| *** r-daneel has quit IRC | 00:01 | |
| *** rloo has quit IRC | 00:02 | |
| *** mtanino has joined #openstack-ironic | 00:03 | |
| EntropyWorks | looking I found the install-guide.html first, just now found the dev-quickstart.html | 00:06 |
|---|---|---|
| *** lauramoore has quit IRC | 00:07 | |
| devananda | EntropyWorks: there's a quickstart for devstack, and one for vagrant | 00:12 |
| devananda | EntropyWorks: I recommend the devstack one if you want to try out "openstack with ironic". and as jroll says, it should "just work" | 00:13 |
| *** naohirot has joined #openstack-ironic | 00:15 | |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Copy amttool from amtterm project ** DO NOT MERGE ** https://review.openstack.org/193765 | 00:17 |
| *** Nisha_away has joined #openstack-ironic | 00:19 | |
| *** Nisha has quit IRC | 00:22 | |
| devananda | lintan__: have you seen https://bugs.launchpad.net/ironic/+bug/1454492 ? | 00:22 |
| openstack | Launchpad bug 1454492 in Ironic "openwsman library is not greenthread-aware, leads to driver lockup with both DRAC and AMT drivers" [Medium,Triaged] - Assigned to Imre Farkas (ifarkas) | 00:22 |
| EntropyWorks | devananda: ok I'll check it out. thanks | 00:23 |
| *** jamielennox|away is now known as jamielennox | 00:29 | |
| *** smoriya has joined #openstack-ironic | 00:30 | |
| jroll | devananda: turns out we are going to need some work on ironic side for this nova thing | 00:38 |
| jroll | need to be able to filter in API by "flavor" | 00:38 |
| jroll | (or alternatively iterate over all nodes and look at them, which might be fine) | 00:38 |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Remove not-currently-required amtterm dependency https://review.openstack.org/194458 | 00:39 |
| TheJulia | querying the API seems to be a little more efficent | 00:39 |
| jroll | TheJulia: assuming node.properties is indexed, yes :) | 00:39 |
| jroll | otherwise it's just python doing the same thing, pick which end it should be on | 00:40 |
| TheJulia | yeah, well python doing the same thing without transferring what could be a huge hardware list over the wire | 00:40 |
| jroll | eh, true | 00:40 |
| jroll | though we do it today already :) | 00:40 |
| jroll | so I think the latter comes first, then optimize to the former | 00:41 |
| TheJulia | true, the client is evil like that | 00:41 |
| jroll | I mean we do it in nova | 00:41 |
| jroll | but yeah | 00:41 |
| TheJulia | yeah | 00:41 |
| *** ijw_ has quit IRC | 00:54 | |
| *** bradjones has quit IRC | 00:54 | |
| *** ijw_ has joined #openstack-ironic | 00:54 | |
| *** bradjones has joined #openstack-ironic | 00:59 | |
| *** bradjones has quit IRC | 00:59 | |
| *** bradjones has joined #openstack-ironic | 00:59 | |
| *** mitchjameson has quit IRC | 01:02 | |
| *** bradjones has quit IRC | 01:07 | |
| *** bradjones has joined #openstack-ironic | 01:09 | |
| *** bradjones has quit IRC | 01:09 | |
| *** bradjones has joined #openstack-ironic | 01:09 | |
| openstackgerrit | Yuiko Takada proposed openstack/ironic-inspector: Migrate to oslo_db https://review.openstack.org/181905 | 01:19 |
| *** cing has joined #openstack-ironic | 01:21 | |
| *** alexpilotti has quit IRC | 01:40 | |
| *** spandhe has quit IRC | 01:44 | |
| *** chenglch has joined #openstack-ironic | 01:48 | |
| *** Nisha_away has quit IRC | 02:02 | |
| *** Marga_ has joined #openstack-ironic | 02:12 | |
| *** zhenguo has joined #openstack-ironic | 02:40 | |
| *** ramineni has joined #openstack-ironic | 02:46 | |
| *** yog_ has joined #openstack-ironic | 03:02 | |
| *** ijw_ has quit IRC | 03:10 | |
| *** ijw_ has joined #openstack-ironic | 03:11 | |
| *** ijw_ has quit IRC | 03:16 | |
| *** achanda has joined #openstack-ironic | 03:25 | |
| *** Marga_ has quit IRC | 03:26 | |
| *** mgarza has joined #openstack-ironic | 03:35 | |
| *** achanda has quit IRC | 03:39 | |
| *** spandhe has joined #openstack-ironic | 03:42 | |
| *** harlowja has quit IRC | 03:43 | |
| *** coolsvap|away is now known as coolsvap | 03:47 | |
| *** mgarza has quit IRC | 03:58 | |
| *** Marga_ has joined #openstack-ironic | 04:14 | |
| *** mtanino has quit IRC | 04:15 | |
| *** mitchjameson has joined #openstack-ironic | 04:18 | |
| *** rameshg871 has joined #openstack-ironic | 04:27 | |
| *** mitchjameson has quit IRC | 04:28 | |
| *** Marga_ has quit IRC | 04:31 | |
| *** Marga_ has joined #openstack-ironic | 04:36 | |
| *** Marga_ has quit IRC | 04:38 | |
| *** mgarza_ has joined #openstack-ironic | 04:38 | |
| *** Marga_ has joined #openstack-ironic | 04:38 | |
| *** rameshg871 has quit IRC | 04:39 | |
| *** mgarza__ has joined #openstack-ironic | 04:41 | |
| *** mgarza_ has quit IRC | 04:43 | |
| *** rameshg87 has joined #openstack-ironic | 04:43 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic-specs: Make ilo drivers standalone work without swift https://review.openstack.org/193478 | 04:45 |
| *** mitchjameson has joined #openstack-ironic | 04:45 | |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: Add RIS support for firmware update https://review.openstack.org/194491 | 04:52 |
| devananda | jroll: query API ++ | 04:54 |
| jroll | devananda: baby steps :) | 04:55 |
| jroll | but yeah, needed | 04:55 |
| * jroll goes to meeting room | 04:58 | |
| *** rameshg87 has quit IRC | 05:00 | |
| devananda | mmm meeting | 05:00 |
| *** martini has joined #openstack-ironic | 05:01 | |
| *** rameshg87 has joined #openstack-ironic | 05:02 | |
| *** Nisha has joined #openstack-ironic | 05:03 | |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: Add RIS support for firmware update https://review.openstack.org/193952 | 05:04 |
| *** krtaylor has quit IRC | 05:05 | |
| *** boris-42 has quit IRC | 05:12 | |
| *** Marga_ has quit IRC | 05:17 | |
| openstackgerrit | Merged openstack/ironic: Add option to configure passes in erase_devices https://review.openstack.org/191750 | 05:24 |
| *** krtaylor has joined #openstack-ironic | 05:25 | |
| *** korean101 has joined #openstack-ironic | 05:27 | |
| korean101 | HI ironic! | 05:28 |
| Haomeng|2 | korean101: we are in meeting room, welcome | 05:29 |
| korean101 | oh, i see! | 05:30 |
| Haomeng|2 | korean101: so will be back to you here once meeting end after 30mins | 05:30 |
| korean101 | ok. got it~ | 05:31 |
| *** korean101 has quit IRC | 05:31 | |
| *** stendulker has joined #openstack-ironic | 05:56 | |
| jroll | naohirot: whatcha got? | 05:58 |
| naohirot | jroll: Oh yes | 06:01 |
| naohirot | devananda: jroll: I'd like to discuss liberty-1 release and irmc vmedia | 06:01 |
| jroll | I keep hearing liberty-1 in this channel | 06:01 |
| naohirot | devananda: since vmedia was bumped from kilo-3 | 06:01 |
| jroll | I prefer to think liberty-1 doesn't exist | 06:02 |
| *** radek__ has joined #openstack-ironic | 06:02 | |
| jroll | but that's an aside, go on | 06:02 |
| naohirot | jroll: devananda : I expected irmc vmedia becomes a part of liberty-1 | 06:02 |
| naohirot | jroll: devananda : As I wrote in the status report, I believe review process has been iterated enough to reach acceptable level of quality. | 06:03 |
| devananda | naohirot: I do not control the reviewers or when they approve patches | 06:03 |
| jroll | naohirot: if we've reached an acceptable level of quality, why is there no +1 or +2 votes on the patch? | 06:03 |
| naohirot | devananda: Maybe I still didn't get how oss deveopment is proceeding | 06:04 |
| naohirot | jroll: I think the last -1 dmitry voted is not valid, so I asked him to explain and clarify his point. | 06:05 |
| naohirot | jroll: but so far I didn't get his answer or reply | 06:06 |
| devananda | naohirot: demanding that folks review your patch(es) does not go very far. we all must agree with and help each other to move forward, so, community participation in reaching common goals is much better and generally results in others wanting to help you, too. | 06:06 |
| devananda | naohirot: also, most of the core team is *VERY* busy | 06:06 |
| *** radek__ has quit IRC | 06:06 | |
| naohirot | devananda: If I want to merge certain feature in to the master, should I ask each core to review each time? | 06:06 |
| jroll | naohirot: sure. that's fair. dmitry is a busy person. | 06:06 |
| *** radek__ has joined #openstack-ironic | 06:07 | |
| devananda | naohirot: if a core does not seem to have time to review your patch, perhaps you can find out what you can do to help them -- so that they have more time and can review your patch | 06:07 |
| naohirot | devananda: The common goal is important, I agree 100% | 06:07 |
| devananda | naohirot: that is what we all do | 06:08 |
| *** martini has quit IRC | 06:08 | |
| jroll | naohirot: we try to move fast, but we can't always do so. especially with non-essential code like a driver. lots of important work to do that affects everything rather than just a certain type of hardware. | 06:08 |
| mrda | Thanks for staying up peoples for the meeting | 06:09 |
| naohirot | devananda: The common goal which I understand is that to provide a product which enable customer free from vendor locking | 06:09 |
| jroll | mrda: thanks for being one of the actually awake people :) | 06:09 |
| jroll | naohirot: as an aside, you keep saying customer. that's a very confusing word -- let's use deployer, operator, and user if we could. | 06:10 |
| jroll | all of those could be considered customers :) | 06:10 |
| *** ukalifon1 has joined #openstack-ironic | 06:10 | |
| * jroll gets back on topic | 06:10 | |
| naohirot | devananda: In your PTL Candidacy mail, http://lists.openstack.org/pipermail/openstack-dev/2015-April/060783.html | 06:10 |
| devananda | naohirot: your patch is also 2k LOC -- that's a lot to review! | 06:11 |
| devananda | naohirot: generally speaking, smaller patches are easier to review & land ... | 06:12 |
| naohirot | devananda: you said that "I'd like to see more consistency in feature coverage across hardware | 06:12 |
| naohirot | drivers, as well as better tracking and communication about each drivers' | 06:12 |
| naohirot | capabilities." | 06:12 |
| devananda | yup, I did say that | 06:12 |
| devananda | naohirot: there are several inconsistencies between drivers that are already in tree today | 06:12 |
| devananda | naohirot: adding another driver doesn't reduce that -- it makes the problem worse, and increases the burden on the core team. while I want more drivers in tree, and support for more hardware, that needs to be balanced with the workload on the current team | 06:13 |
| naohirot | devananda: but the prirority you assigned to irmc vmedia is low, I'd like to ask you really its reason. | 06:13 |
| devananda | naohirot: more people getting deeply involved with the project would help its velocity | 06:14 |
| devananda | naohirot: more people contributing drivers and asking us to land them decreases our velocity | 06:14 |
| devananda | naohirot: because it only helps a small percentage of our user base | 06:14 |
| devananda | naohirot: and it is not addressing the problems which our current user base has | 06:14 |
| devananda | naohirot: you may note that in previous cycles, I also assigned Low priority to other vendor's drivers | 06:15 |
| naohirot | naohirot: Yes, but I didn't know the reason. | 06:16 |
| *** mgarza__ has quit IRC | 06:16 | |
| devananda | naohirot: I would like to knokw why it is so important to you to land this driver in liberty-1 | 06:16 |
| jroll | another thought slightly derailing the current thread: I'd also like to point out that responding to reviewers with the equivalent of "that's your opinion" and ignoring the feedback isn't helpful for getting code landed either. those same reviewers need to maintain your code for the foreseeable future, readability is important. | 06:17 |
