| *** Poornima has joined #openstack-ironic | 01:37 | |
| *** nosnos has joined #openstack-ironic | 01:51 | |
| *** HenryG_ is now known as HenryG | 02:01 | |
| *** jrist has quit IRC | 02:13 | |
| *** rushiagr_away is now known as rushiagr | 02:52 | |
| *** jrist has joined #openstack-ironic | 02:56 | |
| *** rushiagr is now known as rushiagr_away | 03:03 | |
| *** rushiagr_away is now known as rushiagr | 03:05 | |
| *** rushiagr is now known as rushiagr_away | 03:16 | |
| *** yarker has quit IRC | 03:24 | |
| *** lazy_prince is now known as killer_prince | 03:32 | |
| *** Poornima has quit IRC | 03:37 | |
| *** rushiagr_away is now known as rushiagr | 04:11 | |
| *** Poornima has joined #openstack-ironic | 04:13 | |
| *** eghobo has joined #openstack-ironic | 04:14 | |
| *** eghobo has quit IRC | 04:30 | |
| *** Haomeng|2 has joined #openstack-ironic | 04:45 | |
| *** Haomeng has quit IRC | 04:45 | |
| *** eghobo has joined #openstack-ironic | 04:47 | |
| *** nikunj2513 has joined #openstack-ironic | 04:48 | |
| *** eguz has joined #openstack-ironic | 04:48 | |
| *** eghobo has quit IRC | 04:49 | |
| *** eguz has quit IRC | 05:06 | |
| *** eghobo has joined #openstack-ironic | 05:06 | |
| *** rakesh_hs has joined #openstack-ironic | 05:29 | |
| *** vinbs has joined #openstack-ironic | 05:41 | |
| openstackgerrit | Syed Ismail Faizan Barmawer proposed a change to openstack/ironic: Add uefi boot mode support in IloVirtualMediaIscsiDeploy https://review.openstack.org/116561 | 05:52 |
|---|---|---|
| *** nikunj2513 has quit IRC | 05:55 | |
| *** nosnos has quit IRC | 05:57 | |
| *** nosnos_ has joined #openstack-ironic | 06:00 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/118548 | 06:10 |
| *** ifarkas has joined #openstack-ironic | 06:11 | |
| *** ndipanov has joined #openstack-ironic | 06:28 | |
| *** eguz has joined #openstack-ironic | 06:46 | |
| *** eghobo has quit IRC | 06:49 | |
| *** eguz has quit IRC | 06:55 | |
| *** bmahalakshmi has joined #openstack-ironic | 06:57 | |
| *** jcoufal has joined #openstack-ironic | 07:05 | |
| *** bmahalakshmi has quit IRC | 07:09 | |
| *** bmahalakshmi has joined #openstack-ironic | 07:09 | |
| *** bluex-pl has joined #openstack-ironic | 07:22 | |
| *** jcoufal has quit IRC | 07:23 | |
| *** killer_prince has quit IRC | 07:23 | |
| *** jcoufal has joined #openstack-ironic | 07:25 | |
| *** bmahalakshmi2 has joined #openstack-ironic | 07:29 | |
| *** jistr has joined #openstack-ironic | 07:29 | |
| *** rakesh_hs2 has joined #openstack-ironic | 07:29 | |
| *** bmahalakshmi has quit IRC | 07:30 | |
| *** rakesh_hs has quit IRC | 07:30 | |
| *** athomas has joined #openstack-ironic | 07:58 | |
| *** andreykurilin_ has joined #openstack-ironic | 07:58 | |
| dtantsur | Morning Ironic | 08:07 |
| dtantsur | mrda, great news! | 08:07 |
| *** derekh has joined #openstack-ironic | 08:07 | |
| mrda | hey dtantsur | 08:09 |
| mrda | are you talking about driver? | 08:09 |
| dtantsur | yeah | 08:09 |
| mrda | so we're 4 patches merged | 08:10 |
| mrda | One has been in gate, failed, retrying all day today | 08:10 |
| mrda | so 3 more to go | 08:10 |
| mrda | but good progress, s'sure | 08:10 |
| *** viktors|afk is now known as viktors | 08:11 | |
| romcheg | Good morning all! | 08:12 |
| mrda | hi romcheg | 08:12 |
| dtantsur | mrda, should we approve https://review.openstack.org/#/c/119165/ ? | 08:12 |
| dtantsur | romcheg, morning | 08:13 |
| mrda | dtantsur: I would like to see that pass dsvm_virtual-irtonic, but if it does, sure! | 08:14 |
| dtantsur | mrda, well, it won't merge if it does not :) | 08:14 |
| mrda | We should discuss in the meeting in 11 hours about the timeline for removing the driver from our code | 08:14 |
| mrda | then we can stop on these backports | 08:14 |
| dtantsur | heh yeah | 08:15 |
| mrda | It's dinner time here, so I'll take a break. If people can keep an eye on 111429, that's be great. If it fails in the gate (again) please 'reverify'. Once that merges then we can try again with 115540 and 115549 (bit no point until the previous one in the chain merges) | 08:16 |
| mrda | s/bit/but/ | 08:17 |
| mrda | thanks in advance, and see you at the team meeting :) | 08:17 |
| *** lucasagomes has joined #openstack-ironic | 08:21 | |
| dtantsur | see you | 08:27 |
| lucasagomes | jroll, hi there, did you get a change to check https://review.openstack.org/#/c/118326/ out? | 08:37 |
| *** MattMan has joined #openstack-ironic | 08:38 | |
| openstackgerrit | A change was merged to openstack/ironic: Updated from global requirements https://review.openstack.org/119141 | 08:47 |
| mrda | woohoo! 111429 just merged | 08:58 |
| mrda | Two left - 115540 and 115549 | 08:58 |
| *** Poornima has quit IRC | 08:59 | |
| *** Poornima_ has joined #openstack-ironic | 08:59 | |
| dtantsur | \o/ | 09:01 |
| lucasagomes | w00t! | 09:13 |
| *** dguerri`afk is now known as dguerri | 09:13 | |
| *** andreykurilin_ has quit IRC | 09:13 | |
| dtantsur | lucasagomes, mind having a look at https://review.openstack.org/116303 ? I'd prefer to land this kind of potential merge hell asap :) | 09:20 |
| *** eghobo has joined #openstack-ironic | 09:22 | |
| lucasagomes | dtantsur, will do, I'm just finishing the UEFI one | 09:23 |
| dtantsur | ack yeah | 09:23 |
| *** eghobo has quit IRC | 09:23 | |
| lucasagomes | dtantsur, we should target it to RC https://bugs.launchpad.net/ironic/+bug/1364813 | 09:32 |
| lucasagomes | do you think it's an RC bug? | 09:32 |
| dtantsur | lucasagomes, targeted. I think it's a plus to do cleanups for graduation | 09:33 |
| lucasagomes | dtantsur, cool yeah | 09:33 |
| lucasagomes | I'm just checking whether this is not going to cause merge conflicts with other patches like uefi | 09:34 |
| lucasagomes | which is priority | 09:34 |
| lucasagomes | I think it's fine | 09:34 |
| dtantsur | lucasagomes, at least first one should not | 09:34 |
| lucasagomes | yup yeah | 09:34 |
| lucasagomes | they have changes at common/ but won't conflict | 09:35 |
| lucasagomes | dtantsur, +a | 09:35 |
| dtantsur | thnx | 09:36 |
| *** jcoufal has quit IRC | 09:42 | |
| *** Poornima_ has quit IRC | 09:45 | |
| *** Haomeng has joined #openstack-ironic | 09:52 | |
| *** Haomeng|2 has quit IRC | 09:52 | |
| *** dtantsur is now known as dtantsur|lunch | 09:53 | |
| *** Poornima_ has joined #openstack-ironic | 09:59 | |
