strongly opposed. I believe there is a proposal for DRAC RAID config .. .lemme find it | 20:18 |
| devananda | SpamapS: https://review.openstack.org/#/c/107981 | 20:19 |
| jroll | SpamapS: we're working on making image writing pluggable in ironic-python-agent, you could make a plugin to build a raid real quick before you write the image, though not sure if that's upstreamable | 20:19 |
| SpamapS | NobodyCam: rdesktop -> win terminal server -> internet explorer -> ilo ... | 20:19 |
| devananda | jroll: "build a raid real quick" LOL | 20:19 |
| SpamapS | it's like all my nightmares from old IT in one sweep | 20:19 |
| jroll | devananda: :P | 20:20 |
| devananda | SpamapS: can you package the raid tools into an image? | 20:20 |
| SpamapS | devananda: yeah | 20:20 |
| NobodyCam | SpamapS: oh I feel your pain from here! | 20:20 |
| SpamapS | hpacucli | 20:20 |
| devananda | SpamapS: like a ramdisk element | 20:20 |
| devananda | SpamapS: then feed in the desired topology via kernel params at boot | 20:20 |
| SpamapS | It's just a C++ program that makes ioctls | 20:20 |
| devananda | SpamapS: you can totally cludge it in to ironic today, as long as you can build an image with that tool and call it from init | 20:21 |
| devananda | SpamapS: the challenges are in how we make a stable API for doing that | 20:21 |
| SpamapS | Can you have multiple deploy ramdisks? | 20:21 |
| SpamapS | that might be the simplest stable API for it. | 20:22 |
| devananda | and how that API could support both in- and out-of-band RAID config | 20:22 |
| SpamapS | Is just "put it in the ramdisk" | 20:22 |
| *** hemna has joined #openstack-ironic | 20:22 | |
| SpamapS | actually really I could just have a regular OS image that does it. | 20:22 |
| devananda | SpamapS: you, the operator, can change the ramdisk at will. it's not tied to nova flavor | 20:22 |
| SpamapS | devananda: well but I want to have a ramdisk per flavor. :) | 20:22 |
| SpamapS | hypothetically | 20:22 |
| SpamapS | If I used that method for raid configuration | 20:23 |
| pensu | jroll: I have to provide that myself? I thought ironic will gather the info from node itself... | 20:23 |
| devananda | SpamapS: you could add a different property to the node, like "preboot ramdisk", but that'd take some changes to the driver to do it | 20:23 |
| jroll | pensu: not today, no | 20:23 |
| NobodyCam | pensu: not at this time | 20:23 |
| devananda | SpamapS: or use IPA instead of the old ramdisk, and add a predeploy hook in there, though jroll could tell you more about hwo to do that | 20:24 |
| SpamapS | devananda: seems like it would be a simple addition to the API to allow specifying the ramdisk image to use (optionally). | 20:24 |
| NobodyCam | pensu: see the discovery specs I think there are a few up from last cycle | 20:24 |
| jroll | SpamapS: the ramdisk ID is in driver_info for each node these days | 20:24 |
| devananda | SpamapS: yup. "predeploy-ramdisk" and "decom-ramdisk" could be options | 20:24 |
| jroll | oh, ignore me, I see | 20:25 |
| devananda | SpamapS: though again, IPA has already implemented decom slightly differently | 20:25 |
| jroll | differently than? | 20:25 |
| devananda | jroll: AIUI you do decom directly in the agent, not by booting a separate ramdisk | 20:25 |
| SpamapS | that would work | 20:25 |
| devananda | SpamapS: however ... the problem here is that it's not going to expose any means to Nova to know (or change) what the RAID topology is | 20:26 |
| jroll | devananda: oh, I see, decom-ramdisk could just point to the agent, though :) | 20:26 |
| SpamapS | just add a piece that pivots root off /, and then messes with the disks. | 20:26 |
| devananda | I want ironic to expose more "capabilitie" metadata to Nova, and this should inform that in some way | 20:26 |
| SpamapS | devananda: well if the RAID topology is just burned into my decom ramdisk, then as long as I can select decom ramdisk by flavor extra specs.. thats enough for most environments I think. | 20:27 |
| pensu | NobodyCam: jroll: okay, will see how to do that....thanks...:) | 20:27 |