| naohirot | devananda: If there is certain reasonable reason, it's fine. I can corporate that | 06:17 |
| devananda | jroll: whoa - where's that? | 06:17 |
| jroll | see also line 438/431 https://review.openstack.org/#/c/151958/24..25/ironic/drivers/modules/irmc/deploy.py | 06:17 |
| naohirot | devananda: that's because I worked for a company. | 06:18 |
| naohirot | devananda: as well as Ironic community. | 06:18 |
| jroll | devananda: and the related https://github.com/NaohiroTamura/python-scciclient/blob/master/scciclient/irmc/scci.py#L284 | 06:18 |
| naohirot | devananda: as you know, company has its own plan. | 06:19 |
| jroll | naohirot: so why is it important to land this in liberty-1? | 06:19 |
| devananda | naohirot: what does that have to do with liberty-1? | 06:19 |
| naohirot | devananda: because originally company planned it lands in Kilo-3. | 06:20 |
| devananda | naohirot: a company can't determine when code lands in an opensource community project ............. | 06:20 |
| naohirot | devananda: company is trying to catch up the delay | 06:20 |
| jroll | delay? | 06:21 |
| naohirot | devananda: and that plan is out of control of me, as you know. | 06:21 |
| jroll | naohirot: we can't fix the fact that your company is making broken assumptions | 06:22 |
| jroll | naohirot: and we won't change something about the project to fix your company's broken assumptions | 06:22 |
| *** saripurigopi has joined #openstack-ironic | 06:23 | |
| naohirot | jroll: that's fine, if there is a reasonable reason. I'll report the reason. | 06:23 |
| saripurigopi | Hello Ironic.. | 06:23 |
| jroll | naohirot: is "no corporation controls this project" a reasonable reason? | 06:24 |
| *** spandhe has quit IRC | 06:25 | |
| naohirot | jroll: this is oss project, no corporation controls. but how can you answer why irmc vmedia hasn't merged even if ptl said that | 06:25 |
| naohirot | jroll: drivers features should be equal. | 06:26 |
| jroll | naohirot: the real answer is: because the code isn't ready to be merged | 06:26 |
| jroll | naohirot: and that is partially because there haven't been enough reviews, partially because the code is not ready. | 06:26 |
| jroll | naohirot: the primary metric for a merge is code quality, not "driver features should be equal" | 06:27 |
| naohirot | jroll: As I reported, we iterated enough review, and I incorporated most of them. | 06:27 |
| devananda | naohirot: because the community of developers whom I trust -- and who trust each other -- to review and land code in this project has not decided that it is sufficiently "ready" to be merged and maintained by them | 06:27 |
| devananda | naohirot: you, as the author of a patch, do not get to decide when it is "enough reviews" | 06:28 |
| jroll | naohirot: you don't decide what is ... that | 06:28 |
| devananda | naohirot: the core review team decides that. | 06:28 |
| naohirot | jroll: why some vendor feature is merged quickly, but others aren't. | 06:28 |
| devananda | naohirot: they review my work too, and make the same judgement call -- and often enough, they -1 my code because they dont feel it's ready either | 06:28 |
| jroll | naohirot: my team came to ironic with a new driver. it took 100-150 patch sets before merging. | 06:28 |
| jroll | naohirot: I feel your pain. it isn't easy. it isn't quick. | 06:29 |
| devananda | naohirot: the IPA driver took more than six months to merge, if memory serves | 06:29 |
| devananda | naohirot: HP's iLO driver took about that long as well | 06:29 |
| jroll | ~5 months, yeah | 06:29 |
| naohirot | jroll: can you answer it? as you meitioned in the gerrit, if some vendor could sneak in, it's absolutely unfair. | 06:29 |
| jroll | naohirot: answer what? | 06:29 |
| naohirot | jroll: answer for why some vendor feature is merged quickly, but others aren't. | 06:30 |
| jroll | naohirot: new deploy drivers never merge quickly | 06:31 |
| jroll | naohirot: some vendors write good code the first time and it only needs a few iterations | 06:31 |
| naohirot | jroll: that's the reason I proposed a standard rule. | 06:31 |
| jroll | naohirot: rule? what? | 06:31 |
| devananda | naohirot: "standard rule"??? What?? | 06:31 |
| jroll | I missed this proposal. | 06:32 |
| devananda | so did I | 06:32 |
| *** wanghua has joined #openstack-ironic | 06:32 | |
| naohirot | devananda: jroll: https://review.openstack.org/#/c/185171/5/specs/liberty/feature-based-releases.rst | 06:32 |
| devananda | oh | 06:32 |
| devananda | naohirot: https://wiki.openstack.org/wiki/Open | 06:33 |
| naohirot | devananda: why don't we put a sentence which I proposed | 06:33 |
| devananda | naohirot: see my reponse yet? | 06:33 |
| naohirot | """ Ironic team holds a meeting which invites openly all contributors including cores and non cores to decide what features make a release. """ | 06:33 |
| devananda | naohirot: there are several problems with that sentence | 06:34 |
| naohirot | devananda: sorry I havent't | 06:34 |
| *** wanghua has left #openstack-ironic | 06:34 | |
| naohirot | devananda: the last comment I have read is Jim's. | 06:34 |
| devananda | 1. we already have meetings where discussion and design occurs. this is on IRC, at the summits, and at the mid cycles | 06:35 |
| devananda | 2. core reviewers are over burdened with the review queue and writing the bulk of the code for the "core" of the project. non-cores do not get to dictate what cores do -- in fact, none of us dictate what anyone else does at all! | 06:35 |
| *** dtantsur|afk is now known as dtantsur | 06:36 | |
| dtantsur | Morning Ironic! | 06:36 |
| devananda | 3. we can never decide, in advance, what feature will be done in what release. we can GUESS about it, but IT IS ONLY A GUESS | 06:36 |
| jroll | morning dtantsur :) | 06:37 |
| jroll | so besides why this proposal doesn't work -- how does this proposal help a driver to merge? | 06:37 |
| naohirot | devananda: So as I mentioned in the Ironic Official IRC meeting, so all I can do is just wait for a good luck? | 06:38 |
| devananda | naohirot: no - you can participate in the project | 06:38 |
| devananda | naohirot: by helping others, they in turn have more time, and are more likely, to help you | 06:38 |
| devananda | naohirot: beyond that -- yes. be patient and keep working on it | 06:39 |
| jroll | naohirot: I will say, ignoring reviewer feedback is not a good way to get more reviews | 06:39 |
| devananda | true that ^ | 06:40 |
| naohirot | devananda: Yes, I'll help others. but I believe that is not logically connected. | 06:40 |
| *** ndipanov has joined #openstack-ironic | 06:40 | |
| devananda | naohirot: believe it or not -- it is | 06:40 |
| devananda | naohirot: openstack is a socially interconnected meritocracy | 06:40 |
| devananda | naohirot: contributions are judged on their merits, but the project is driven by humans, and humans are social creatures | 06:41 |
| rameshg87 | naohirot: honestly i am in same position as you | 06:42 |
| rameshg87 | naohirot: i am payed for working on vendor features and it takes 2 releases to merge my complete features most of the time :) | 06:42 |
| naohirot | jroll: I believe you did "ignoring reviewer feedback" to me, sorry for nit picking but I'd like to say to same thing. because core and non core in the same situation. | 06:42 |
| rameshg87 | but from the community-wise, we are making lesser impact because our features are vendor specific | 06:43 |
| rameshg87 | ilo virtual media driver went through several iterations, was first proposed in mid of juno | 06:43 |
| rameshg87 | it finally ended up merging on last day of kilo feature freeze (iirc) | 06:44 |
| rameshg87 | naohirot: one option that you can try out (i don't know how to make all ends work) | 06:44 |
| jroll | naohirot: I thought we discussed my spec and you understand why that happened. | 06:44 |
| rameshg87 | naohirot: is that you can keep your driver out of the tree working version for whomever you want to deliver | 06:45 |
| rameshg87 | naohirot: and try to merge it in tree with reviews upstream. and finally get it merged | 06:45 |
| jroll | rameshg87: +1 | 06:45 |
| jroll | we launched our production environment with the agent driver out of tree. | 06:45 |
| * rameshg87 messed up on release names it was mid of icehouse and ended up merging on end of juno :) | 06:45 | |
| naohirot | rameshg87: As dtantsur said that openstack and stackforge is quite different from customer's point of view | 06:46 |
| rameshg87 | and trust me doing reviews and helping others does work ;-) | 06:46 |
| devananda | rameshg87: I was gonna say, I thought it was icehouse ... | 06:46 |
| rameshg87 | devananda: just figured it out, ramineni said in private chat :) | 06:47 |
| naohirot | rameshg87: Driver inside upstream and downstream is quite different from customer's point of view | 06:47 |
| jroll | naohirot: stop saying customers. nobody knows what that means. deployers, operators, users. | 06:47 |
| *** ndipanov has quit IRC | 06:47 | |
| naohirot | rameshg87: that's reality | 06:47 |
| rameshg87 | naohirot: i agree. code in openstack/ironic is more widely reviewed and scrutinized. but still stackforge is not *so* bad :) | 06:48 |
| devananda | naohirot: why do fujitsu customers care if your driver has landed in openstack or not? | 06:49 |
| rameshg87 | devananda: some do care. trust me ;-) | 06:49 |
| * rameshg87 knows it | 06:49 | |
| naohirot | jroll: customer I meant in this case is decision maker who don't have technical knowledge | 06:50 |
| rameshg87 | but i believe it's up to we as citizens of community to educate others | 06:50 |
| jroll | naohirot: "decision maker" means nothing to me. what do these people do with ironic? | 06:50 |
| jroll | naohirot: install it, run it, use it? | 06:50 |
| naohirot | devananda: that's brand. openstack brand is important. | 06:51 |
| naohirot | jroll: "decision maker" doesn't care about Ironic | 06:52 |
| naohirot | jroll: they purchase a system | 06:52 |
| jroll | naohirot: so why does "decision maker" care if a driver is in the openstack namespace or not | 06:53 |
| naohirot | jroll: a system includes all hardware and software as a whole | 06:53 |
| devananda | naohirot: if they are buying software from fujitsu, what prevents you from including an out-of-tree driver, or a driver from stackforge? | 06:53 |
| naohirot | jroll: they compare HP openstack, rackspace openstack, fujitsu openstack by a catalog | 06:54 |
| naohirot | jroll: comparison table has significant impact for them. | 06:54 |
| naohirot | jroll: so we need to be compete fairly | 06:55 |
| jroll | naohirot: rackspace doesn't sell hardware so I hope that you're just using the name as an example :) | 06:55 |
| * rameshg87 notes jroll removed from competition ;-) | 06:55 | |
| naohirot | jroll: that is just an example :) | 06:55 |
| jroll | rameshg87: you have a public cloud, don't remove me :) | 06:56 |
| rameshg87 | lol :) | 06:56 |
| jroll | naohirot: ok. so. ironic, the project, doesn't care *at all* about whether a given vendor can check a box to make a sale. that is not what we're here for. | 06:56 |
| jroll | naohirot: we're here to promote open hardware provisioning, common apis, etc. | 06:56 |
| *** lazy_prince has joined #openstack-ironic | 06:57 | |
| naohirot | jroll: yes, that's the costumer wants I think. | 06:57 |
| jroll | naohirot: we are not your lever for making money. we are your lever for fixing the hardware industry which is made up of a bunch of proprietary bullshit and vendor lock-in. we want to make hardware open and secure. | 06:57 |
| *** Nisha_away has joined #openstack-ironic | 06:58 | |