| openstackgerrit | Szymon Wróblewski proposed a change to openstack/ironic-python-agent: Enhance decorators in agent and automate creation of command_map https://review.openstack.org/117529 | 10:07 |
| openstackgerrit | A change was merged to openstack/ironic: Translator functions cleanup part 1 https://review.openstack.org/116303 | 10:17 |
| *** Haomeng has quit IRC | 10:23 | |
| *** Haomeng has joined #openstack-ironic | 10:24 | |
| *** pelix has joined #openstack-ironic | 10:30 | |
| *** vdrok_afk is now known as vdrok | 10:35 | |
| *** vinbs_ has joined #openstack-ironic | 10:42 | |
| *** vinbs has quit IRC | 10:43 | |
| *** vinbs_ is now known as vinbs | 10:43 | |
| *** jcoufal has joined #openstack-ironic | 10:47 | |
| *** andreykurilin has joined #openstack-ironic | 10:47 | |
| *** Haomeng|2 has joined #openstack-ironic | 10:55 | |
| *** saripurigopi has joined #openstack-ironic | 10:55 | |
| *** bluex-pl has quit IRC | 10:55 | |
| *** Haomeng has quit IRC | 10:55 | |
| *** bluex-pl has joined #openstack-ironic | 10:55 | |
| saripurigopi | I'm trying to load new driver. I've added the entry in setup.cfg and specified the driver name in enabled_drivers in ironic.conf file.But ironic-conductor is still failing to load. The same was working fine earlier juno versions. | 10:57 |
| saripurigopi | Is something changed recently? | 10:57 |
| *** bluex-pl has quit IRC | 11:01 | |
| *** bluex-pl has joined #openstack-ironic | 11:02 | |
| lucasagomes | saripurigopi, hmm nothing changed in that part afair :/ | 11:07 |
| *** bluex-pl has quit IRC | 11:08 | |
| lucasagomes | saripurigopi, was it re installed after the entry was added (python setup.py install) | 11:08 |
| *** bluex-pl has joined #openstack-ironic | 11:08 | |
| saripurigopi | lucasagomes, yes I've reinstalled after patching my changes. | 11:13 |
| openstackgerrit | Vladyslav Drok proposed a change to openstack/ironic: Use oslo.utils instead of ironic.openstack.common https://review.openstack.org/119732 | 11:13 |
| lucasagomes | saripurigopi, right hmm... off the top of my head I dunno what is going on then. Do you have the logs? it's a simply "loading driver X failed?" | 11:14 |
| saripurigopi | lucasagomes, 2014-08-08 13:40:11.283 14160 INFO ironic.common.driver_factory [-] Loaded the following drivers: ['fake', 'pxe_ssh', 'fake', 'pxe_ssh'] | 11:14 |
| saripurigopi | lucasagomes, it doesn't throw any error wrt new driver. | 11:15 |
| lucasagomes | hmm odd | 11:17 |
| *** Poornima_ has quit IRC | 11:19 | |
| lucasagomes | it should have throw an exception if driver failed to load https://github.com/openstack/ironic/blob/master/ironic/common/driver_factory.py#L120 | 11:19 |
| lucasagomes | saripurigopi, maybe adding a log on that _catch_driver_not_found to see if stevedore is calling that for you driver but we are actually doing something wrong with the conditionals there? | 11:19 |
| lucasagomes | your* | 11:20 |
| *** vinbs has quit IRC | 11:20 | |
| saripurigopi | lucasagomes, looks like re installing messed up the packages, it is not picking my changes. Could not load 'pxe_cisco': <module 'ironic.drivers.pxe' from '/usr/local/lib/python2.7/dist-packages/ironic/drivers/pxe.pyc'> has no 'PXEAndCiscoUCSMDriver' attribute | 11:21 |
| lucasagomes | saripurigopi, oh I see | 11:21 |
| lucasagomes | hmm | 11:21 |
| lucasagomes | right, I'll go for lunch before it gets too late here, let me know if you found the error later if not we can continue to investigate it | 11:24 |
| *** lucasagomes is now known as lucas-hungry | 11:24 | |
| saripurigopi | lucasagomes, sure thank you. | 11:25 |
| *** vinbs has joined #openstack-ironic | 11:27 | |
| *** saripurigopi has quit IRC | 11:33 | |
| *** vinbs has quit IRC | 11:41 | |
| *** bluex-pl has quit IRC | 12:02 | |
| *** enterprisedc has joined #openstack-ironic | 12:02 | |
| *** bluex-pl has joined #openstack-ironic | 12:09 | |
| *** enterprisedc has quit IRC | 12:10 | |
| *** enterprisedc has joined #openstack-ironic | 12:11 | |
| *** nosnos_ has quit IRC | 12:13 | |
| *** nosnos has joined #openstack-ironic | 12:14 | |
| *** enterprisedc has quit IRC | 12:15 | |
| *** enterprisedc has joined #openstack-ironic | 12:15 | |
| *** nosnos has quit IRC | 12:18 | |
| *** bmahalakshmi2 has quit IRC | 12:19 | |
| NobodyCam | woo hoo $5 landed :) | 12:22 |
| *** dtantsur|lunch is now known as dtantsur | 12:29 | |
| *** lucas-hungry is now known as lucasagomes | 12:30 | |
| openstackgerrit | Vladyslav Drok proposed a change to openstack/ironic: Use oslo.utils instead of ironic.openstack.common https://review.openstack.org/119732 | 12:30 |
| lucasagomes | NobodyCam, yup w00t! | 12:44 |
| lucasagomes | morning | 12:44 |
| Shrews | morning lucasagomes, NobodyCam, et.al. | 12:48 |
| lucasagomes | Shrews, morning :) | 12:48 |
| bluex-pl | hi | 12:49 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Fix race conditions running pxe_utils tests in parallel https://review.openstack.org/119746 | 12:50 |
| NobodyCam | morning lucasagomes :) | 12:53 |
| NobodyCam | morning Shrews | 12:54 |
| NobodyCam | morning dtantsur | 12:54 |
| NobodyCam | hehehe | 12:54 |
| * NobodyCam is almost awake | 12:54 | |
| bluex-pl | jroll: fixed cause of pep8 internal crash in https://review.openstack.org/#/c/117529 - period at the end of commit first line | 13:04 |
| bluex-pl | anyway, I would appreciate if you take a look at it | 13:05 |
| *** Poornima has joined #openstack-ironic | 13:08 | |
| *** rloo has joined #openstack-ironic | 13:08 | |
| *** linggao has joined #openstack-ironic | 13:15 | |
| *** rakesh_hs2 has quit IRC | 13:31 | |
| dtantsur | NobodyCam, morning! | 13:39 |
| openstackgerrit | A change was merged to openstack/ironic: Switch to "incubating" doc theme https://review.openstack.org/119576 | 13:42 |
| NobodyCam | :) | 13:45 |
| *** r-daneel has joined #openstack-ironic | 13:45 | |
| rloo | morning, afternoon Ironickers | 14:03 |
| *** bluex-pl has quit IRC | 14:04 | |
| Shrews | rloo: hi, morning | 14:05 |
| rloo | morning Shrews! | 14:05 |
| NobodyCam | morning rloo | 14:06 |
| lucasagomes | morning rloo | 14:06 |
| rloo | morning NobodyCam, afternoon lucasagomes! | 14:06 |
| NobodyCam | :) | 14:06 |
| *** chuckC has quit IRC | 14:10 | |
| openstackgerrit | Mike Durnosvistov proposed a change to openstack/ironic: Check whether specified FS is supported https://review.openstack.org/98102 | 14:13 |
| NobodyCam | lol watching 115540 fail / restart automaticly makes me really question the true value of the gate checks | 14:19 |