| SpamapS | anyway | 20:27 |
| SpamapS | I see how this goes around in circles :) | 20:27 |
| devananda | heh | 20:28 |
| SpamapS | nothing today.. count me as somebody interested in having it "soonish" | 20:28 |
| jroll | pensu: for example, ironic node-update $uuid add properties/cpus 4 | 20:28 |
| jroll | pensu: afaik you will need cpus, local_gb, memory_mb, and cpu_arch | 20:29 |
| devananda | SpamapS: hope you'll join our design session. I'm fairly sure this will "make the cut" and be one of the hot topics | 20:29 |
| SpamapS | devananda: yes I am going to be in most of yours I believe. :) | 20:29 |
| SpamapS | devananda: 2 hours on ilo remote console has convinced me I need moar robots | 20:30 |
| devananda | moar robots is always the right answer | 20:30 |
| NobodyCam | odd we dont have a add-a-node example | 20:31 |
| devananda | except when you run out of ammo during the robot apocalypse .... | 20:31 |
| SpamapS | except when moar beer can also solve the problem. | 20:31 |
| pensu | jroll: yeah, these things only.....let me try that out...:) | 20:31 |
| pensu | NobodyCam: Yeah, I agree! | 20:31 |
| SpamapS | suffice to say there are exceptions to that always, some fun, some not so fun | 20:31 |
| devananda | SpamapS: that too | 20:31 |
| pensu | NobodyCam: Will add it, provided I am able to add a node successfully! | 20:31 |
| rloo | pensu, NobodyCam: take a look at https://review.openstack.org/#/c/118614/ | 20:34 |
| rloo | pensu: specifically http://docs-draft.openstack.org/14/118614/7/check/gate-ironic-docs/5360cf7/doc/build/html/deploy/install-guide.html#flavor-creation | 20:34 |
| rloo | pensu: there's an example there. and yeah, we need to add something to that doc (or somewhere else) about adding nodes, etc. | 20:35 |
| rloo | we should get 118614 merged soon because it has a lot of useful info that people are asking for. | 20:36 |
| *** rainya has quit IRC | 20:38 | |
| pensu | rloo: Nice! that looks very helpful. Thanks. :) | 20:40 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Add a node-create example to quick-start https://review.openstack.org/127343 | 20:45 |
| NobodyCam | oh that might actually be better then what i just put up | 20:46 |
| rloo | NobodyCam: maybe you can quickly review 118614. You had +2'd it before and the latest revision is only better ;) | 20:50 |
| NobodyCam | rloo: will right after I do a quick walkies | 20:55 |
| rloo | thx NobodyCam | 20:55 |
| *** rainya has joined #openstack-ironic | 20:56 | |
| *** jcoufal has joined #openstack-ironic | 20:58 | |
| *** jjohnson2 has quit IRC | 21:00 | |
| *** jjohnson2 has joined #openstack-ironic | 21:04 | |
| NobodyCam | rloo: your ok with landing with the white space? | 21:07 |
| rloo | NobodyCam: yeah. this is a doc. it can be cleaned up later. We seem to be getting more questions and I think it is more important for this to go out. | 21:07 |
| NobodyCam | :) | 21:08 |
| NobodyCam | get her' done | 21:08 |
| rloo | NobodyCam: I'd like to get questions about the latest version, not an old version ;) | 21:08 |
| NobodyCam | yes | 21:08 |
| rloo | also, this way people can test the docn ;) | 21:08 |
| *** jjohnson2 has quit IRC | 21:09 | |
| *** pensu has quit IRC | 21:10 | |
| openstackgerrit | David Shrewsbury proposed a change to openstack/ironic: Enable H904 PEP8 check https://review.openstack.org/127348 | 21:12 |
| Shrews | well there's one of the pep8 checks fixed | 21:12 |
| *** andreykurilin_ has quit IRC | 21:13 | |
| *** HenryG has quit IRC | 21:14 | |
| rloo | Shrews: I thought you were slow? | 21:14 |
| NobodyCam | rloo: +a'd :) witha comment | 21:15 |
| devananda | rloo: yes, i am eager to get that landed. any word from Vinay? | 21:15 |
| rloo | thx NobodyCam! | 21:15 |
| rloo | devananda: I think NobodyCam just landed it. | 21:15 |
| devananda | ah :) | 21:15 |
| rloo | devananda: https://review.openstack.org/118614 | 21:16 |
| rloo | oh that reminds me (I think I've been ignoring it). devananda, wrt diagrams, the doc folks want SVG output | 21:17 |