| persia | naohirot: Change the conversation with the "customer": the Ironic project *can't* care about commercials (it breaks the social model), but the customers who use a checksheet usually end up with a project that requires custom bespoke solutions or failure. You can probably source some statistics on that, if it helps you. Operators that can consume upstream are unlikely to care (because they can do the maintenance), and operators who cannot | 06:58 |
| persia | consume upstream cannot usefully contribute so should not be concerned with the source of the code. | 06:58 |
| naohirot | jroll: that's the way we should go | 06:58 |
| jroll | naohirot: what is the way who should go | 06:59 |
| naohirot | persia: hi | 06:59 |
| naohirot | persia: Yeah, I think the situation is different among regions | 07:00 |
| naohirot | persia: In Japan, everything IT related things are managed by IT vendor. customer doesn't do themselves. | 07:02 |
| *** Nisha has quit IRC | 07:02 | |
| naohirot | persia: but in U.S., I believe that customer does most of IT things themselves, generally right? | 07:03 |
| devananda | naohirot: not necessarily. there are a lot of companies in the US who buy IT services from other companies | 07:04 |
| dtantsur | ++ | 07:04 |
| jroll | rackspace exists for that reason | 07:04 |
| jroll | so does dtantsur^Wred hat | 07:05 |
| dtantsur | we too, to some extend :) | 07:05 |
| dtantsur | right :) | 07:05 |
| *** rameshg871 has joined #openstack-ironic | 07:05 | |
| *** rameshg87 has quit IRC | 07:06 | |
| dtantsur | re timing: it took me 2 cycles to land inspection :) | 07:06 |
| dtantsur | and it did require me to tamper with stackforge, now I have very painful renaming | 07:06 |
| jroll | so I think I need to go to bed, it seems this conversation is waning | 07:06 |
| * rameshg871 wonders if 2 cycles can be considered as a minimum requirement ;-) | 07:06 | |
| dtantsur | it's not always smooth for us, despite being cores | 07:06 |
| devananda | on that note, I need to sleep. have a meeting in 6hrs | 07:07 |
| *** rameshg871 is now known as rameshg87 | 07:07 | |
| dtantsur | devananda, jroll, g'night! | 07:07 |
| jroll | devananda: lucky me, mine is 7 hours away! | 07:07 |
| rameshg87 | good night folks ... | 07:07 |
| jroll | night rameshg87 dtantsur devananda naohirot etc etc | 07:08 |
| *** chenglch|2 has joined #openstack-ironic | 07:08 | |
| *** chenglch has quit IRC | 07:08 | |
| *** yuanying_ has joined #openstack-ironic | 07:08 | |
| naohirot | jroll: devananda: good night, thanks for sparing your time for me | 07:09 |
| *** yuanying_ has quit IRC | 07:10 | |
| *** korean101 has joined #openstack-ironic | 07:10 | |
| *** yuanying has quit IRC | 07:12 | |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: When boot option is not persisted, set boot on next power on https://review.openstack.org/177642 | 07:20 |
| *** romainh has joined #openstack-ironic | 07:24 | |
| *** boris-42 has joined #openstack-ironic | 07:32 | |
| *** lucasagomes has joined #openstack-ironic | 07:32 | |
| *** ifarkas has joined #openstack-ironic | 07:36 | |
| *** yuanying has joined #openstack-ironic | 07:38 | |
| openstackgerrit | Merged openstack/python-ironic-inspector-client: Updated from global requirements https://review.openstack.org/194421 | 07:41 |
| openstackgerrit | Merged openstack/ironic-inspector: Updated from global requirements https://review.openstack.org/194347 | 07:45 |
| rameshg87 | dtantsur: hi | 07:45 |
| dtantsur | rameshg87, o/ | 07:45 |
| rameshg87 | dtantsur: how does https://review.openstack.org/#/c/192254/1 help ? | 07:45 |
| dtantsur | rameshg87, it allows `heat stack-delete` with broken nodes (by allowing nova to dissociate instances) | 07:46 |
| dtantsur | (at least I assume it) | 07:46 |
| rameshg87 | dtantsur: but if the node is to be retained, internally the problem has to be fixed and teared down, right ? | 07:47 |
| rameshg87 | dtantsur: so even if nova says delete fail | 07:47 |
| rameshg87 | dtantsur: isn't better for the operator to fix the actual problem (w.r.t delete) and then try asking user to delete it again | 07:47 |
| *** romcheg has joined #openstack-ironic | 07:47 | |
| rameshg87 | dtantsur: because on a critical problem, if the node is to be retained in ironic environment, it has to be teared down before it can be reused again, right ? | 07:48 |
| rameshg87 | dtantsur: will be back after lunch | 07:50 |
| *** rameshg87 is now known as rameshg87-lunch | 07:50 | |
| dtantsur | rameshg87-lunch, so you suggest an operator just delete a node with instance? because that's the only thing that is possible with a deeply broken node | 07:51 |
| dtantsur | i.e. you can't tear it down, you even can't bring it our of maintenance | 07:52 |
| *** persia has quit IRC | 07:53 | |
| *** persia has joined #openstack-ironic | 07:53 | |
| *** persia has quit IRC | 07:53 | |
| *** persia has joined #openstack-ironic | 07:53 | |
| *** romcheg has quit IRC | 07:59 | |
| *** dtantsur is now known as dtantsur|brb | 08:01 | |
| lucasagomes | is gerrit really slow today!? | 08:03 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Add iBoot driver documentation https://review.openstack.org/193546 | 08:10 |
| *** yuanying has quit IRC | 08:13 | |
| openstackgerrit | Merged openstack/ironic: Switch from MySQL-python to PyMySQL https://review.openstack.org/184378 | 08:16 |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Do not preserve partitions after ephemeral https://review.openstack.org/182688 | 08:17 |
| *** Marga_ has joined #openstack-ironic | 08:28 | |
| *** Marga_ has quit IRC | 08:28 | |
| openstackgerrit | Merged openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/194023 | 08:28 |
| *** Marga_ has joined #openstack-ironic | 08:29 | |
| *** rameshg87-lunch is now known as rameshg87 | 08:31 | |
| *** e0ne has joined #openstack-ironic | 08:32 | |
| rameshg87 | dtantsur|brb: I didn't actually suggest that. I suggested shouldn't they try fixing why it can't be deleted after putting node into maintenance mode ? | 08:32 |
| *** e0ne is now known as e0ne_ | 08:38 | |
| *** athomas has joined #openstack-ironic | 08:39 | |
| *** e0ne_ has quit IRC | 08:44 | |
| *** e0ne has joined #openstack-ironic | 08:48 | |
| *** dguerri` is now known as dguerri | 08:48 | |
| *** pelix has joined #openstack-ironic | 08:51 | |
| *** dtantsur|brb is now known as dtantsur | 08:51 | |
| dtantsur | rameshg87, so the thing is: node BMC is broken, not responding correctly | 08:52 |
| dtantsur | and it happened in the middle of deploy | 08:52 |
| dtantsur | which means: node can no longer get out of maintenance | 08:52 |
| dtantsur | but on the other hand, it's still associated with an instance | 08:52 |
| dtantsur | what to do? | 08:52 |
| rameshg87 | dtantsur: hmm okay, so we just want to delete the instance and get on with it ? | 08:53 |
| rameshg87 | got it .. | 08:53 |
| dtantsur | rameshg87, yep. to delete nova instance, we need to dissociate it (at least that's what our driver does) | 08:54 |
| dtantsur | otherwise deletion fails | 08:54 |
| rameshg87 | okay | 08:54 |
| *** mitchjameson has quit IRC | 08:54 | |
| openstackgerrit | Merged openstack/ironic: API: Get a subset of fields from Nodes https://review.openstack.org/183491 | 08:59 |
| *** e0ne is now known as e0ne_ | 09:06 | |
| *** dtantsur is now known as dtantsur|brb | 09:07 | |
| *** e0ne_ has quit IRC | 09:11 | |
| *** MattMan has quit IRC | 09:11 | |
| *** e0ne has joined #openstack-ironic | 09:15 | |
| *** athomas has quit IRC | 09:22 | |
| *** jcoufal has joined #openstack-ironic | 09:23 | |
| *** MattMan has joined #openstack-ironic | 09:23 | |
| *** marios_ is now known as marios | 09:26 | |
| *** athomas has joined #openstack-ironic | 09:30 | |
| *** alexpilotti has joined #openstack-ironic | 09:36 | |
| openstackgerrit | Shivanand Tendulker proposed stackforge/proliantutils: Adding RIS support for virtual media devices https://review.openstack.org/194570 | 09:37 |
| sambetts | Morning all o/ | 09:41 |
| *** rsalevsky has joined #openstack-ironic | 09:42 | |
| lucasagomes | sambetts, morning | 09:44 |
| *** marzif has joined #openstack-ironic | 09:48 | |
| *** oomichi has joined #openstack-ironic | 09:51 | |
| saripurigopi | hello sambetts | 09:54 |
| *** chenglch|2 has quit IRC | 09:54 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Clear ilo_boot_iso before deploy for glance images https://review.openstack.org/193436 | 09:54 |
| *** oomichi has quit IRC | 09:54 | |
| *** chenglch|2 has joined #openstack-ironic | 09:54 | |
| *** viktors has joined #openstack-ironic | 09:55 | |
| *** naohirot has quit IRC | 09:56 | |
| *** oomichi has joined #openstack-ironic | 09:57 | |
| *** rsalevsky has left #openstack-ironic | 10:05 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Refactor method to add or update capability string https://review.openstack.org/162451 | 10:07 |
| *** ndipanov has joined #openstack-ironic | 10:07 | |
| *** MattMan has quit IRC | 10:09 | |
| *** e0ne is now known as e0ne_ | 10:10 | |
| *** romcheg has joined #openstack-ironic | 10:14 | |
| *** Nisha_away has quit IRC | 10:19 | |
| *** cing has quit IRC | 10:20 | |
| *** e0ne_ has quit IRC | 10:20 | |
| *** cing has joined #openstack-ironic | 10:20 | |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: Common Changes required for firmware update https://review.openstack.org/194589 | 10:28 |
| *** dtantsur|brb is now known as dtantsur | 10:29 | |
| dtantsur | sambetts, morning! | 10:29 |
| sambetts | Morning lucasagomes, saripurigopi, dtantsur | 10:30 |
| *** oomichi has quit IRC | 10:31 | |
| TheJulia | good morning | 10:32 |
| sambetts | Morning TheJulia o/ | 10:32 |
| dtantsur | lucasagomes, could you also consider https://review.openstack.org/#/c/191736 please? You've approved it's child patch | 10:36 |
| dtantsur | TheJulia, morning! | 10:36 |
| openstackgerrit | Merged openstack/ironic-inspector: Move client out of tree https://review.openstack.org/193150 | 10:36 |
| openstackgerrit | Anusha Ramineni proposed stackforge/proliantutils: Common Changes required for firmware update https://review.openstack.org/194589 | 10:36 |
| lucasagomes | dtantsur, sure, 1 sec... finishing up debugging something | 10:37 |
| dtantsur | ack | 10:37 |
| *** chenglch|2 has quit IRC | 10:39 | |
| lucasagomes | iptables rules *sigh* | 10:42 |
| TheJulia | good morning | 10:46 |
| saripurigopi | Hi TheJulia | 10:47 |
| * TheJulia disappears for 2 hours | 10:53 | |
| *** coolsvap is now known as coolsvap|away | 10:55 | |
| lucasagomes | https://review.openstack.org/#/c/194598 | 10:56 |
| lucasagomes | devstack stuff, if someone is interested in taking a look :-) ^ | 10:56 |
| lucasagomes | morning saripurigopi TheJulia | 10:56 |
| *** ramineni has quit IRC | 10:56 | |
| saripurigopi | morning lucasagomes | 10:57 |
| lucasagomes | dtantsur, that sounds like we are using maintenance as a mean to --force delete | 10:59 |
| *** jamielennox is now known as jamielennox|away | 11:00 | |
| *** e0ne has joined #openstack-ironic | 11:09 | |
| dtantsur | lucasagomes, --force everything :) | 11:09 |
| dtantsur | yeah, kind of | 11:09 |
| *** romcheg has quit IRC | 11:09 | |
| lucasagomes | yeah I wonder if we need to think about a better api for that | 11:09 |
| lucasagomes | dtantsur, like a real force passed to the delete command | 11:09 |
| dtantsur | lucasagomes, I don't think it should be an API, people are going to abuse it... | 11:09 |
| dtantsur | lucasagomes, I have enough people using node-delete to solve inspector problems... | 11:10 |
| lucasagomes | right, but aren't we abusing of maintenance by doing it? | 11:10 |