| *** dlaube has joined #openstack-ironic | 14:20 | |
| rloo | NobodyCam: I couldn't even figure out what the Hper-V CI failure meant. Did it fail and do we care? | 14:24 |
| NobodyCam | rloo I didn't actually check, the job keeps restarting with every patch that lands :-p | 14:27 |
| rloo | NobodyCam: ok. I wasn't sure if I should do a recheck or not cuz of that Hyper-V failure, but it looks like it is continuing so cross-my-fingers... | 14:28 |
| *** eghobo has joined #openstack-ironic | 14:30 | |
| NobodyCam | yep :) | 14:32 |
| jroll | morning ironic :) | 14:45 |
| jroll | lucasagomes: will try to look today :) | 14:45 |
| lucasagomes | jroll, morning! aight! thanks | 14:45 |
| NobodyCam | morning jroll | 14:46 |
| dtantsur | rloo, jroll, morning | 14:47 |
| rloo | hi dtantsur, morning jroll. | 14:48 |
| *** eghobo has quit IRC | 14:50 | |
| *** jcoufal has quit IRC | 14:56 | |
| Shrews | lucasagomes: great find on 119746. are you ok with merging with Yuriy's nit, or want to clean that up first? | 14:58 |
| Shrews | we do hardcode the node uuid there | 14:58 |
| *** chuckC has joined #openstack-ironic | 14:59 | |
| lucasagomes | Shrews, hey... so that nits are not related to the patch, I would include it tho if I needed another patch set | 14:59 |
| lucasagomes | I can do it now if you think it's important | 14:59 |
| rloo | lucasagomes: wrt https://review.openstack.org/#/c/119357/, have we discussed with tripleo folks. Is this blocking them? | 14:59 |
| Shrews | yeah, true. nah, not important i think | 14:59 |
| lucasagomes | rloo, yup, it breaks TripleO | 15:00 |
| lucasagomes | rloo, afaik derekh is reverting that bit in Ironic in the tripleO CI | 15:00 |
| rloo | lucasagomes: ok, so even if 119357 isn't approved, they're "ok" | 15:01 |
| lucasagomes | rloo, the tripleO CI yes, but tripleO per si will fail | 15:01 |
| lucasagomes | so I don't think they are OK | 15:01 |
| jroll | well | 15:01 |
| jroll | tripleo itself depends how the deployer configures things | 15:01 |
| Shrews | lucasagomes: +A'd | 15:01 |
| jroll | if the deployer uses x86_64 for everything, it won't break | 15:02 |
| jroll | (for example) | 15:02 |
| lucasagomes | Shrews, cheers | 15:02 |
| jroll | tripleo CI happens to use amd64, which is not canonicalized properly | 15:02 |
| jroll | (patches are up for that) | 15:02 |
| lucasagomes | jroll, rloo ^ yeah... problem is that they usually use things like amd64 | 15:02 |
| lucasagomes | jroll, nice! | 15:02 |
| jroll | I'm sure there is someone out there using 'donkeycpu' or something that will never work | 15:03 |
| lucasagomes | heh | 15:03 |
| rloo | jroll: so those patches to fix amd64 thing will fix their CI for them. but for people using tripleo, it could still break. | 15:03 |
| lucasagomes | jroll, https://review.openstack.org/#/c/117555/ merge failed | 15:03 |
| jroll | we should probably put this in the meeting agenda and grab derekh or lifeless for that meeting | 15:03 |
| jroll | ugh | 15:04 |
| jroll | thanks lucas | 15:04 |
| jroll | rloo: it could, yes | 15:04 |
| lucasagomes | np just saw it now | 15:04 |
| jroll | rloo: need docs... we fixed a bug (imho) that people were relying on | 15:04 |
| rloo | so the change we made isn't backwards compatible. hmm. | 15:04 |
| jroll | technically :| | 15:05 |
| jroll | so there's also a scheduler filter change | 15:05 |
| jroll | that starts canonicalizing things | 15:05 |
| jroll | that recently landed in nova | 15:05 |
| rloo | yeah, on the nova side. i looked before, and it seemed like ironic driver was the only one using it :-( | 15:05 |
| jroll | in which case we *have* to canonicalize the arch, or else things could break in a slightly different way | 15:06 |
| jroll | also, this isn't specific to tripleo, an existing ironic install could also break due to this | 15:06 |
| rloo | if other drivers in nova use the canonicalize thing, how are they going to be backwards compatible if it breaks their old usage. | 15:06 |
| rloo | jroll: yes, that's true. any ironic install. | 15:06 |
| rloo | if you fix a bug, does it have to be backwards compatible then? I don't think so? | 15:07 |
| rloo | we need to file a bug... | 15:07 |
| *** ifarkas has quit IRC | 15:08 | |
| jroll | right, idk if it needs to be backwards compat | 15:08 |
| jroll | like, windows has a bunch of bugs that they intentionally leave because devs depend on the buggy behavior | 15:08 |
| jroll | and I wouldn't want openstack to turn into that :( | 15:08 |
| jroll | hm, what if when InvalidArch is raised, we keep the arch we tried to canonicalize | 15:09 |
| rloo | jroll: yeah, I agree. is that something we need to bring up. I wonder if we need to start listing/working on docn. | 15:09 |
| jroll | rather than making it un-deployable | 15:09 |
| * jroll wonders how danpb would feel about that | 15:10 | |
| lucasagomes | jroll, sounds like a good plan | 15:10 |
| lucasagomes | yeah idk how the nova guys will react to that | 15:10 |
| *** jcoufal has joined #openstack-ironic | 15:11 | |
| dtantsur | anyway I see no point in reverting a patch in our code, which we're probably deleting in a couple of days | 15:13 |
| rloo | I was thinking that maybe we can't delete so soon. Do we need to give some X time for people to migrate? | 15:14 |
| *** bluex-pl has joined #openstack-ironic | 15:14 | |
| dtantsur | rloo, which people? those who use master will have a conflicting copy of code from nova | 15:14 |
| jroll | indeed | 15:15 |
| jroll | rloo: it's only a config option | 15:15 |
| jroll | that needs to be changed | 15:15 |
| rloo | true. master is fine. those that use j-1 or j-2? | 15:15 |
| jroll | (and deploy new nova etc) | 15:15 |
| jroll | rloo: same thing, no? | 15:15 |
| rloo | what am i saying? if they use j1 or j2, they'll have the driver there so no problem. | 15:15 |
| dtantsur | rloo, the same. actually everyone who'll be not careful enough, will get two versions of source code | 15:16 |
| rloo | good. we can get rid of it on master soon! | 15:16 |
| rloo | praying to jenkins... | 15:16 |
| jroll | dtantsur: it doesn't matter, really, because there's a config option that points at which driver to use | 15:16 |
| dtantsur | well right. anyway changing one config is the easiest upgrade path I've heard of :) upgrade from nova-bm will be harder | 15:17 |
| jroll | indeed | 15:18 |