| *** jgrimm is now known as zz_jgrimm | 21:17 | |
| rloo | devananda. and Jay had suggested graphviz because he thinks it is a good idea to keep the source files | 21:17 |
| rloo | but I looked into graphviz and it is good for graphs, not 'any' kind of diagram. | 21:17 |
| rloo | and Sam-I-Am (or something like that) from docs, said he likes omnigraffle, which isn't free and maybe only runs on a few platforms (not sure). | 21:18 |
| Shrews | omnigraffle is OS X only, but it is awesome | 21:18 |
| rloo | devananda: so... I'm not sure what we want to do here. And Sam-I-Am said he was supposed to come up with diagram standards, but he hasn't yet | 21:18 |
| devananda | !free !linux means some of us can't use it | 21:18 |
| openstack | devananda: Error: "free" is not a valid command. | 21:18 |
| devananda | heh | 21:19 |
| rloo | ha ha | 21:19 |
| NobodyCam | lol | 21:19 |
| rloo | So, I'm not sure what you want to do. My thinking is that the folks that want to modify the diagrams, should use whatever tool they want. | 21:19 |
| openstackgerrit | A change was merged to openstack/ironic: Add documentation for PXE UEFI setup https://review.openstack.org/127137 | 21:19 |
| rloo | For the record, I am happy NOT to modify any diagrams. (Although I happen to have omnigraffle, which I've never used.) | 21:20 |
| devananda | rloo: my concern with "use what ever they want" is it is not open to collaboration if someone else doesn't have access to the same software | 21:20 |
| NobodyCam | nice MORE docs | 21:20 |
| NobodyCam | hahahaha | 21:20 |
| devananda | rloo: I believe our doc writing need to be open to any contributor, not limited to those on a specific platform | 21:20 |
| devananda | that said | 21:21 |
| rloo | devananda: yup, I realize that. But I'm not sure what the solution is, unless someone takes the time to review all free s/w and recommend something. | 21:21 |
| devananda | as long as the _original_ files are contributed (perhaps in a different dir - that's fine) | 21:21 |
| rloo | at this point, I feel somewhat desparate to have someone volunteer to do the diagrams period :-) | 21:21 |
| devananda | then it is close | 21:21 |
| NobodyCam | only folks with amiga OS can write docs | 21:22 |
| NobodyCam | http://www.amigaos.net | 21:22 |
| NobodyCam | lol :-p | 21:22 |
| rloo | yeah, we only have one source file a .doc I think; and the rest are images in .png | 21:22 |
| rloo | NobodyCam: ++ | 21:22 |
| devananda | rloo: right. that's a bigger problem IMO | 21:22 |
| devananda | adam_g: what is going on here? https://review.openstack.org/#/c/126822 | 21:23 |
| devananda | adam_g: with the jenkins test results | 21:24 |
| devananda | why is the last vote +1 without running any of the functional tests | 21:24 |
| adam_g | devananda, b/c tempest API do not pass against stable/icehouse, we discussed a couple of weeks back | 21:24 |
| mrda | Morning ironic | 21:25 |
| NobodyCam | morning mrda :) | 21:25 |
| adam_g | devananda, trying to unjam the stable/icehouse branch since lucas and others have been trying to land a few patches there, we disabled the tempest runs against stable/icehouse | 21:25 |
| devananda | adam_g: oh. it's just reporting oddly now? | 21:27 |
| adam_g | devananda, how do you mean? i think gerrit's reporting of CI results has changed in the past week or so | 21:27 |
| mrda | Just for the record, I like omnigraffle, but since it's platform specific and non-free, we need to look elsewhere. | 21:28 |
| devananda | adam_g: ok. somehow I didn't notice until this one showed up as +1 but still failing tests | 21:28 |
| adam_g | devananda, but that patch is meant to fix the test suite in stable/icehouse, the stable/icehouse *pxe_ssh* jobs were just disabled, so its now getting a +1 from jenkins | 21:28 |
| devananda | oooh. ok. the sidebar looks like it shows >1 set of tests | 21:29 |
| devananda | adam_g: I'm with you now :) | 21:29 |
| adam_g | devananda, yeah i think the CI report at the top of the page now accumulates all the tests that have been run throughout the reviews life | 21:29 |