| dtantsur | lucasagomes, I thought it's exactly why we have maintenance: to tell Ironic not to interfere, no? | 11:11 |
| lucasagomes | "node delete will fail if node is on XYZ states but will succeed if in maintenance" | 11:11 |
| lucasagomes | dtantsur, yeah, it's like to call the attention of the operator | 11:11 |
| lucasagomes | so he can check what's wrong with the node | 11:11 |
| dtantsur | lucasagomes, and if everything is wrong, to delete it, right? | 11:12 |
| lucasagomes | dtantsur, sure, but and that's where I'm worrying whether we should a better interface to do that | 11:13 |
| lucasagomes | if operator sees BMC is failing etc... and can't delete normally he would force a deletion | 11:13 |
| lucasagomes | but the API for forcing the deletion is hidden in the maintenance | 11:13 |
| * rameshg87 goes home | 11:13 | |
| *** rameshg87 has quit IRC | 11:13 | |
| lucasagomes | dtantsur, I'm not against the change, but I want to have more eyes on that | 11:14 |
| dtantsur | lucasagomes, he/she just issues node-delete and it works | 11:14 |
| dtantsur | lucasagomes, as opposed to figure out which microversion to use to opt-in new API (yeah, versioning for the win again).. | 11:14 |
| lucasagomes | dtantsur, right but you understand that a deletion like that may have consequences, like tenant data won't be cleaned from the disk | 11:14 |
| dtantsur | lucasagomes, it's already inevitable at this point | 11:15 |
| dtantsur | (and the same applied for delete --force) | 11:15 |
| lucasagomes | yeah, I will put some thought into it after lunch | 11:21 |
| lucasagomes | I'm not against it, I just feel uncomfortable approving it as-is | 11:22 |
| lucasagomes | so I will leave it +0 | 11:22 |
| lucasagomes | and think more | 11:22 |
| *** yuikotakada has quit IRC | 11:24 | |
| *** jcoufal has quit IRC | 11:29 | |
| *** e0ne is now known as e0ne_ | 11:36 | |
| *** stendulker has quit IRC | 11:37 | |
| *** thrash|g0ne is now known as thrash | 11:40 | |
| *** lucasagomes is now known as lucas-hungry | 11:42 | |
| *** zhenguo has quit IRC | 11:43 | |
| *** trown|outttypeww is now known as trown | 11:45 | |
| *** e0ne_ has quit IRC | 11:46 | |
| *** dprince has joined #openstack-ironic | 11:49 | |
| *** cing has quit IRC | 11:51 | |
| *** romcheg has joined #openstack-ironic | 11:51 | |
| *** mtanino has joined #openstack-ironic | 12:05 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Make task_manager logging more helpful https://review.openstack.org/194619 | 12:08 |
| dtantsur | lucas-hungry, ^^ might help with these nasty "node locked" errors debugging | 12:08 |
| *** romainh has quit IRC | 12:09 | |
| *** dguerri is now known as dguerri` | 12:15 | |
| *** dguerri` is now known as dguerri | 12:15 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Always allow removing instance_uuid from node in maintenance mode https://review.openstack.org/192254 | 12:15 |
| dtantsur | lucas-hungry, no changes rebase ^^^ could you please reapprove? | 12:16 |
| *** mgarza_ has joined #openstack-ironic | 12:19 | |
| *** killer_prince has joined #openstack-ironic | 12:20 | |
| *** killer_prince has quit IRC | 12:21 | |
| *** killer_prince has joined #openstack-ironic | 12:22 | |
| *** lazy_prince has quit IRC | 12:23 | |
| *** jjohnson2 has joined #openstack-ironic | 12:26 | |
| *** killer_prince has quit IRC | 12:27 | |
| *** romainh has joined #openstack-ironic | 12:32 | |
| *** e0ne has joined #openstack-ironic | 12:37 | |
| NobodyCam | oh its early morning | 12:39 |
| * NobodyCam is heading into the office today | 12:39 | |
| dtantsur | NobodyCam, morning :) | 12:40 |
| NobodyCam | morning dtantsur | 12:40 |
| *** saripurigopi has quit IRC | 12:42 | |
| *** boris-42 has quit IRC | 12:42 | |
| *** lucas-hungry is now known as lucasagomes | 12:43 | |
| lucasagomes | dtantsur, ++ | 12:43 |
| lucasagomes | NobodyCam, morning | 12:43 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Use LOGDIR instead of SCREEN_LOGDIR in docs https://review.openstack.org/194638 | 12:46 |
| *** amotoki has joined #openstack-ironic | 12:51 | |
| *** coolsvap|away is now known as coolsvap | 12:51 | |
| BadCub | morning folks | 12:53 |
| dtantsur | BadCub, o/ | 12:54 |
| BadCub | hiya dtantsur :) | 12:54 |
| lucasagomes | BadCub, good ugt morning | 12:55 |
| BadCub | hiya lucasagomes :) | 12:55 |
| * BadCub must deliver NobodyCam tpo the shuttle. | 12:55 | |
| *** smoriya has quit IRC | 12:59 | |
| *** coolsvap is now known as coolsvap|away | 13:01 | |
| *** coolsvap|away is now known as coolsvap | 13:08 | |
| *** rameshg87 has joined #openstack-ironic | 13:17 | |
| *** openstack has quit IRC | 13:17 | |
| *** openstack has joined #openstack-ironic | 13:19 | |
| lucasagomes | yet another devstack patch for those interested: https://review.openstack.org/194656 | 13:19 |
| lucasagomes | it's fixing the logs for ir-cond | 13:19 |
| *** lazy_prince has joined #openstack-ironic | 13:20 | |
| rameshg87 | lucasagomes: oh thanks. we actually have an ironic patch proposed | 13:20 |
| lucasagomes | rameshg87, for that ^? | 13:21 |
| rameshg87 | lucasagomes: to chttps://review.openstack.org/#/c/184714/ | 13:21 |
| devananda | g'morning, all | 13:21 |
| devananda | rameshg87: btw, we should move proliantutils out of stackforge and into openstack/ | 13:21 |
| rameshg87 | \o/ | 13:21 |
| lucasagomes | w00t | 13:21 |
| lucasagomes | devananda, morning | 13:21 |
| rameshg87 | devananda: with pleasure :) | 13:21 |
| lucasagomes | rameshg87, oh I have seem this... hmmmmm | 13:22 |
| rameshg87 | and good morning | 13:22 |
| rameshg87 | lucasagomes: I had commented the same on that patch | 13:22 |
| rameshg87 | devananda: but do projects like proliantutils solely aimed at vendor hardware fit into openstack namespace ? | 13:22 |
| lucasagomes | rameshg87, I wonder tho if the change is needed on both places | 13:22 |
| rameshg87 | lucasagomes: I don't think ironic needed a change honestly | 13:22 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Make task_manager logging more helpful https://review.openstack.org/194619 | 13:23 |
| devananda | rameshg87: look at neutron project -- they have split all their drivers out recently into separate projects | 13:23 |
| dtantsur | devananda, morning! | 13:23 |
| devananda | rameshg87: short answer is "yes" | 13:23 |
| lucasagomes | rameshg87, right, I have to look more into it | 13:23 |
| rameshg87 | devananda: oh yes, but it's still it's a extension confirming to an openstack project (neutron) | 13:23 |
| dtantsur | devananda, btw can we (i.e. ifarkas :) propose an empty project for dracclient? | 13:23 |
| lucasagomes | the fix for devstack works without that change FYI | 13:23 |
| * lucasagomes have tested it here just now | 13:24 | |
| rameshg87 | I mean I am not against it ;-) | 13:24 |
| rameshg87 | by any means | 13:24 |
| lucasagomes | me neither :-) I have to read the ironic patch again | 13:25 |
| rameshg87 | and lucasagomes w.r.t your comment on https://review.openstack.org/#/c/192936/1/specs/liberty/ironic-generic-raid-interface.rst | 13:27 |
| rameshg87 | lucasagomes: we have /states/console which has both GET and PUT | 13:27 |
| lucasagomes | x.x | 13:27 |
| rameshg87 | L 293 https://review.openstack.org/#/c/192936/1/specs/liberty/ironic-generic-raid-interface.rst | 13:27 |
| lucasagomes | rameshg87, what does GET console returns? | 13:28 |
| rameshg87 | lucasagomes: I think console url | 13:28 |
| lucasagomes | GET /states/console | 13:28 |
| * rameshg87 confirms | 13:28 | |
| lucasagomes | right, and this can't be GET'd from GET v1/nodes/NNNN nor GET v1/nodes/NNNN/states right? | 13:29 |
| rameshg87 | yeah today yes :( - from 3 places, same information | 13:29 |
| lucasagomes | cause the problem with the power and provision is that the current and target information can be get from those 2 places | 13:29 |
| lucasagomes | so it's redudant | 13:29 |
| lucasagomes | rameshg87, we can get the console URL from those 3 places? | 13:29 |
| rameshg87 | lucasagomes: oh right ! | 13:30 |
| rameshg87 | only console_enabled from first two | 13:30 |
| rameshg87 | and actual console url from 3rd | 13:30 |
| rameshg87 | lucasagomes: so shouldn't we restrict just to GET /v1/nodes/NNNN/states/raid instead | 13:31 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Add iBoot driver documentation https://review.openstack.org/193546 | 13:31 |
| rameshg87 | instead of returning information on GET /v1/nodes/NNNN and GET /v1/nodes/NNNN/states | 13:31 |
| lucasagomes | rameshg87, that would be ideal, but it's incosistent with what we already have | 13:31 |
| lucasagomes | historical problems | 13:31 |
| lucasagomes | rameshg87, so that's what I was proposing, keep it consistent with the rest of the states | 13:32 |
| lucasagomes | because otherwise | 13:32 |
| lucasagomes | GET /v1/nodes/NNNN/states/raid - WORKS | 13:32 |
| lucasagomes | GET /v1/nodes/NNNN/states/power - FAIL | 13:32 |
| lucasagomes | GET /v1/nodes/NNNN/states/provision - FAIL | 13:32 |
| lucasagomes | but PUT works for all | 13:32 |
| lucasagomes | it looks very inconsistent | 13:32 |
| * lucasagomes is in a call, gimme 10 min please | 13:33 | |
| rameshg87 | sure .. | 13:33 |
| openstackgerrit | Jarrod Johnson proposed stackforge/pyghmi: Provide useful error on bad username https://review.openstack.org/194660 | 13:38 |
| rameshg87 | lucasagomes: on the counter argument, we could start fixing things from now on | 13:44 |
| rameshg87 | lucasagomes: so may be start putting up things on the same GET and PUT url | 13:45 |
| *** coolsvap is now known as coolsvap|away | 13:46 | |
| *** ppal has joined #openstack-ironic | 13:51 | |
| *** ppal has left #openstack-ironic | 13:52 | |
| lucasagomes | rameshg87, back | 13:54 |
| rameshg87 | lucasagomes: I am here too | 13:54 |
| lucasagomes | rameshg87, right, so, even if we add GET v1/states/{power, provision} | 13:54 |
| lucasagomes | it's duplicated, with GET /v1/nodes/NNN | 13:54 |
| rameshg87 | lucasagomes: honestly, in the API asking people to GET and PUT information in different endpoints sounds weird :( | 13:54 |
| lucasagomes | *v1/nodes/NNN/states/{power,provision} | 13:55 |
| rameshg87 | lucasagomes: yeah I agree with that too. | 13:55 |
| lucasagomes | rameshg87, yeah | 13:55 |
| lucasagomes | the console is different because the GET returns something that is not duplicated | 13:55 |
| lucasagomes | but power, provision and the proposed raid all the GET will return can already be fetched from the node representation | 13:56 |
| lucasagomes | + with us allowing getting a subset of the attribtues of the resources it's even more odd | 13:56 |
| rameshg87 | hmm .. | 13:57 |
| rameshg87 | so can't we just accept the difference for now | 13:58 |
| rameshg87 | lucasagomes: yes, power are provision are there in /v1/nodes/NNN and /v1/nodes/NNN/states because of historical reasons | 13:59 |
| rameshg87 | *power and provision | 13:59 |
| rameshg87 | lucasagomes: but from a particular minor api version | 13:59 |
| rameshg87 | we start switching the states to their own GET and PUT | 13:59 |
| rameshg87 | like as you said, we can add GET v1/states/{power, provision} as well | 13:59 |
| rameshg87 | should be straightforward | 13:59 |
| rameshg87 | lucasagomes: wdyt ? | 14:00 |
| rameshg87 | with the same minor api version, we could start doing them .. | 14:00 |
| lucasagomes | rameshg87, does it worth it? Cause I don't think it solves the problem | 14:01 |
| *** rwsu has joined #openstack-ironic | 14:01 | |
| lucasagomes | I mean, unless we are thinking of removing those info for the node resource too | 14:01 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Refactor method to add or update capability string https://review.openstack.org/162451 | 14:01 |