| * jroll can't wait to do that upgrade, one less large dependency on the nova-compute hosts :) | 15:18 | |
| NobodyCam | so once all the driver patches land ... whos going to put up the delete patch for our repo | 15:20 |
| Shrews | not it | 15:20 |
| rloo | NobodyCam: mrda can do the honours! | 15:20 |
| jroll | I can! | 15:20 |
| jroll | I like free commits | 15:20 |
| jroll | oh yeah mrda totally should | 15:20 |
| jroll | he earned it | 15:20 |
| NobodyCam | :) | 15:20 |
| lucasagomes | yeah +1 mrda | 15:21 |
| rloo | (if he wants to) | 15:21 |
| jroll | oh, I'm sure he's ready to delete that code :P | 15:21 |
| NobodyCam | yes he did, thou every has been envoulved, and done a awesome job | 15:21 |
| NobodyCam | s/every/everyone/ | 15:23 |
| *** andreykurilin has quit IRC | 15:23 | |
| dtantsur | 117555 approved, I guess we can stop worring too much about canonicalization thing | 15:30 |
| jroll | hopefully | 15:31 |
| jroll | danprince said it was still giving him issues, idk | 15:31 |
| jroll | I told him to poke me | 15:31 |
| dtantsur | stepping afk before meeting | 15:39 |
| *** mikedillion has joined #openstack-ironic | 15:40 | |
| devananda | morning, all | 15:42 |
| jroll | mornin' devananda | 15:42 |
| lucasagomes | devananda, morning | 15:42 |
| NobodyCam | good morning devananda | 15:42 |
| rloo | morning devananda | 15:44 |
| dtantsur | devananda, morning | 15:45 |
| NobodyCam | one more check left for 115540 ( | 15:49 |
| NobodyCam | gate-tempest-dsvm-postgres-full:) | 15:49 |
| NobodyCam | a little quick on that .. | 15:50 |
| dtantsur | devananda and others: should we start retargeting driver bugs to nova? | 15:50 |
| *** bluex-pl has quit IRC | 15:50 | |
| NobodyCam | dtantsur: +1 sound about right to me | 15:51 |
| devananda | looks like patches 6 and 7 are still in teh merge queues? | 15:51 |
| lucasagomes | yeah | 15:52 |
| lucasagomes | loads of random failures | 15:52 |
| NobodyCam | yep watching them now | 15:52 |
| devananda | cool. let's prepare two more patches (not for nova) | 15:52 |
| devananda | 1. switch devstack config to use the driver from the nova tree | 15:52 |
| devananda | 2. delete the driver from ironic's tree | 15:52 |
| dtantsur | \o/ | 15:52 |
| lucasagomes | devananda, btw, graduation wise. Do we need something else? | 15:52 |
| lucasagomes | devananda, +1! we were talking about leaving 2) to mrda | 15:52 |
| devananda | lucasagomes: i'm going to go through that today -- our graduation review is *tomorrow* | 15:53 |
| NobodyCam | devananda: we were going to let mrda put up the delete patch | 15:53 |
| lucasagomes | devananda, uu I see | 15:53 |
| * lucasagomes fingers crossed | 15:53 | |
| Shrews | tomorrow? won't we draw attention away from apple's new iphone release??? | 15:53 |
| Shrews | lol | 15:53 |
| devananda | heh | 15:53 |
| NobodyCam | lol | 15:53 |
| devananda | dtantsur: can you do an audit of https://bugs.launchpad.net/ironic/+bugs?field.tag=nova-driver and make sure they're all current | 15:54 |
| dtantsur | devananda, will do | 15:54 |
| devananda | that should, in theory, be the ones that will need retargetign to nova | 15:54 |
| devananda | eg, https://review.openstack.org/#/c/115540/ appears related to https://bugs.launchpad.net/ironic/+bug/1289048 but is not tagged | 15:55 |
| devananda | (that just jumped out at me) | 15:55 |
| *** viktors is now known as viktors|afk | 16:01 | |
| Shrews | stepping out for lunch. bbiab | 16:02 |
| NobodyCam | enjoy Shrews | 16:03 |
| *** mikedillion has quit IRC | 16:04 | |
| *** pcrews has joined #openstack-ironic | 16:05 | |
| *** ChuckC_ has joined #openstack-ironic | 16:07 | |
| *** chuckC has quit IRC | 16:08 | |
| NobodyCam | brb | 16:08 |
| linggao | Hi jroll | 16:09 |
| *** mikedillion has joined #openstack-ironic | 16:09 | |
| devananda | adam_g: what's the status of the sideways upgrade testing? | 16:10 |
| *** rushiagr is now known as rushiagr_away | 16:14 | |
| devananda | russell_h: was someone on your team working with Horizon folks? How's that going? | 16:17 |
| russell_h | devananda: yeah, JoshNang and aweeks have been working on that | 16:17 |
| russell_h | I had understood it to be mostly working | 16:18 |
| JoshNang | devananda: i've got the api portion finished and got a LGTM from a horizon core. https://review.openstack.org/#/c/117376/ | 16:18 |
| JoshNang | aweeks is working on the frontend/panel part. not sure if a patchset is up yet | 16:19 |
| devananda | great ,thanks! | 16:19 |
| devananda | please let me know if so -- i'm referencing all of these in teh graduation review | 16:19 |
| JayF | When is the graduation review? | 16:19 |
| jroll | linggao: hi! | 16:20 |
| JoshNang | will do | 16:20 |
| jroll | JayF: tomorrow | 16:20 |
| * jroll forgets what time | 16:20 | |
| JayF | Yeah I was going to make a calendar entry for it to make sure I didn't miss it:) | 16:20 |
| JoshNang | ahh. i'll make sure at least something is up for horizon panels today | 16:20 |
| jroll | it's TC meeting, whenever that is | 16:20 |
| devananda | 1pm pacific, tomorrow | 16:20 |
| jroll | thanks | 16:21 |
| linggao | Does deployment with IPA work now on the trunk? | 16:21 |
| devananda | fwiw, making my notes here: https://etherpad.openstack.org/p/IronicGraduationDiscussion | 16:21 |
| *** eghobo has joined #openstack-ironic | 16:21 | |
| jroll | linggao: yes | 16:21 |
| linggao | Is there a doc on how to set it up? | 16:21 |
| jroll | linggao: (or, it did, I haven't tested recently and we *still* don't have CI) | 16:21 |
| jroll | linggao: it's in ironic's devstack docs | 16:22 |
| *** jistr has quit IRC | 16:22 | |
| linggao | I installed a clound with devstack following "Deploying Ironic with DevStack" in Developer Quick-Start | 16:23 |
| jroll | ok, there are some agent-specific bits there | 16:23 |
| jroll | just localrc things | 16:23 |
| linggao | In /etc/ironic/ironic.conf it has enabled_drivers = fake,agent_ssh,agent_ipmitool | 16:23 |
| linggao | Is that the right doc I should follow? | 16:24 |
| *** penick has joined #openstack-ironic | 16:24 | |
| jroll | yes | 16:24 |
| linggao | http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#deploying-ironic-with-devstack | 16:24 |
| jroll | linggao: yes, that's it, see the part in localrc "# If running with the agent driver:" | 16:25 |
| *** romcheg1 has joined #openstack-ironic | 16:26 | |
| NobodyCam | devananda: are you adding anything about the proxy api commands stuff? | 16:26 |