| adam_g | so yeah.. *confusing* | 21:29 |
| devananda | adam_g: also, gertty still shows the old failures | 21:29 |
| *** jcoufal has quit IRC | 21:29 | |
| devananda | or perhaps i need to update gertty too | 21:29 |
| devananda | heh | 21:29 |
| adam_g | speaking of which, is there a equiv of 'gertty-manage db sync' ? | 21:30 |
| adam_g | i keep getting a db schema error trying to start a fresh gertty install | 21:31 |
| adam_g | http://paste.openstack.org/show/120000/ | 21:31 |
| devananda | not afaik | 21:31 |
| devananda | i'd delete the gertty.db file | 21:32 |
| devananda | which is terrible | 21:32 |
| devananda | since it takes for ever to resync | 21:32 |
| devananda | so there's probably a way I dont know :) | 21:32 |
| adam_g | yeah.. ive deleted it, it recreates it and throws that error.. ill check -infra | 21:36 |
| *** praneshp has quit IRC | 21:37 | |
| *** praneshp has joined #openstack-ironic | 21:38 | |
| NobodyCam | wow first time I've see this in bright red: Patch in Merge Conflict | 21:38 |
| rloo | NobodyCam: my bad. I approved another patch that conflicts. Do you want to rebase 118614 or shall I? | 21:41 |
| *** praneshp has quit IRC | 21:42 | |
| NobodyCam | if your up for it i'll reapprove it | 21:42 |
| NobodyCam | :-p | 21:42 |
| rloo | NobodyCam: ok, will do it in a few somethings | 21:42 |
| openstackgerrit | A change was merged to openstack/ironic: Fix pep8 test https://review.openstack.org/127311 | 21:49 |
| openstackgerrit | Ruby Loo proposed a change to openstack/ironic: Adds instructions for deploying instances on real hardware https://review.openstack.org/118614 | 21:51 |
| rloo | NobodyCam: I'm not sure why it wasn't smart enough to do the merge but anyway. I'd wait to see what the generated doc looks like. I think I did it correctly ;) | 21:52 |
| NobodyCam | :) I'm sure it came out good... but I'll wait and see :) | 21:54 |
| rloo | mrda. hi, glad you like omnigraffle too ;) | 21:56 |
| rloo | fwiw, I added my notes about the diagram stuff to the etherpad: https://etherpad.openstack.org/p/IronicDocumentationTasks | 21:57 |
| mrda | Not that I've used it in anger, but could we use Inkscape for diagramming? | 22:01 |
| NobodyCam | Ditaa | 22:03 |
| rloo | mrda: oh, inkscape is free. the doc person I talked with said that it 'halfway decent' and that it took him longer to get a decent diagram from that vs omnigraffle. | 22:04 |
| rloo | mrda: so maybe that's what we could use. if people want to use it. | 22:04 |
| rloo | I think JayF made some negative remark about inkscape yest, but i'm not sure. I'm not volunteering him for anything though. not yet. | 22:05 |
| *** praneshp has joined #openstack-ironic | 22:07 | |
| NobodyCam | rloo: http://ditaa.sourceforge.net | 22:08 |
| *** achanda has joined #openstack-ironic | 22:08 | |
| mrda | it's not meant for diagramming, but AFAICS it's the most advanced and stable open source graphics program (well, besides the GIMP, but that's just bitmap graphics) | 22:08 |
| rloo | NobodyCam: I added that to 'the list'. Have you used it? | 22:10 |
| NobodyCam | no... Just remembered it from some of ascii art searching | 22:11 |
| rloo | NobodyCam: ha ha. | 22:11 |
| JayF | mrda: my alternative suggestion was graphviz :) | 22:17 |
| * JayF just generally dislikes using drawing programs to make diagrams | 22:17 | |
| NobodyCam | JayF: look at the Ditaa source file format :-p (it actually kinda kelw) :-p | 22:19 |
| mrda | JayF: I like graphviz for state transition diagrams etc, but for general diagramming? Not so much | 22:19 |
| JayF | I mean honestly | 22:19 |
| JayF | the best graphing program | 22:19 |
| JayF | is the one the first person just uses and makes our diagrams with | 22:20 |
| JayF | because it'll be way better than any current diagram we have | 22:20 |
| * JayF doesn't want to impose restrictions on whoever might volunteer | 22:20 | |
| * NobodyCam digs up his copy of visio and offeres it up | 22:21 | |
| NobodyCam | lol ... j/k | 22:21 |