| rameshg87 | lucasagomes: but at least doesn't things start getting consistent ? | 14:01 |
| lucasagomes | this info will still be present on multiple endpoints | 14:01 |
| rameshg87 | lucasagomes: I don't think removing the info is a good idea, but we will breaking third party libraries for no reason | 14:02 |
| rameshg87 | lucasagomes: but can't power and provision be duplicated (as they are anyway today) ? | 14:02 |
| lucasagomes | rameshg87, so the propose is just to add GET to the states/{provision, power} and continue to allow GET /states ? | 14:02 |
| rameshg87 | allow the following: GET /v1/nodes/NNN/states/raid, PUT /v1/nodes/NNN/states/raid. to maintain consistency add GET /v1/nodes/NNN/states/power, GET /v1/nodes/NNN/states/provision | 14:03 |
| lucasagomes | rameshg87, right but it doesn't maintain consistency because GET v1/nodes/NNN/states will work for power and provision but not raid | 14:04 |
| lucasagomes | there's no consistent there | 14:04 |
| *** Nisha has joined #openstack-ironic | 14:04 | |
| lucasagomes | that's why I proposed to just mimic what power and provision does in raid | 14:05 |
| lucasagomes | so, although it's ugly, at least it's consistent with what we have | 14:05 |
| lucasagomes | but it will be more ugly imo to duplicate power and provision once more in another endpoint | 14:05 |
| rameshg87 | hmm .. okay | 14:07 |
| lucasagomes | rameshg87, I'm looking ftw to have a api v2 tho | 14:07 |
| lucasagomes | cause many of those decisions were taking when we didn't know anything about ReST or how an API for baremetal would look like | 14:08 |
| rameshg87 | lucasagomes: yeah, so v2 should be more consistent and better for the current ironic | 14:08 |
| lucasagomes | definitely | 14:08 |
| *** zz_jgrimm is now known as jgrimm | 14:09 | |
| lucasagomes | rameshg87, but you got why I think that it would bring more incosistency if we start allowing GET's for power and provision there? | 14:09 |
| rameshg87 | lucasagomes: yeah, I got that part. but still ... | 14:10 |
| lucasagomes | yeah | 14:12 |
| lucasagomes | still not great | 14:12 |
| rameshg87 | lucasagomes: idk. still we decide we could say "{power,provision} were available in /v1/nodes/NNN and /v1/nodes/NNN/states for *historical reasons*, but if you are reading it now, read it from GET /v1/nodes/NNN/states/{power,provision} because that's the right place" | 14:12 |
| rameshg87 | I agree adding one more place is weird, but to me we are at least correcting something w.r.t the current ironic | 14:13 |
| lucasagomes | right, but who will know that? Once they read the documentation of the api (which is auto generated) he will be like | 14:13 |
| lucasagomes | wtf | 14:13 |
| rameshg87 | hmm :D | 14:13 |
| lucasagomes | he/she won't know the history around it | 14:13 |
| lucasagomes | so it's better to have only one way to do it for all the states which has target_ and current_ | 14:14 |
| lucasagomes | even if not great | 14:14 |
| lucasagomes | than 2 ways and blame history | 14:14 |
| jroll | morning y'all | 14:14 |
| rameshg87 | jroll: o/ | 14:14 |
| jroll | heya rameshg87 lucasagomes dtantsur devananda and anyone else lurking :) | 14:14 |
| lucasagomes | jroll, yo! morning (ugt) | 14:15 |
| dtantsur | morning jroll! | 14:15 |
| rameshg87 | lucasagomes: yeah we could count on v2 to correct it then. | 14:15 |
| jroll | lucasagomes: can I have your eyes on this at some point https://review.openstack.org/#/c/194453/2 | 14:15 |
| lucasagomes | rameshg87, ++ | 14:15 |
| jroll | and everyone else with opinions on nova things | 14:15 |
| lucasagomes | jroll, ++ I will have a call (quick) very soon | 14:15 |
| lucasagomes | I will review it just after | 14:15 |
| jroll | yeah no rush | 14:16 |
| jroll | just a heads up | 14:16 |
| lucasagomes | s/just/right | 14:16 |
| lucasagomes | I mean right after* | 14:16 |
| lucasagomes | urgh haha damn... I can't type today | 14:16 |
| dtantsur | jroll, problem description is awesome! I would approve the patch just for that | 14:17 |
| jroll | dtantsur: :) | 14:17 |
| jroll | dtantsur: needs more gif though | 14:17 |
| *** mgarza_ has quit IRC | 14:17 | |
| dtantsur | ++ | 14:18 |
| lucasagomes | yuriyz, hi there, re https://review.openstack.org/#/c/194132/ so the suggestion is to move it after it's registered? | 14:19 |
| lucasagomes | to move that block of code* | 14:19 |
| yuriyz | hi lucasagomes yes move after registering | 14:19 |
| lucasagomes | yuriyz, ack, thanks for the review. I will update it soon | 14:20 |
| jroll | lucasagomes: hm, is that true for agent driver? | 14:23 |
| jroll | I think it will succeed if a conductor dies | 14:23 |
| lucasagomes | jroll, because the image is copied in the node? + hearbeat? | 14:24 |
| jroll | lucasagomes: yeah | 14:24 |
| devananda | jroll: g'morning | 14:25 |
| lucasagomes | but it's the conductor that moves the node from DEPLOYING to ACTIVE | 14:25 |
| lucasagomes | jroll, + the node will have a lock so other conductors won't touch it | 14:25 |
| jroll | lucasagomes: mmmm, locks | 14:25 |
| lucasagomes | jroll, gimme 10~15 min i'm going join the call | 14:25 |
| lucasagomes | brb | 14:25 |
| jroll | yeah np | 14:25 |
| jroll | me too :P | 14:25 |
| lucasagomes | jroll, yeah so it will never get out of the DEPLOYING state | 14:25 |
| jroll | depends how long conductor was down :) | 14:26 |
| lucasagomes | there's no timeout | 14:26 |
| lucasagomes | there's no timeout | 14:26 |
| devananda | random ? - have we fix the 'node stuck in cleaning' issues? | 14:27 |
| jroll | not really :/ | 14:27 |
| devananda | gah | 14:28 |
| devananda | so I keep hitting that in my NUC tests locally | 14:28 |
| jroll | devananda: yeah, it's becoming painful down here | 14:29 |
| devananda | the API is frustrating -- even if I set maintenance mode, I still can't get out of CLEANING without mucking with the DB | 14:29 |
| jroll | I have a feeling JoshNang has patches incoming | 14:29 |
| jroll | devananda: that's true for all states fwiw :) | 14:29 |
| jroll | but some have a timeout | 14:29 |
| jroll | transient states* | 14:29 |
| openstackgerrit | Merged openstack/ironic: Always allow removing instance_uuid from node in maintenance mode https://review.openstack.org/192254 | 14:30 |
| devananda | jroll: nova spec looks good | 14:31 |
| jroll | \o/ | 14:32 |
| devananda | jroll: though you could add some verbiage to History | 14:32 |
| jroll | shipit.gif | 14:32 |
| jroll | yeah, I wasn't sure what that was supposed to look like | 14:32 |
| jroll | template was confusing | 14:32 |
| jroll | and it was like 6:30pm :P | 14:32 |
| dtantsur | devananda, that's why (at least looks similar) I proposed https://review.openstack.org/#/c/192254/ and https://review.openstack.org/#/c/191736/ | 14:32 |
| devananda | .... maybe describe the aegean war :) | 14:33 |
| TheJulia | That seems like it could be a very long description... | 14:34 |
| devananda | dtantsur: hm, i woul dhave bumped the api versoin for https://review.openstack.org/#/c/192254/2/ironic/api/controllers/v1/node.py,cm | 14:34 |
| devananda | dtantsur: because it is a behavior change | 14:34 |
| dtantsur | well... | 14:35 |
| dtantsur | bug fixes can also change the behavior, and this is a bug fix from my pov | 14:35 |
| dtantsur | and as I have to backport it, bumping API in it brings us back to "downstream backport" heated discussion | 14:36 |
| dtantsur | i.e. downstream vendors backporting something that bumps API version | 14:36 |
| dtantsur | devananda, ^^^ | 14:36 |
| devananda | dtantsur: sure, bug fix can change behavior. but how will client know that behavior is changed otherwise? | 14:36 |
| dtantsur | devananda, how can a client detect if we fix a crash in Ironic? well, it stopped crashing :) in this case, we'll be able to `nova delete` finally... | 14:37 |
| devananda | dtantsur: "it stopped crashing" is completely different from "I no longer get an HTTP ERROR from this request ..." | 14:38 |
| dtantsur | ok, will a traceback with HTTP 500 from a server be a better analogy? | 14:38 |
| devananda | dtantsur: also, backports of API fixes are expressely not allowed -- https://wiki.openstack.org/wiki/StableBranch#Appropriate_Fixes | 14:38 |
| devananda | " Changes to the external HTTP APIs" | 14:39 |
| devananda | dtantsur: no - this isn't a trace or a crash | 14:39 |
| dtantsur | devananda, I meant downstream | 14:39 |
| * rameshg87 is done for the day | 14:40 | |
| rameshg87 | good night folks | 14:40 |
| *** rameshg87 has quit IRC | 14:40 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Clear ilo_boot_iso before deploy for glance images https://review.openstack.org/193436 | 14:40 |
| devananda | dtantsur: sorry, i dont understand what you're asking re: downstream | 14:41 |
| jroll | devananda: +1 for a version bump there | 14:41 |
| dtantsur | only after we agree on version bumps for driver_info/instance_info/properties changes | 14:42 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Clean nodes stuck in DEPLOYING state when ir-cond restarts https://review.openstack.org/194132 | 14:42 |
| dtantsur | otherwise it's clearly insane, sorry | 14:42 |
| devananda | dtantsur: no | 14:42 |
| lucasagomes | devananda, ^ | 14:42 |
| lucasagomes | jroll, and back | 14:42 |
| devananda | dtantsur: i'm sorry you disagree, but srsly, changing the behavior of the basic API itself requires a version bump. we dont need to hide old behavior -- but we DO NEED TO SIGNAL THE NEW ONE | 14:42 |
| *** lazy_prince has quit IRC | 14:43 | |
| dtantsur | devananda, I can propose a patch if you feel like (read: "you all force on me"), but I will never understand, why we don't bump things for driver_info changes which are both API-visible, change behavior and are features | 14:44 |
| *** e0ne is now known as e0ne_ | 14:44 | |
| devananda | dtantsur: drivers expose a means to discover what driver_info they require | 14:44 |
| devananda | dtantsur: were we to change the method of that discovery -- i would definitely want a version bump | 14:45 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Clean nodes stuck in DEPLOYING state when ir-cond restarts https://review.openstack.org/194132 | 14:45 |
| devananda | dtantsur: but if a driver changes what it requires (not how we ask it to validate that) -- well, ironic has no control over that | 14:45 |
| devananda | dtantsur: an out of tree driver can change its driver_info any time it wants | 14:45 |
| devananda | *its required driver_info parameteres | 14:45 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Addition of shade support to bifrot_inventory.py https://review.openstack.org/194684 | 14:46 |
| dtantsur | devananda, also re "dont need to hide old behavior", aren't we in (informal) agreement that we hide any changes under microversions right now? | 14:46 |
| TheJulia | devananda: ^^^ | 14:47 |
| dtantsur | (which I object to, but folks are still doing) | 14:47 |
| rloo | morning everyone | 14:47 |
| dtantsur | rloo, morning! | 14:47 |
| devananda | lucasagomes: I dont see where _fail_if_in_state is defined? | 14:47 |
| *** mgarza_ has joined #openstack-ironic | 14:48 | |
| lucasagomes | devananda, manager.py L1734 | 14:48 |
| devananda | dtantsur: I replied to the discussion on the ML about that -- I agree with you, no need to hide additions to the API, but we haven't formally agreed (ie, changed the spec) yet | 14:48 |
| *** e0ne_ is now known as e0ne | 14:49 | |
| *** heyongli has quit IRC | 14:49 | |
| dtantsur | devananda, so if I bump version right now, I will have to hide this change (which I still treat as a bug fix)? | 14:49 |