| linggao | jroll, yes, I did that part. | 16:26 |
| jroll | linggao: great, now run nova boot! | 16:26 |
| linggao | But the nova boot part failed. | 16:26 |
| jroll | linggao: which image did you boot with? | 16:26 |
| NobodyCam | 115540 Merged | 16:27 |
| devananda | NobodyCam: that's more detail that the TC will likely want, but it may come up | 16:27 |
| NobodyCam | devananda: fyi: https://review.openstack.org/#/c/116316/ | 16:27 |
| linggao | jroll, I was just following the instruction, it complains about the flavor baremetal does not exist. | 16:27 |
| jroll | uhhhh | 16:28 |
| devananda | NobodyCam: ack, ty. fyi, the deprecate-baremetal spec wasn't approved yet | 16:28 |
| jroll | linggao: that's interesting | 16:28 |
| linggao | jroll, the image is cirros-0.3.2-x86_64-disk | 16:28 |
| jroll | linggao: ok, that's right | 16:28 |
| *** mikedillion has quit IRC | 16:28 | |
| jroll | weird that the baremetal flavor doesn't exist | 16:28 |
| devananda | NobodyCam: there's some internal disagreements within nova, it seems. or they just forgot to update their comments on the spe | 16:28 |
| devananda | spec | 16:28 |
| linggao | jroll, what does the baremetal flavor look like? I can add it manually. | 16:29 |
| devananda | NobodyCam: https://review.openstack.org/#/c/95025/ | 16:29 |
| devananda | NobodyCam: at this point, the actual deprecation work is almost definitely going to be bumped to Kilo, which is weird, but what ever | 16:30 |
| linggao | jroll, I also remember you said IPA uses coreOS as the deploy_kernel and deploy_ramdisk. I do not see them in the image table. | 16:30 |
| *** mikedillion has joined #openstack-ironic | 16:30 | |
| jroll | linggao: something like this: https://github.com/openstack-dev/devstack/blob/master/lib/ironic#L498-505 | 16:30 |
| jroll | linggao: hmm | 16:31 |
| jroll | linggao: they should just be called ir-deploy.* | 16:31 |
| linggao | jroll, here is my image table: | 16:31 |
| linggao | # nova image-list | 16:31 |
| linggao | +--------------------------------------+---------------------------------+--------+--------+ | 16:31 |
| linggao | | ID | Name | Status | Server | | 16:31 |
| linggao | +--------------------------------------+---------------------------------+--------+--------+ | 16:31 |
| linggao | | eab26b35-38ab-4aae-aecc-23c077108497 | Fedora-x86_64-20-20140618-sda | ACTIVE | | | 16:31 |
| linggao | | 406b3179-d32a-4262-ad20-b06e1625739b | cirros-0.3.2-x86_64-disk | ACTIVE | | | 16:31 |
| linggao | | c5492218-e665-4dfa-bd04-51b19568efae | cirros-0.3.2-x86_64-uec | ACTIVE | | | 16:31 |
| linggao | | 54c03f52-83df-4a8d-9f60-64cb7286a97d | cirros-0.3.2-x86_64-uec-kernel | ACTIVE | | | 16:31 |
| linggao | | 4db2f7bd-0f1b-4b03-8db0-b14e183b59d9 | cirros-0.3.2-x86_64-uec-ramdisk | ACTIVE | | | 16:31 |
| linggao | +--------------------------------------+---------------------------------+--------+--------+ | 16:31 |
| linggao | jroll, that's all. | 16:31 |
| NobodyCam | lol nice... well we have it if they want it | 16:31 |
| jroll | please you paste.openstack.org or similar :/ | 16:31 |
| JayF | jroll: I know what broke | 16:31 |
| jroll | huh. | 16:31 |
| devananda | linggao: please use a pastebin, and do not spam the channel with large pastes | 16:31 |
| jroll | JayF: what broke | 16:31 |
| JayF | jroll: linggao: We migrated Devstack to look for the individually published files, not a tarball | 16:32 |
| linggao | ooops, sorry devananda | 16:32 |
| JayF | and nothing in the agent has landed since then to publish the individual files | 16:32 |
| jroll | ugh | 16:32 |
| * jroll proposes something silly | 16:32 | |
| JayF | I have probably non-silly changes I can make | 16:32 |
| JayF | like updating the document to point to the individual files instead of the tarball | 16:32 |
| openstackgerrit | Jim Rollenhagen proposed a change to openstack/ironic-python-agent: Change rebuilding to building in README https://review.openstack.org/119805 | 16:33 |
| jroll | JayF: +a that, pleeeease | 16:33 |
| jroll | that would work too | 16:33 |
| jroll | JayF: surprised it didn't error somewhere :| | 16:34 |
| linggao | jroll, JayF, I am confused here. | 16:34 |
| NobodyCam | 115549 Merged that's all seven!!!!! | 16:34 |
| JoshNang | \o/ | 16:34 |
| jroll | linggao: there's a broken thing | 16:34 |
| linggao | jroll, in devstack? | 16:35 |
| jroll | linggao: technically it's an IPA problem | 16:36 |
| openstackgerrit | Jay Faulkner proposed a change to openstack/ironic-python-agent: Fix incorrect documentation https://review.openstack.org/119806 | 16:36 |
| JayF | we're fixing it right meow | 16:36 |
| JayF | jroll: ^ no, you +2A that! | 16:36 |
| JayF | :P | 16:36 |
| jroll | linggao: in devstack, you can run "git checkout 5ecb1ef86ae49272d43db0dff2d619b211d439ea" to roll back for now | 16:36 |
| JayF | jroll: isn't that broken too? | 16:36 |
| JayF | jroll: because pathing? | 16:37 |
| jroll | linggao: we publish files after merge | 16:37 |
| jroll | JayF: oh, you're right | 16:37 |
| kylestev | JayF: nice one :P | 16:37 |
| JayF | jroll: I'd suggest directing her to download and put the files in the tarball in the right place | 16:37 |
| JayF | jroll: or merge my patch :) | 16:37 |
| JayF | quick merge to unbreak IPA in devstack (we need the post job to run) --> https://review.openstack.org/#/c/119806/ | 16:37 |
| jroll | JayF: +A'd | 16:37 |
| linggao | jroll, JayF, now what am I supposed to do? | 16:38 |
| jroll | linggao: need to wait until that patch is merged and the post jobs run, sorry :/ | 16:39 |
| jroll | luckily IPA gate is much faster | 16:39 |
| jroll | (for now) | 16:39 |
| linggao | jroll, that patch has only 2 readme file changed. | 16:41 |
| jroll | linggao: after any IPA patch merges, we build the ramdisk and upload it | 16:42 |
| jroll | linggao: devstack pulls the ramdisk down and loads it into glance | 16:42 |
| jroll | linggao: and we recently changed which files are uploaded, so devstack is looking for files that haven't been uploaded yet | 16:42 |
| jroll | (tis my fault0 | 16:42 |
| jroll | ) | 16:42 |
| NobodyCam | I've lost track, whats the status of the migration scripts? | 16:43 |
| jroll | it landed afaik | 16:44 |
| linggao | jroll, okay, thanks for the help. I'll switch to pxe_ipminative driver for now. Is it just as simple of change the conf file or I need to reinstall the stack? | 16:44 |