| devananda | we should all use http://asciiflow.com/ | 22:24 |
| NobodyCam | nice | 22:24 |
| devananda | no. really. | 22:24 |
| NobodyCam | i like it | 22:24 |
| NobodyCam | this is slick | 22:24 |
| devananda | I mean, our spec repo even says, if you're going to include a diagram, it MUST be drawn using http://asciiflow.com/ | 22:25 |
| devananda | so why does our doc repo not do the same thing? | 22:25 |
| devananda | (or some other ascii diagraming method) | 22:25 |
| devananda | ok, really, our spec template says diagrams are required to be in ascii. it doesn't say which tool we draw them in. | 22:26 |
| mrda | thanks devananda for pointing this out - I've never seen this | 22:34 |
| * devananda returns to hacking on ansible modules | 22:42 | |
| * NobodyCam needs to run to the ups store ... back in a bit | 22:48 | |
| *** romcheg1 has quit IRC | 22:53 | |
| *** romcheg1 has joined #openstack-ironic | 22:53 | |
| *** romcheg1 has quit IRC | 22:54 | |
| jroll | wow, asciiflow is so cool | 22:59 |
| dlaube | indeed | 22:59 |
| * dlaube starts free-handing some funny art for /etc/motd | 23:00 | |
| * devananda looks forward to seeing all jroll's summit presentation diagrams in ascii | 23:03 | |
| jroll | devananda: dunno, I'm between that and https://www.websequencediagrams.com/# | 23:04 |
| JayF | we'll have those sequence diagrams in our decom talk for sure | 23:04 |
| devananda | jroll: yah. i've used that befoer for sequence dia, too | 23:04 |
| jroll | cool | 23:05 |
| *** HenryG has joined #openstack-ironic | 23:07 | |
| openstackgerrit | Jim Mankovich proposed a change to openstack/ironic-specs: Send iLO health metrics to ceilometer https://review.openstack.org/127378 | 23:07 |
| *** igordcard has quit IRC | 23:16 | |
| devananda | Jmank: hello and welcome! :) | 23:16 |
| Jmank | hi, thanks for the welcome, I'm looking forward to some feedback :-) | 23:17 |
| devananda | Jmank: I very quickly skimmed it -- looks great | 23:18 |
| devananda | Jmank: you'll want to get eyes from Haomeng|2 as he did the ceilometer integration for IPMI, possibly also linggao. | 23:19 |
| Jmank | thanks, I'll put together the other needed specs asap | 23:19 |
| devananda | aweeks: and I think you might be interested in ^ as well | 23:19 |
| aweeks | devananda: hey, thanks for the ping, I'll take a look | 23:19 |
| Jmank | There are three pieces to this puzzle, shall see how it pans out | 23:19 |
| Jmank | I do have a full proto of this up and running with devstack | 23:22 |
| devananda | Jmank: that's always good to hear :) | 23:23 |
| * NobodyCam is back | 23:28 | |
| jroll | devananda: wdyt about when ironic-conductor receives a SIGINT, it cleaning things up and releasing locks etc? | 23:28 |
| jroll | or anyone ^ | 23:29 |
| devananda | jroll: gentle shutdown. yes | 23:29 |
| jroll | do we have some form of that already? | 23:29 |
| JayF | jroll: honestly it should do it on a TERM as well | 23:30 |
| JayF | jroll: because that's what an init system would send it | 23:30 |
| jroll | sure | 23:30 |
| jroll | signals in general | 23:30 |
| devananda | jroll: actually should be (SIGINT,SIGTERM,SIGQUIT) | 23:31 |
| devananda | those three are the standard ones | 23:31 |
| devananda | SIGHUP should be handled differently, and probably, if we do ^ gracefully, we dont need SIGHUP as much | 23:31 |
| devananda | and we should definitely not catch SIGKILL | 23:31 |
| JayF | You can't :) | 23:32 |
| devananda | well. not in python ... | 23:32 |
| devananda | *maybe | 23:32 |
| devananda | jroll: see conductor/manager.py: del_host | 23:33 |
| *** hemna has quit IRC | 23:33 | |
| jroll | devananda: cool, ty | 23:33 |
| devananda | jroll: and ironic/openstack/common/service.py fo rthe signal handling itself | 23:34 |
| devananda | jroll: i wonder how long its been since we repulled that from oslo ... | 23:34 |
| devananda | only 2 months - go GheRivero! :) | 23:35 |
| NobodyCam | :) | 23:36 |
| *** rainya has quit IRC | 23:41 | |
| *** zz_jgrimm is now known as jgrimm | 23:43 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!