| devananda | dtantsur: also, on further thought, if we were really following libtool versioning, we should hide the change you made above | 14:49 |
| devananda | dtantsur: because it's not an addition, but a behavior change | 14:49 |
| devananda | yea | 14:49 |
| devananda | it's tiny -- I know | 14:49 |
| devananda | but a good test of our principles on this | 14:49 |
| rloo | dtantsur: meant to ask you. Wasn't 1.3 the breaking change? http://lists.openstack.org/pipermail/openstack-dev/2015-June/067613.html | 14:49 |
| dtantsur | *sigh* | 14:49 |
| devananda | dtantsur: I know :( | 14:49 |
| dtantsur | rloo, oh, I'm a bad reader :( | 14:50 |
| *** heyongli has joined #openstack-ironic | 14:50 | |
| dtantsur | devananda, which means that we got into situation "downstream backports version-gated changes" even faster than I expected... | 14:50 |
| devananda | yup | 14:51 |
| devananda | building software is hard :( | 14:51 |
| dtantsur | so it sucks, it makes little practical sense, but we're still doing it? | 14:51 |
| devananda | dtantsur: but if I'm a user-of-ironic, and there are two companies providing me ironic API endpoints, how do I tell the behavior of a kilo+your_fix and liberty-trunk-with-fix ? | 14:51 |
| devananda | dtantsur: that is the way it makes sense. a user who builds their application to talk to our APIs needs to be informed of behavior changes | 14:52 |
| devananda | otherwise, as you know very well, their software may break in unexpected ways | 14:52 |
| dtantsur | devananda, informed != explicitly opted in | 14:52 |
| devananda | dtantsur: imagine if discoverd relied (for some reason) on this particular functionality (API returns error from this command) and then one day that behavior changes | 14:53 |
| devananda | dtantsur: we developers can call this a bug fix, but from user POV, it doesn't matter what we call it -- the behavior just changed and their tooling broke | 14:53 |
| dtantsur | devananda, the same as if I relied on rest endpoint being absent? driver_info field never appear? | 14:54 |
| devananda | nope | 14:54 |
| devananda | I made that argument several months back, and iwas wrong about it | 14:54 |
| devananda | relying on a REST endpoint *not* existing isn't a thing. relying on an existing REST endpoint doing X, and then suddenly it does Y -- that hurts users | 14:55 |
| *** mariojv- is now known as mariojv | 14:55 | |
| devananda | I need to catch a bus into the office soon :( | 14:55 |
| jroll | what is this office | 14:55 |
| dtantsur | devananda, I don't know which "user" we're talking about. User Dmitry has a bug reported that we have undeletable Nova instances. I'm still do see users who will benefit from whatever we're doing about hiding features though.. | 14:55 |
| devananda | jroll: the place where i talk to people instead of typing to people | 14:56 |
| * devananda goes afk for a bit | 14:56 | |
| jroll | devananda: seems weird. :) | 14:56 |
| dtantsur | ++ office is a strange thing. on a bright side, here you can find people wanting to drink some beer with you in the evening :) | 14:57 |
| rloo | jroll: qq about https://review.openstack.org/#/c/180014/. I don't think driver_internal_info should be modified in parse_instance_info(). | 14:58 |
| rloo | jroll: want to check if others agree (ramesh did) | 14:58 |
| rloo | jroll: before I try to fix it | 14:59 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Clear ilo_boot_iso before deploy for glance images https://review.openstack.org/193436 | 15:00 |
| *** ijw_ has joined #openstack-ironic | 15:02 | |
| jroll | rloo: :( you're probably right | 15:03 |
| rloo | jroll: :) :( | 15:04 |
| *** absubram has joined #openstack-ironic | 15:05 | |
| *** r-daneel has joined #openstack-ironic | 15:07 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: API: Get a subset of fields from Ports and Chassis https://review.openstack.org/183549 | 15:09 |
| lucasagomes | folks can I get some eyes on ^ | 15:09 |
| lucasagomes | the Node bits was merged already | 15:09 |
| *** yog_ has quit IRC | 15:14 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Add ENROLL and related states to the state machine https://review.openstack.org/192269 | 15:21 |
| dtantsur | devananda, btw lucasagomes fields patch changed 'field' API parameter from ignored to giving "not acceptable", which IMO is just as well a behavior change (let's have metamicroversion to guard it :) | 15:23 |
| rloo | dtantsur: what's a metamicroversion? | 15:25 |
| *** amotoki has quit IRC | 15:26 | |
| dtantsur | rloo, microversion introduced to guard changes introduced by microversions | 15:26 |
| dtantsur | i.e. I'm kidding | 15:26 |
| rloo | dtantsur: is there a spec for that? | 15:26 |
| dtantsur | we should write one, I believe | 15:26 |
| rloo | dtantsur: i don't even get the joke. but I wasn't paying attention. | 15:26 |
| dtantsur | rloo, we're again changing error codes based on microversions, so I'm joking about need to microversion microversions | 15:27 |
| lucasagomes | dtantsur, but that's how the microversion works no? I mean it's true for all changes that we bump | 15:27 |
| lucasagomes | say updating "name" before it existed it was a 400 bad request | 15:28 |
| lucasagomes | now it's a 406 | 15:28 |
| dtantsur | was just going to ask what we did with names | 15:28 |
| rloo | dtantsur: wrt the microversions, until i hear otherwise, i'm going with 'status quo' (what we have done to-date) so that we don't stall development. | 15:28 |
| *** yuikotakada has joined #openstack-ironic | 15:28 | |
| devananda | I will work on taking the ML discussion and updating the spec -- unless someone else wants to do that | 15:32 |
| dtantsur | devananda, ++ please do | 15:32 |
| devananda | may not happen today. more likely tomorrow | 15:32 |
| * dtantsur is working on a change that DEFINITELY requires both version bump and gating, i.e. ENROLL one | 15:33 | |
| rloo | devananda: we don't really have a spec, except for the one that discusses client/server interaction. | 15:34 |
| *** amotoki has joined #openstack-ironic | 15:35 | |
| *** Nisha has quit IRC | 15:40 | |
| *** romcheg has quit IRC | 15:41 | |
| *** mgarza_ has quit IRC | 15:42 | |
| *** romcheg has joined #openstack-ironic | 15:42 | |
| *** mgarza_ has joined #openstack-ironic | 15:43 | |
| *** romcheg has quit IRC | 15:44 | |
| openstackgerrit | Merged stackforge/pyghmi: Provide useful error on bad username https://review.openstack.org/194660 | 15:45 |
| *** e0ne is now known as e0ne_ | 15:46 | |
| *** mgarza_ has quit IRC | 15:47 | |
| *** e0ne_ is now known as e0ne | 15:47 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: [WIP] Start using new ENROLL state https://review.openstack.org/194722 | 15:50 |
| *** mariojv has left #openstack-ironic | 15:51 | |
| *** lynxman has quit IRC | 15:51 | |
| *** lynxman has joined #openstack-ironic | 15:51 | |
| *** mgarza_ has joined #openstack-ironic | 15:51 | |
| dtantsur | see you all tomorrow, g'night! | 15:52 |
| sambetts | see you later dtantsur o/ | 15:52 |
| *** yuikotakada has quit IRC | 15:54 | |
| lucasagomes | ramesh<tab>... | 15:55 |
| *** rwsu has quit IRC | 15:58 | |
| *** dtantsur is now known as dtantsur|afk | 16:01 | |
| *** viktors is now known as viktors|afk | 16:01 | |
| *** jgrimm is now known as zz_jgrimm | 16:03 | |
| *** ifarkas has quit IRC | 16:05 | |
| *** spandhe has joined #openstack-ironic | 16:09 | |
| *** achanda has joined #openstack-ironic | 16:12 | |
| *** romainh has left #openstack-ironic | 16:13 | |
| *** ijw_ has quit IRC | 16:17 | |
| *** romcheg has joined #openstack-ironic | 16:17 | |
| lucasagomes | folks I will have to call a day a bit earlier today | 16:28 |
| lucasagomes | have a good night everyone! | 16:29 |
| *** e0ne has quit IRC | 16:29 | |
| *** lucasagomes is now known as lucas-dinner | 16:29 | |
| *** krtaylor has quit IRC | 16:29 | |
| *** spandhe has quit IRC | 16:30 | |
| *** ijw_ has joined #openstack-ironic | 16:35 | |
| *** davideagnello has quit IRC | 16:43 | |
| *** davideagnello has joined #openstack-ironic | 16:44 | |
| *** marzif_ has joined #openstack-ironic | 16:50 | |
| *** krtaylor has joined #openstack-ironic | 16:51 | |
| *** ukalifon1 has quit IRC | 16:54 | |
| *** ukalifon1 has joined #openstack-ironic | 16:55 | |
| *** ukalifon1 has quit IRC | 16:55 | |
| *** marzif has quit IRC | 16:56 | |
| SpamapS | TheJulia: btw, on bifrost, we need to figure out how to have a local mod to the settings without having a diff in git. | 16:57 |
| cinerama | SpamapS: oh like in playbooks/inventory/group_vars/all? | 16:59 |
| SpamapS | cinerama: correct | 17:00 |
| SpamapS | the README has you editting that | 17:01 |
| *** zz_jgrimm is now known as jgrimm | 17:01 | |
| SpamapS | right choice is probably to generate that file from user input + template | 17:01 |
| *** dontalton has joined #openstack-ironic | 17:01 | |
| *** Nisha has joined #openstack-ironic | 17:08 | |
| *** pelix has quit IRC | 17:15 | |
| *** mitchjameson has joined #openstack-ironic | 17:16 | |
| TheJulia | SpamapS: yes, that. or json based inventory | 17:18 |
| SpamapS | TheJulia: the inventory isn't the problem for me. Its the network settings. | 17:18 |
| SpamapS | so really _and_ :) | 17:18 |
| TheJulia | yeah, overides, and the dynamic stuff lets you override | 17:19 |
| * TheJulia laughhs evilly | 17:19 | |
| SpamapS | TheJulia: also I've figured out how to iterate a little faster by getting to the pxe commandline so at least I can try something twice before reset. :) | 17:19 |
| TheJulia | \o/ | 17:20 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Image service should not be set in ImageCache constructor https://review.openstack.org/194752 | 17:20 |
| *** dguerri is now known as dguerri` | 17:21 | |
| TheJulia | SpamapS: the plus of identifying the user class is that the only thing it gets is direct to the ipxe image | 17:23 |
| SpamapS | Ok my ipxe command line knowledge is too low | 17:24 |
| SpamapS | so I couldnt' actually do the same thing on the second try :-P | 17:24 |
| * SpamapS just waits the 8 minutes | 17:24 | |
| sambetts | Night all o/ | 17:24 |
| TheJulia | goodnight sambetts | 17:25 |
| *** harlowja has joined #openstack-ironic | 17:26 | |
| SpamapS | TheJulia: the chain load worked right! | 17:29 |
| TheJulia | \o/ | 17:29 |
| SpamapS | TheJulia: weird though, you end up doing DHCP on that interface 3 times. | 17:30 |
| SpamapS | which takes a while because the other two interfaces have no link | 17:30 |
| SpamapS | TheJulia: does "deploying" mean the agent is doing something? | 17:31 |
| *** rwsu has joined #openstack-ironic | 17:31 | |
| TheJulia | yeah, the i ages dont know about the last dhcp request... :( | 17:31 |
| SpamapS | oo | 17:31 |
| SpamapS | its rebooting | 17:31 |
| SpamapS | zomg | 17:31 |
| TheJulia | SpamapS: that it is writing out the image | 17:31 |
| *** rwsu has quit IRC | 17:33 | |
| SpamapS | TheJulia: question is, does that patch make the moonshot/otherstuff not work? | 17:34 |
| SpamapS | TheJulia: mmmm... boot blocks | 17:36 |
| TheJulia | well ci and local vm resting worked, i actually need to test it on physical gear, but realisitcally they should be sending their user definable class as gpxe | 17:37 |
| TheJulia | err, should not be | 17:37 |
| SpamapS | hm | 17:37 |
| TheJulia | hmm? | 17:38 |
| SpamapS | chain loading.. so much chain loading | 17:38 |
| TheJulia | :( yeah | 17:39 |
| *** rwsu has joined #openstack-ironic | 17:39 | |
| SpamapS | TheJulia: how do I log in? | 17:39 |
| TheJulia | it should have burned your ssh key in | 17:39 |
| SpamapS | it looks like it booted the deploy ramdisk again | 17:39 |
| SpamapS | not the image on disk | 17:39 |
| TheJulia | could the mellonoxes be overriding boot order? | 17:40 |