| jroll | flavor update script did not, but I think there was discussion if that was needed | 16:44 |
| jroll | linggao: I would restack | 16:44 |
| jroll | linggao: if you're testing with VMs as fake baremetal servers, ipmi won't work | 16:45 |
| linggao | jroll, so for Juno IPA will replace pxe driver or they co-exist? | 16:45 |
| jroll | co-exist | 16:45 |
| jroll | pxe is still default | 16:45 |
| jroll | I have a goal for kilo to make agent the default, we shall see how that goes :) | 16:45 |
| linggao | jroll, I plan to test with real machines, thanks for the help. | 16:46 |
| NobodyCam | call for agenda updates! | 16:47 |
| jroll | linggao: no problem :) | 16:47 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Updated from global requirements https://review.openstack.org/119809 | 16:49 |
| *** ndipanov has quit IRC | 16:49 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Update ironic.conf.sample https://review.openstack.org/119816 | 16:53 |
| lucasagomes | pep8 gate is broken due ^ | 16:54 |
| lucasagomes | FYI http://logs.openstack.org/46/119746/1/gate/gate-ironic-pep8/4a54d56/console.html | 16:55 |
| jroll | :/ | 16:55 |
| jroll | lucasagomes: should we pin to the latest oslo.messaging, then? | 16:55 |
| lucasagomes | jroll, we have to go according with the global requirements | 16:55 |
| NobodyCam | nice chatch lucasagomes | 16:55 |
| lucasagomes | so there's nothing we can do really | 16:55 |
| NobodyCam | catch even | 16:55 |
| lucasagomes | devananda, ^ how other projects deal with that? | 16:56 |
| lucasagomes | I think some of then just do not check the sample in gate | 16:56 |
| jroll | yeah | 16:56 |
| jroll | maybe time to remove the sample :P | 16:56 |
| JayF | I think the proposed patch to get rid of the sample config althogether is a better fix | 16:56 |
| JayF | we should generate it in the post job if possible? | 16:56 |
| lucasagomes | JayF, yeah well if we decide to do it | 16:56 |
| rloo | thx lucasagomes. I was in the middle of opening a bug about it, but I'm fine with a patch only ;) | 16:56 |
| jroll | JayF: idk if post jobs can commit though | 16:56 |
| NobodyCam | bbt brb | 16:56 |
| JayF | if it's entirely a generated file, having humans keep it in sync, especially when a computer cna break it, seems silly | 16:57 |
| lucasagomes | rloo, yeah it just conflicted with the one fixing the race condition I had | 16:57 |
| jroll | NobodyCam: I have to ask since I've always been curious, what is bbt? | 16:57 |
| lucasagomes | JayF, yeah... it wasn't a problem before because the oslo code was in our tree | 16:57 |
| lucasagomes | not it's not | 16:57 |
| lucasagomes | now* | 16:57 |
| lucasagomes | but yeah we have to think about a nice way to fix that | 16:57 |
| lucasagomes | maybe having a jenkins job that syncs it time to time | 16:58 |
| lucasagomes | just like we have for requirements and i18n | 16:58 |
| rloo | i like seeing the .sample when reviewing code that adds configs, so I can see what it looks like. | 16:58 |
| jroll | we could also have a post job that uploads it to the docs site or something | 16:58 |
| lucasagomes | right, yeah that's quite nice | 16:58 |
| rloo | jroll: I think bbt == bubbies break time. (I forgot the name of the dogs but something like that) | 16:58 |
| jroll | and have a preview | 16:59 |
| jroll | rloo: ah! :) | 16:59 |
| lucasagomes | let's talk about it in the meeting today? | 16:59 |
| lucasagomes | but fix that soon so we can move on with the gate | 16:59 |
| * lucasagomes will add to the calendar | 16:59 | |
| devananda | lucasagomes: nova stopped checking the sample in the gate during Juno because of this | 16:59 |
| jroll | yeah | 16:59 |
| devananda | and now actually has no sample config *at all* | 16:59 |
| lucasagomes | devananda, oh hmm | 16:59 |
| lucasagomes | idk if I'd like that | 16:59 |
| devananda | in their tree. instead, nova requries users to generate sample.conf on demand if they want to see it | 17:00 |
| jroll | can always generate it if needed | 17:00 |
| devananda | which takes quite a while, at least for me | 17:00 |
| jroll | yeah | 17:00 |
| JayF | devananda: is it possible for the ironic.conf.sample to be *committed* to the git repo in a post job? | 17:00 |
| devananda | JayF: committed - no. proposed - yes. | 17:00 |
| devananda | well. | 17:00 |
| JayF | Having it committed back in post seems glorious to me | 17:00 |
| devananda | i may be wrong. infra knows magic. | 17:00 |
| * lucasagomes do not add to the meeting since we are discussing it now | 17:00 | |
| NobodyCam | Haomeng put this up for us: https://review.openstack.org/#/c/106493/ | 17:00 |
| *** derekh has quit IRC | 17:00 | |
| JayF | that's like the A+ solution to the problem, right? | 17:00 |
| devananda | JayF: sounds fine to me. want to follow up with infra? | 17:01 |
| JayF | I' | 17:01 |
| JayF | I can try :) | 17:01 |
| *** MattMan has quit IRC | 17:03 | |
| *** vdrok_ has joined #openstack-ironic | 17:04 | |
| *** Poornima has quit IRC | 17:04 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/python-ironicclient: List resources with detail https://review.openstack.org/119825 | 17:05 |
| lucasagomes | alright I will be back for the meeting later on | 17:06 |
| *** lucasagomes is now known as lucas-brb | 17:06 | |
| *** rushiagr_away is now known as rushiagr | 17:06 | |
| Shrews | rloo: jroll: i thought it was "breakfast begal time" :) | 17:07 |
| NobodyCam | lol rloo had it right. Bubbie Breakfest time :) | 17:10 |
| NobodyCam | breakfast even | 17:10 |
| JayF | Infra basically said what I expected; we can propose changes back but not commit directly | 17:10 |
| JayF | they suggest going to a model of: 1) Don't test config matching and/or 2) Remove config completely from git, Nova-style | 17:11 |
| devananda | JayF: ack. sounds like 1 + a post job would meet our desires | 17:11 |
| jroll | NobodyCam: lol | 17:11 |
| devananda | adam_g: around? | 17:12 |
| JayF | devananda: I forsee us having quite a few additional 'fix config' patches if we go that route | 17:12 |
| JayF | devananda: given how difficult it can be to land things in the integrated gate, I question if that's a good path to go down | 17:12 |
| devananda | JayF: ah. integrated gate. | 17:13 |
| JayF | As much as I like having the conf sample in fit, it seems more valuable to reduce brokenness... that to me lends itself to the nova path | 17:13 |
| JayF | but we could add something of our own like generating it and publishing it in the docs or some other place, so it's still viewable by those who don't have working ironic dev environments | 17:14 |
| *** harlowja_away is now known as harlowja | 17:14 | |
| Shrews | devananda: adam_g posted this over the weekend: <adam_g> Shrews, post-grenade, parallel tempest run /w isolated credentials \o/ http://logs.openstack.org/15/118515/15/experimental/check-grenade-dsvm-ironic-sideways/d581408/logs/testr_results.html.gz | 17:15 |
| devananda | Shrews: woot! | 17:16 |
| openstackgerrit | A change was merged to openstack/ironic-python-agent: Fix incorrect documentation https://review.openstack.org/119806 | 17:24 |
| jroll | ^ JayF | 17:25 |
| * jroll watches post jobs | 17:25 | |
| adam_g | devananda, hey | 17:27 |
| dtantsur | lucas-brb, please see my comment for 119825 tl;dr we have a huge overlap :( | 17:27 |
| kylestev | jroll: JayF \o/ | 17:27 |
| devananda | adam_g: hi there! got an update from shrews above. Also, will you be able to make the TC meeting tomorrow? | 17:28 |
| devananda | adam_g: in case they raise concerns about our tempest coverage, I'd like you to be around. if not, let's chat today so I can prepare | 17:28 |
| adam_g | devananda, yeah, i should be able to make that | 17:29 |
| adam_g | 2000 UTC? | 17:29 |
| devananda | sounds right. an hour later than our meeting (but on tuesday) | 17:29 |
| adam_g | hmm actually i have appointment at 1PM on tuesdays. ill see if i can move it an hour | 17:30 |
| JayF | jroll: we're still queued :( | 17:33 |
| jroll | JayF: yeah | 17:34 |
| *** vdrok_ has quit IRC | 17:35 | |
| devananda | ooh. looks like all the noav driver patches just finished merging! | 17:44 |
| dtantsur | \o/ | 17:45 |
| * devananda proposes https://review.openstack.org/#/c/119844/ | 17:48 | |
| NobodyCam | :) | 17:49 |
| NobodyCam | devananda: +1'd | 17:49 |
| NobodyCam | :-p | 17:49 |
| jroll | ooo, agent_ssh is broken :) | 17:57 |
| jroll | lovely | 17:58 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic-specs: Support Decommissioning of Nodes https://review.openstack.org/102685 | 17:59 |
| NobodyCam | :( | 17:59 |
| NobodyCam | JoshNang: is that for kilo? | 18:00 |
| jroll | juno! | 18:00 |
| * jroll applies for ffe | 18:00 | |
| JoshNang | NobodyCam: yeah, i didn't switch the directories yet :) | 18:00 |
| NobodyCam | :) | 18:00 |
| JayF | jroll: how is it broken? | 18:01 |
| jroll | JayF: pxe template is borked | 18:01 |
| jroll | I'll fix | 18:01 |
| JayF | ah :(. I bet it happened during the ilo ipa patch | 18:02 |
| jroll | it did | 18:02 |
| jroll | I was wary of it, and then for whatever reason decided it should work | 18:02 |
| NobodyCam | what borked? | 18:02 |
| jroll | NobodyCam: {{ options.ipa-api-url }} | 18:02 |
| jroll | isn't valid | 18:02 |
| jroll | it does options.ipa - api - url | 18:03 |
| NobodyCam | ipa_api_url? | 18:03 |
| jroll | right... but he changed it | 18:03 |
| NobodyCam | ahh | 18:03 |
| jroll | for valid reasons, of course | 18:03 |
| jroll | but now I have to fix for both of us :( | 18:03 |
| NobodyCam | :( | 18:03 |
| JayF | I'll take Agent CI for $100, Alex | 18:03 |
| JayF | "This thing would've prevented this bad patch from breaking agent_ssh" | 18:04 |
| JayF | jroll you're the first to the buzzer | 18:04 |
| NobodyCam | please answer in the form of a question | 18:04 |
| jroll | what is reading the docs? | 18:04 |
| NobodyCam | lol | 18:04 |
| JayF | ooh, I'm sorry, that's the answer to every other question on the board except that one | 18:05 |
| JayF | lol | 18:05 |
| JayF | the correct response was "What is IPA CI?" | 18:05 |
| jroll | that too | 18:05 |
| jroll | JayF: I'm not going to have these jobs vote for the first run, even in IPA | 18:05 |
| jroll | because I'm not confident right now | 18:05 |
| JayF | Obviously :D | 18:05 |
| JayF | jroll: https://jenkins01.openstack.org/job/ironic-python-agent-buildimage-coreos/4/ | 18:06 |
| NobodyCam | http://www.sushoo.com/Entities/IPA-Clinical-Integration.html ?????? | 18:06 |
| jroll | JayF: whee. | 18:06 |
| JayF | jroll: 2014-09-08 18:07:28.946 | [SCP] ‘imagebuild/coreos/UPLOAD/coreos_production_pxe-oem.cpio.gz’ doesn’t match anything: ‘imagebuild’ exists but not ‘imagebuild/coreos/UPLOAD/coreos_production_pxe-oem.cpio.gz’ | 18:07 |
| JayF | we did not pop that error for the vmlinuz | 18:08 |
| * JayF looks for the typo | 18:08 | |
| JayF | coreos_production_pxe_image-oem.cpio.gz | 18:08 |
| JayF | ugh | 18:08 |
| jroll | o.o | 18:09 |
| jroll | I'm sorry | 18:09 |
| jroll | I thought I copy/pasted that | 18:09 |
| JayF | that one is me man | 18:09 |
| jroll | oh wait, I did the devstack side | 18:09 |
| * jroll uh oh | 18:09 | |
| JayF | I wrote the openstack-infra/config | 18:09 |
| jroll | I got it right | 18:10 |
| * jroll yay | 18:10 | |
| JayF | jroll: https://review.openstack.org/119853 | 18:11 |
| jroll | ECANNOT+2 | 18:12 |
| *** penick has quit IRC | 18:13 | |
| *** penick has joined #openstack-ironic | 18:13 | |
| NobodyCam | TY mordred !! :) | 18:14 |
| jroll | woo | 18:14 |
| JayF | now jroll you have about 15m to wait for a puppet run | 18:14 |
| *** jcoufal has quit IRC | 18:14 | |
| JayF | enough time to find more doc mistakes | 18:14 |
| jroll | lol | 18:14 |
| JayF | jroll: protip: I copied the URL for the IPA ramdisk from the infra/config | 18:14 |
| JayF | jroll: so it's likely wrong :P | 18:15 |
| JayF | as far as a doc mistake you can merge in | 18:15 |
| JayF | lol | 18:15 |
| jroll | can you look for something? | 18:15 |
| jroll | I'm fixing our driver :P | 18:15 |
| JayF | I see, it's your driver when it works, and our driver when it's broken | 18:15 |
| JayF | typical! :P | 18:15 |
| kylestev | :P | 18:17 |
| jroll | lol | 18:18 |
| NobodyCam | JayF: at least it's not "YOUR" driver when it dosen't work | 18:18 |
| openstackgerrit | Jay Faulkner proposed a change to openstack/ironic-python-agent: Fix minor documentation issues https://review.openstack.org/119859 | 18:19 |
| NobodyCam | JayF: so I can have a config iso? | 18:21 |
| JayF | NobodyCam: the iLO IPA driver boots IPA off an ISO | 18:21 |
| NobodyCam | with a config floppy? | 18:21 |
| JayF | NobodyCam: and embeds configuration in a virtual config floppy iirc | 18:21 |
| jroll | yep | 18:22 |
| JayF | I don't know why you'd ever want to PXE the agent but still interact with virtual media for configuration | 18:22 |