| SpamapS | Oh thats very likely that they try net before disk | 17:40 |
| SpamapS | I thought the pxe config would be removed, so it would fall through. | 17:40 |
| SpamapS | which is generally how I'd want things to work so I can regain control without having to 0 the boot block. ? | 17:41 |
| TheJulia | or maybe setting boot device failed, can you check the conductor log at /var/log/upstart/ironic-conductor.log? | 17:41 |
| SpamapS | TheJulia: there's a lot in that file. Search key? | 17:42 |
| *** Nisha has quit IRC | 17:42 | |
| *** rwsu has quit IRC | 17:42 | |
| TheJulia | ironic, should be toggling the default boot dev for the bios | 17:42 |
| TheJulia | SpamapS: look for ipmitool and bootdev i think | 17:42 |
| *** rwsu has joined #openstack-ironic | 17:43 | |
| SpamapS | Oh I never knew that we could do that with IPMI | 17:44 |
| SpamapS | 2015-06-23 10:31:27.727 10383 DEBUG ironic.common.utils [-] Command stdout is: "Set Boot Device to disk | 17:44 |
| SpamapS | so yeah, tha tdidn't work :-P | 17:45 |
| SpamapS | let me force it via the boot menu | 17:45 |
| TheJulia | any stderr? | 17:45 |
| SpamapS | TheJulia: nope | 17:46 |
| TheJulia | oh... awesome. | 17:46 |
| SpamapS | also how fun. deploy ramdisk kernel panics on nmi | 17:46 |
| TheJulia | SpamapS: sounds like a defect for coreos :( | 17:47 |
| SpamapS | Oh yeah let me get right on that. | 17:47 |
| TheJulia | lol | 17:47 |
| JayF | well hold on | 17:47 |
| JayF | we have a PR up to bump the CoreOS version | 17:47 |
| TheJulia | hi JayF | 17:47 |
| JayF | if it's passing tests now, I'll merge it and maybe it'll fix your issue SpamapS | 17:47 |
| SpamapS | JayF: its not something that is super necessary, just that I tend to try and use the graceful-est mechanism first. :) | 17:48 |
| TheJulia | SpamapS: just power it off :) | 17:48 |
| * TheJulia ducks | 17:48 | |
| SpamapS | yeah reset works fine, :) | 17:49 |
| SpamapS | TheJulia: its entirely possible that we did something in setting these boxes up that makes them ignore ipmi's instructions on how to boot | 17:49 |
| SpamapS | I'm trying with dnsmasq turned off, just to see if the hard drive works at all. | 17:49 |
| SpamapS | What I see though, the order is not changed. Net is happening before Disk. | 17:50 |
| JayF | jroll: FYI https://review.openstack.org/#/c/191505/ still failing tests :( | 17:50 |
| SpamapS | TheJulia: just wondering if it wouldn't make sense to also have ironic clean up the symlinks in pxelinux.cfg to side-step this problem. | 17:50 |
| jroll | JayF: oh yeah, need to pick that up. | 17:50 |
| TheJulia | SpamapS: for some reason, i thought it should be, but i've noticed some local artifacts and always thought it was a result of my regular behavior of "drop database ironic;\q" | 17:52 |
| SpamapS | lrwxrwxrwx 1 ironic ironic 53 Jun 23 10:05 78e7d12295ed -> /httpboot/52171505-a9ac-4c03-a707-a327f4fcb088/config | 17:53 |
| SpamapS | lrwxrwxrwx 1 ironic ironic 53 Jun 23 10:05 78-e7-d1-22-95-ed -> /httpboot/52171505-a9ac-4c03-a707-a327f4fcb088/config | 17:53 |
| SpamapS | TheJulia: been there since the first deploy | 17:53 |
| SpamapS | TheJulia: ah so I think my image is going to need dhcp-all-interfaces ... | 17:54 |
| SpamapS | oh! it has something like that | 17:55 |
| SpamapS | err, no it doesn't Hrm | 17:55 |
| SpamapS | TheJulia: does it just assume eth0 only? | 17:56 |
| TheJulia | yeah :( you can burn in dhcp-all-interfaces to the image | 17:57 |
| EntropyWorks | so I'm still trying to get a test env for ironic and now I'm following http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html and did the "vagrant up" looks like it completed without error but I'm not finding the " ./etc/ironic/ironic.conf.local" file mentioned in the dev-quickstart.html | 17:57 |
| TheJulia | SpamapS: stepping away for a couole minutes, will bbiam | 17:57 |
| SpamapS | TheJulia: np. :) thanks for helping me get further. | 17:57 |
| *** mtanino has quit IRC | 18:00 | |
| *** spandhe has joined #openstack-ironic | 18:03 | |
| TheJulia | EntropyWorks: try removing the leading . | 18:04 |
| TheJulia | EntropyWorks: warning: i've never tried the vagrent stuff, so I'm just tossing out an idea | 18:04 |
| devananda | EntropyWorks: vagrant sets up a tiny VM for running the mysql and rabbit services -- not the ironic services | 18:04 |
| *** absubram has quit IRC | 18:05 | |
| devananda | EntropyWorks: if you want to test a full openstack+ironic, use the devstack walkthrough | 18:05 |
| *** achanda_ has joined #openstack-ironic | 18:06 | |
| EntropyWorks | well damn... | 18:06 |
| TheJulia | SpamapS: you'll want to clean-up, cleanup-deployment-images.yaml should do that nicely | 18:06 |
| TheJulia | SpamapS: did you find the var to pass to add dhcp-all-interfaces? | 18:06 |
| SpamapS | TheJulia: yes | 18:07 |
| SpamapS | TheJulia: building now | 18:07 |
| TheJulia | awesome | 18:07 |
| *** spandhe has quit IRC | 18:07 | |
| SpamapS | The ipmi boot ordering thing is bothering me. | 18:07 |
| TheJulia | SpamapS: you just need to nuke the old qcow2 image before attempting to build again | 18:07 |
| SpamapS | I want IPMI to be the god that smites the local snowflakes. | 18:08 |
| SpamapS | TheJulia: indeed, thats what I did. :) | 18:08 |
| *** e0ne has joined #openstack-ironic | 18:08 | |
| TheJulia | excellent! | 18:08 |
| *** achanda has quit IRC | 18:08 | |
| SpamapS | TheJulia: how do I get ironic to re-deploy the node though? node-delete and then run deploy playbook again? | 18:09 |
| TheJulia | SpamapS: I feel that it should be, it should have a lightning bolt | 18:09 |
| SpamapS | TheJulia: you should have a lightning bolt | 18:09 |
| TheJulia | SpamapS: you can do that, suddenly realizing that a rebuild playbook would be a good idea | 18:10 |
| SpamapS | InvalidState: Can not delete node "52171505-a9ac-4c03-a707-a327f4fcb088" while it is in provision state "active". Valid provision states to perform deletion are: "('available', None, 'manageable')" | 18:11 |
| SpamapS | I'm Ironic-ignorant | 18:12 |
| SpamapS | (which is why I'm going to have to use a full nova based ironic for infra-cloud.. we're all ironigorant ;) | 18:12 |
| TheJulia | oh... ironic node-set-provision-state uuid deleted | 18:13 |
| *** spandhe has joined #openstack-ironic | 18:13 | |
| SpamapS | TheJulia: thats.. weird | 18:13 |
| SpamapS | I just set it to rebuild | 18:13 |
| SpamapS | which should send my new image. :) | 18:13 |
| * TheJulia wonders if that re-ships the configdrive | 18:14 | |
| EntropyWorks | was hoping that "vagrant up" and maybe use pipework (https://github.com/jpetazzo/pipework) would be enough. Be nice if "vagrant up" and then "vagrant ssh" would give you enough to boot something else via virtualbox or ipmi. I may have to look into that more :-D | 18:14 |
| SpamapS | TheJulia: there's a configdrive? | 18:14 |
| SpamapS | TheJulia: guessing only on ilo driver? | 18:15 |
| TheJulia | yup, we include a configdrive, thats how we get the ssh key(s) to the remote host | 18:15 |
| TheJulia | and any network config | 18:15 |
| jroll | SpamapS: configdrive should work on any agent driver | 18:15 |
| SpamapS | but you said you baked it into my image? | 18:15 |
| TheJulia | SpamapS: nope, works quite well with the agent, just don't try to send it a configdrive over 64 megs | 18:15 |
| SpamapS | OH | 18:16 |
| TheJulia | your key doesn't get baked in to the image it's self | 18:16 |
| SpamapS | agent is der smart | 18:16 |
| SpamapS | kk | 18:16 |
| jroll | :D | 18:16 |
| TheJulia | SpamapS: now, you can use the magical dib elements to do that | 18:16 |
| TheJulia | sort of like importing antigravity | 18:16 |
| *** Marga_ has quit IRC | 18:17 | |
| *** achanda_ has quit IRC | 18:17 | |
| *** amotoki has quit IRC | 18:17 | |
| *** achanda has joined #openstack-ironic | 18:17 | |
| SpamapS | rebuild failed | 18:27 |
| SpamapS | no idea why | 18:27 |
| SpamapS | | 52171505-a9ac-4c03-a707-a327f4fcb088 | None | None | power off | deploy failed | False | | 18:27 |
| SpamapS | deleted.. waiting for cleaning :-P | 18:27 |
| *** e0ne is now known as e0ne_ | 18:27 | |
| *** dguerri` is now known as dguerri | 18:28 | |
| TheJulia | If memory serves, it may bomb if you didn't define a configdrive to rebuild. Cleaning should be disabled.. | 18:29 |
| *** dguerri is now known as dguerri` | 18:29 | |
| SpamapS | | 52171505-a9ac-4c03-a707-a327f4fcb088 | None | None | power on | cleaning | False | | 18:30 |
| *** romcheg has quit IRC | 18:30 | |
| *** e0ne_ has quit IRC | 18:32 | |
| SpamapS | is there any way to peek at what the agent is doing? | 18:33 |
| SpamapS | I just see cleaning | 18:33 |
| *** athomas has quit IRC | 18:34 | |
| jroll | SpamapS: console | 18:34 |
| SpamapS | jroll: just login | 18:34 |
| jroll | SpamapS: alternatively there's ways to inject ssh keys into the image | 18:34 |
| jroll | SpamapS: yeah, and then.... | 18:34 |
| SpamapS | login: last thing on console | 18:34 |
| *** ijw_ has quit IRC | 18:35 | |
| SpamapS | conductor log says cleaning is going | 18:35 |
| jroll | journalctl -u ironic-python-agent.service -f -l | 18:35 |
| jroll | add --no-tail to not tail | 18:35 |
| jroll | oh is the console not logged in? | 18:35 |
| SpamapS | no | 18:35 |
| jroll | grr | 18:35 |
| JayF | for the console to login | 18:36 |
| jroll | add 'coreos.autologin=ttyS4' to kernel cmd line | 18:36 |
| jroll | or whatever tty | 18:36 |
| JayF | jroll: the =ttyS4 is specific to us :) | 18:36 |
| SpamapS | 2015-06-23 11:35:58.154 17723 DEBUG ironic.drivers.modules.agent_base_vendor [-] Clean step still running for node 52171505-a9ac-4c03-a707-a327f4fcb088: None _get_completed_cleaning_command /usr/local/lib/python2.7/dist-packages/ironic/drivers/modules/agent_base_vendor.py:348 | 18:36 |
| JayF | without an = will make it autologin on all TTYs | 18:36 |
| jroll | JayF: oh interesting | 18:36 |
| JayF | coreos.autologin is therefor sufficient | 18:36 |
| jroll | SpamapS: also, 'sshkey=yourpubkeyhere' is a valid kernel cmd line | 18:37 |
| TheJulia | SpamapS: cleaning can take a long time, its zeroing the disks | 18:37 |
| TheJulia | SpamapS: which is why we turn it off for testing | 18:37 |
| SpamapS | TheJulia: yes, cleaning should definitely be off then. :) | 18:37 |
| SpamapS | TheJulia: so -e cleaning=false in install.yaml ? | 18:38 |
| jroll | is nobody using SSDs with secure erase features? :/ | 18:38 |
| TheJulia | SpamapS: yeah, or ./playbooks/roles/ironic-install/defaults/main.yml or /etc/ironic/ironic.conf :) | 18:38 |
| TheJulia | jroll: no | 18:38 |
| TheJulia | :) | 18:39 |
| jroll | sadfaaaaaace. | 18:39 |
| TheJulia | very sadface, if memory serves, these machines have 2TB disks? | 18:39 |
| TheJulia | SpamapS: I think you can still set the power state off, I honestly don't remember the exact state of when you can and cannot abort cleaning | 18:40 |
| *** e0ne has joined #openstack-ironic | 18:41 | |
| SpamapS | how long might it take to 0 2.5TB ? Ugh. Can I cancel cleaning? | 18:41 |
| TheJulia | power it off :) | 18:41 |
| TheJulia | and just... nuke the node from mysql | 18:41 |
| SpamapS | TheJulia: and then I can delete? | 18:41 |
| TheJulia | I honestly don't remember | 18:42 |
| TheJulia | jroll: didn't some stuff land recently to block killing cleaning state? | 18:42 |
| jroll | TheJulia: block killing it? I don't think we ever allowed killing it | 18:42 |
| TheJulia | SpamapS: your combating ERRMIGRAINE at the moment which introduces random recall errors | 18:43 |
| jroll | but I'd love a patch to allow it | 18:43 |
| *** marzif_ has quit IRC | 18:43 | |
| SpamapS | so there's no state transition possible by the user from cleaning? | 18:43 |
| jroll | correct | 18:44 |
| jroll | huge pain point | 18:44 |
| TheJulia | jroll: something like setting the state to "ididntwantcleaning..pleasestop...please" | 18:44 |
| SpamapS | jroll: ow | 18:44 |
| jroll | JoshNang: is this something you're working on? | 18:44 |
| openstackgerrit | Merged openstack/bifrost: Check for and install easy_install https://review.openstack.org/193400 | 18:47 |
| *** spandhe has quit IRC | 18:48 | |
| openstackgerrit | Merged openstack/bifrost: Remove not-currently-required amtterm dependency https://review.openstack.org/194458 | 18:48 |
| SpamapS | ok so hopefully I didn't screw it up too bad by just stopping conductor, updating the table, and then restarting conductor and deleting the node. | 18:48 |
| *** natorious is now known as zz_natorious | 18:49 | |
| TheJulia | Eh... most likely not :) | 18:49 |
| *** rwsu has quit IRC | 18:51 | |
| *** dguerri` is now known as dguerri | 18:53 | |
| *** rwsu has joined #openstack-ironic | 18:53 | |
| *** dguerri is now known as dguerri` | 18:54 | |
| SpamapS | (also turning off cleaning in the middle there) | 18:55 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Add README note about cleaning https://review.openstack.org/194783 | 18:57 |
| SpamapS | so now I'm kind of stumped how to make my BIOS respect the bootdev piece | 18:57 |
| SpamapS | Right now I just turn off dnsmasq until pxe passes on console, and then turn it back on while the kernel boots. | 18:58 |
| TheJulia | SpamapS: so what is the boot order that is defined in the bios? | 18:58 |
| SpamapS | TheJulia: have not checked that yet | 18:58 |
| SpamapS | TheJulia: I would expect it is net->disk | 18:58 |
| TheJulia | those are converged adapters right? | 19:00 |
| SpamapS | I don't know what that means. | 19:00 |
| SpamapS | textcons can't get into the bios, will have to use vsp on the next iteration | 19:01 |
| SpamapS | But really, if we just dropped pxelinux.cfg/${mac} ... wouldn't matter. | 19:02 |
| TheJulia | eh, then you'll never really be able to deploy | 19:02 |
| TheJulia | but yeah | 19:03 |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Save disk layout information when deploying https://review.openstack.org/194786 | 19:03 |
| SpamapS | TheJulia: but then <-- consequence of which condition, converged adapters, or dropping pxelinux.cfg ? | 19:03 |
| *** mgarza_ has quit IRC | 19:04 | |
| TheJulia | SpamapS: converged being delivery of network and disk access | 19:04 |
| TheJulia | SpamapS: but yeah was in reference to pxelinux.cfg/${mac} | 19:05 |
| SpamapS | TheJulia: I think it would work fine. You write it back out again when you want the agent to boot. | 19:05 |
| TheJulia | but I think it is a holdover from the case where nodes don't have a local bootloader burned to disk | 19:05 |
| *** rwsu has quit IRC | 19:05 | |
| TheJulia | Bifrost really wouldn't care though, the playbook could purge the pxelinux.cfg/${mac} symlink after deploy is completed | 19:06 |
| SpamapS | TheJulia: So except for the ipmi thing, it worked | 19:09 |
| SpamapS | I'm in and I have a real Ubuntu box. ;) | 19:09 |
| TheJulia | congrats! | 19:09 |
| TheJulia | Enjoy! | 19:09 |
| SpamapS | TheJulia: and relatively repeatable too :) | 19:10 |
| SpamapS | just need your patch, plus somebody who can fix boot order or toggle dnsmasq off/on at the right time. :) | 19:10 |
| SpamapS | Ah interesting | 19:11 |
| TheJulia | SpamapS: so I have a facts module proposed, we could use that to figure out when the transition to active occurs and nuke the symlink | 19:11 |
| *** hblixt has joined #openstack-ironic | 19:11 | |
| SpamapS | so one thing that failed is that config drive shoved a static config in, but it was assuming eth0 | 19:12 |
| TheJulia | at least, I think... it might get weird. | 19:12 |
| SpamapS | TheJulia: I am not excited about that plan. | 19:12 |
| TheJulia | That figures, its a known issue | 19:12 |
| TheJulia | SpamapS: https://review.openstack.org/#/c/192224/ | 19:13 |
| TheJulia | The world might be happier then :) | 19:13 |
| SpamapS | TheJulia: all of this pain may be for not.. we are getting some feedback on how soon we can have all the nics replaced with intel nics.. and in doing so, they will likely update BIOS. | 19:13 |
| *** rwsu has joined #openstack-ironic | 19:13 | |
| TheJulia | SpamapS: awesome | 19:13 |
| SpamapS | TheJulia: is there enough metadata in the config drive for glean to find the MAC and match it to the physical interface? | 19:14 |
| TheJulia | SpamapS: yes, if glean can execute :) | 19:14 |
| SpamapS | nice | 19:14 |
| TheJulia | hence patch | 19:14 |
| SpamapS | yeah, I dunno about a static PATH, it seems wrong. ;) | 19:15 |
| SpamapS | I'll ponder it over lunch. | 19:15 |
| *** mariojv has joined #openstack-ironic | 19:15 | |
| TheJulia | yeah... except can you trust that your existing path will include /usr/local/[bin|sbin] | 19:15 |
| *** mgarza_ has joined #openstack-ironic | 19:15 | |
| *** hblixt has quit IRC | 19:16 | |
| TheJulia | SpamapS: enjoy, I only tested it with the dynamic env, I may have a problem in the csv path, I didn't manually test it, but I know glean is not executing on stock ubuntu right now :( | 19:16 |
| TheJulia | s/env/inventory/ | 19:16 |
| *** openstackgerrit has quit IRC | 19:21 | |
| *** openstackgerrit has joined #openstack-ironic | 19:22 | |
| *** zz_natorious is now known as natorious | 19:25 | |
| *** achanda has quit IRC | 19:26 | |
| JoshNang | jroll: it's in my todo-list, hoping to put up a clean->fail transition sometime this week. it wouldn't help for skipping it though (without a conductor restart after failing) | 19:38 |
| jroll | JoshNang: cool | 19:38 |
| *** dprince has quit IRC | 19:38 | |
| *** Marga_ has joined #openstack-ironic | 19:39 | |
| *** rwsu has quit IRC | 19:41 | |
| *** achanda has joined #openstack-ironic | 19:43 | |
| *** rwsu has joined #openstack-ironic | 19:45 | |
| NobodyCam | woo hoo /me has connection again | 19:46 |
| TheJulia | yay! | 19:47 |
| *** rwsu has quit IRC | 19:51 | |
| *** romcheg has joined #openstack-ironic | 19:55 | |
| *** Marga_ has quit IRC | 19:56 | |
| *** Marga_ has joined #openstack-ironic | 19:57 | |
| *** e0ne is now known as e0ne_ | 20:05 | |
| *** e0ne_ is now known as e0ne | 20:05 | |
| *** achanda has quit IRC | 20:06 | |
| *** greghaynes has quit IRC | 20:08 | |
| *** ndipanov has quit IRC | 20:10 | |
| *** romainh has joined #openstack-ironic | 20:10 | |
| *** Marga_ has quit IRC | 20:13 | |
| *** Marga_ has joined #openstack-ironic | 20:14 | |
| *** Marga_ has quit IRC | 20:15 | |
| *** Marga_ has joined #openstack-ironic | 20:15 | |
| *** Sukhdev has joined #openstack-ironic | 20:20 | |
| *** dontalton has quit IRC | 20:32 | |
| *** dontalton has joined #openstack-ironic | 20:33 | |
| *** greghaynes has joined #openstack-ironic | 20:33 | |
| *** greghaynes has quit IRC | 20:33 | |
| *** rwsu has joined #openstack-ironic | 20:39 | |
| mrda | Morning Ironic | 20:46 |
| TheJulia | good morning mrda | 20:49 |
| mrda | Hi TheJulia | 20:49 |
| *** romcheg has quit IRC | 20:53 | |
| *** romcheg has joined #openstack-ironic | 20:53 | |
| *** alexpilotti has quit IRC | 20:57 | |
| *** trown is now known as trown|outttypeww | 21:02 | |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Refactor ironic-install playbook to support CentOS & RHEL https://review.openstack.org/193742 | 21:04 |
| NobodyCam | morning mrda | 21:04 |
| mrda | Hey NobodyCam | 21:05 |
| NobodyCam | :) how things down under there mrda | 21:06 |
| *** jjohnson2 has quit IRC | 21:06 | |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Refactor ironic-install playbook to support CentOS & RHEL https://review.openstack.org/193742 | 21:06 |
| mrda | NobodyCam: Everything is fine here, busy but fine :) | 21:07 |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Refactor ironic-install playbook to support CentOS & RHEL https://review.openstack.org/193742 | 21:08 |
| NobodyCam | mrda: busy is good :) | 21:08 |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Refactor ironic-install playbook to support CentOS & RHEL https://review.openstack.org/193742 | 21:09 |
| *** radek__ has quit IRC | 21:10 | |
| *** Marga_ has quit IRC | 21:10 | |
| jroll | NobodyCam: says who :) | 21:13 |
| krotscheck | ok, fixed the webclient project patch. | 21:17 |
| krotscheck | bleargh | 21:17 |
| * krotscheck needs to figure out how to add an ACL group | 21:17 | |
| krotscheck | Unless all y'all want to become experts at Javascript :) | 21:17 |
| jroll | what makes you think I'm not? :) | 21:18 |
| * mrda thinks jroll just volunteered | 21:18 | |
| jroll | I'm fine with reviewing ironic-webclient, though it won't be top priority | 21:19 |
| mrda | :) | 21:19 |
| *** romcheg has quit IRC | 21:21 | |
| NobodyCam | :p | 21:22 |
| JayF | jroll: should I out you as a former frontend dev | 21:22 |
| jroll | I think you just idd :) | 21:23 |
| jroll | did | 21:23 |
| JayF | whoops ;) | 21:23 |
| jroll | I don't mind, to be clear :P | 21:27 |
| *** e0ne is now known as e0ne_ | 21:28 | |
| JayF | It's a great story; it didn't take you long to detox and start working on real infrastructure problems ;) | 21:28 |
| JayF | </troll> | 21:28 |
| *** krtaylor has quit IRC | 21:29 | |
| *** e0ne_ has quit IRC | 21:29 | |
| rloo | mrda: morning. you might be interested in this (having a hard time adding you as a reviewer): https://review.openstack.org/#/c/193587/ | 21:31 |
| NobodyCam | morning rloo, JayF krotscheck and jroll :) | 21:32 |
| *** romainh has quit IRC | 21:32 | |
| JayF | mornign :) | 21:32 |
| jroll | hiya NobodyCam :) | 21:32 |
| mrda | thanks rloo | 21:33 |
| mrda | (it auto completes case sensitively, if that helps :) | 21:34 |
| *** e0ne has joined #openstack-ironic | 21:34 | |
| rloo | mrda: works now, maybe I can't type when in a meeting | 21:36 |
| NobodyCam | :) | 21:41 |
| *** e0ne is now known as e0ne_ | 21:42 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/194320 | 21:43 |
| *** e0ne_ is now known as e0ne | 21:45 | |
| *** e0ne has quit IRC | 21:55 | |
| openstackgerrit | Merged openstack/ironic: Use LOGDIR instead of SCREEN_LOGDIR in docs https://review.openstack.org/194638 | 21:56 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Updated from global requirements https://review.openstack.org/194320 | 22:04 |
| *** Sukhdev has quit IRC | 22:19 | |
| *** dontalton2 has joined #openstack-ironic | 22:19 | |
| *** jgrimm is now known as zz_jgrimm | 22:21 | |
| openstackgerrit | Josh Gachnang proposed openstack/ironic-specs: Implement Zapping States https://review.openstack.org/185122 | 22:27 |
| mrda | rloo: Got a moment to chat? | 22:41 |
| *** rwsu has quit IRC | 22:47 | |
| *** natorious is now known as zz_natorious | 22:51 | |
| *** r-daneel has quit IRC | 23:00 | |
| *** ijw_ has joined #openstack-ironic | 23:13 | |
| *** dontalton2 has quit IRC | 23:16 | |
| *** dontalton has quit IRC | 23:16 | |
| *** lucas-dinner has quit IRC | 23:17 | |
| *** mgarza_ has quit IRC | 23:22 | |
| *** ijw_ has quit IRC | 23:28 | |
| *** ijw_ has joined #openstack-ironic | 23:28 | |
| *** yuanying has joined #openstack-ironic | 23:32 | |
| *** Nisha has joined #openstack-ironic | 23:38 | |
| *** Nisha has quit IRC | 23:42 | |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Make MTU configurable https://review.openstack.org/194876 | 23:47 |
| TheJulia | ohhhh | 23:49 |
| * TheJulia wonders where her glasses have gotten off to | 23:50 | |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Make git URLs consistent with each other https://review.openstack.org/194881 | 23:55 |
| *** ijw_ has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!