| NobodyCam | booting and configuring IPA via virtual media in52 | 18:22 |
| NobodyCam | the form of an ISO image | 18:22 |
| JayF | If you have the ability to do so, remove PXE from the equation altother | 18:22 |
| NobodyCam | seemed a little unclear about the floppy | 18:22 |
| NobodyCam | but would just be a nit | 18:23 |
| JayF | I'll s/and configuring/ | 18:23 |
| JayF | just remove it | 18:23 |
| NobodyCam | :) | 18:23 |
| openstackgerrit | Jay Faulkner proposed a change to openstack/ironic-python-agent: Fix minor documentation issues https://review.openstack.org/119859 | 18:23 |
| *** eghobo has quit IRC | 18:28 | |
| *** eghobo has joined #openstack-ironic | 18:29 | |
| dtantsur | I was approved for travel support program for Paris Summit \o/ | 18:30 |
| rloo | congrats dtantsur!!! | 18:30 |
| JayF | Awesome, will be glad to have you there | 18:30 |
| dtantsur | not to decide about dates: on what days are design sessions? | 18:31 |
| JayF | This year, supposedly, Friday is not a throwaway day | 18:32 |
| JayF | so don't leave too early :) | 18:32 |
| JayF | I'm in for the full five days | 18:32 |
| *** rushiagr is now known as rushiagr_away | 18:32 | |
| dtantsur | JayF, you mean, arrive on Mon, leave on Sat, right? | 18:33 |
| JayF | I arrive on Sun, leave on Saty | 18:33 |
| JayF | actually, I may be Sat-Sat | 18:33 |
| JayF | I purposefully made sure I'd have at least a day to explore Paris :) | 18:33 |
| dtantsur | :) | 18:34 |
| dtantsur | but I guess monday can be skipped | 18:35 |
| NobodyCam | do we have CLI docs? | 18:35 |
| *** penick has quit IRC | 18:36 | |
| rloo | NobodyCam: sort of. | 18:37 |
| rloo | NobodyCam: can't remember where. and not very useful. | 18:37 |
| jroll | dtantsur: monday talks might be fun, though | 18:37 |
| jroll | :) | 18:38 |
| dtantsur | hmm ok | 18:38 |
| JayF | If mine (+joshnang) talk is on Monday, and you miss it, I'll weep a river of tears. | 18:38 |
| jroll | he will | 18:38 |
| jroll | I've seen it | 18:38 |
| JayF | the punishment for missing jroll's talk is less obvious | 18:39 |
| rloo | NobodyCam: do you mean http://docs.openstack.org/developer/python-ironicclient/ | 18:39 |
| JayF | you just get to learn all the scaling lessons the hard way :P | 18:39 |
| jroll | muahaha | 18:39 |
| JoshNang | JayF: I think we're like 11am Wed | 18:41 |
| JayF | We've been scheduled? | 18:41 |
| jroll | I haven't seen a schedule | 18:41 |
| JayF | the schedule is published | 18:41 |
| JayF | Wed, 11am for me and Josh | 18:41 |
| JoshNang | https://openstacksummitnovember2014paris.sched.org/ | 18:42 |
| NobodyCam | rloo: yes I did ... Thank you :) | 18:42 |
| JayF | jroll: is tues, 5:30p | 18:42 |
| rloo | NobodyCam: yw ;) | 18:42 |
| jroll | JayF: oh, neat | 18:43 |
| openstackgerrit | Ruby Loo proposed a change to openstack/ironic: Remove XML from api doc samples https://review.openstack.org/119868 | 18:44 |
| NobodyCam | devananda: would not this count for cli docs? http://docs.openstack.org/developer/python-ironicclient/cli.html | 18:44 |
| openstackgerrit | Jim Rollenhagen proposed a change to openstack/ironic: Fix agent PXE template https://review.openstack.org/119869 | 18:44 |
| openstackgerrit | Jim Rollenhagen proposed a change to openstack/ironic: Use expected, actual order for PXE template test https://review.openstack.org/119870 | 18:44 |
| jroll | ^ fixes agent driver | 18:44 |
| jroll | adam_g: is this new? self.verify_partition(vm_client, 'ephemeral0', '/mnt', eph_size) | 18:45 |
| jroll | adam_g: (in tempest baremetal scenarios) | 18:45 |
| *** romcheg1 has quit IRC | 18:45 | |
| devananda | jroll: that was proposed months ago by lucas. | 18:45 |
| *** romcheg1 has joined #openstack-ironic | 18:46 | |
| jroll | huh | 18:46 |
| jroll | it's now failing for the agent | 18:46 |
| *** romcheg1 has quit IRC | 18:46 | |
| jroll | I never saw it fail there before | 18:46 |
| devananda | ya'll dont support ephemeral partitions? | 18:46 |
| * jroll hides | 18:46 | |
| Shrews | jroll: i added that | 18:46 |
| *** romcheg1 has joined #openstack-ironic | 18:46 | |
| adam_g | i thought that was part of the rebuild testing Shrews added | 18:46 |
| JayF | IPA still only supports full-disk iamges. | 18:46 |
| devananda | Shrews: s/lucas/shrews/ :) | 18:46 |
| jroll | gah | 18:47 |
| jroll | any way I can like... skip that | 18:47 |
| jroll | or somethign? | 18:47 |
| *** penick has joined #openstack-ironic | 18:47 | |
| devananda | jroll: not supporting rebuild --preserve-ephemeral means tripleo can't use IPA | 18:48 |
| *** romcheg1 has joined #openstack-ironic | 18:49 | |
| jroll | devananda: I'm aware | 18:49 |
| *** mikedillion has quit IRC | 18:49 | |
| jroll | devananda: my goal is feature parity in k | 18:49 |
| jroll | meh, this puts a huge dent in my CI plans :/ | 18:50 |
| *** openstackgerrit has quit IRC | 18:51 | |
| devananda | jroll: looks like it merged to tempest on Aug 15 | 18:51 |
| NobodyCam | 9 minutes until meeting | 18:51 |
| Shrews | jroll: it's now part of the basic test. you'd either have to skip the test (not recommended) or code around it or support ephemeral partitions | 18:51 |
| NobodyCam | last chance for coffee refills | 18:51 |
| devananda | Shrews: is there a feature flag for that? | 18:51 |
| Shrews | gah. sorry. SEVERE upload lag while i was pushing a large image somewhere | 18:52 |
| adam_g | could you skip that ephemeral stuff if the configured flavor's ephemeral == 0? | 18:52 |
| Shrews | devananda: jroll: there is a 'rebuild' compute-feature flag | 18:52 |
| jroll | Shrews: where/how do I set that? | 18:52 |
| Shrews | jroll: in devstack-gate config, iirc | 18:53 |
| * jroll sighs | 18:53 | |
| Shrews | jroll: oh, that might be in devstack/lib/ironic, actually | 18:53 |
| Shrews | an iniset command | 18:54 |
| jroll | hmm | 18:54 |
| jroll | idk, I'll look after meeting I geuss | 18:54 |
| Shrews | but then the entire scenario test will be skipped | 18:54 |
| jroll | yeah | 18:54 |
| jroll | so then we aren't actually testing anything, right? | 18:54 |
| Shrews | we won't be testing spinning up an instance | 18:54 |
| jroll | right | 18:55 |
| jroll | which defeats the purpose | 18:55 |
| Shrews | which, ya know, we probably should | 18:55 |
| jroll | ugh, would love a way to skip the rebuild bit | 18:55 |
| russell_h | that feels more like a tempest bug than anything | 18:55 |
| jroll | it's more an assumption in ironic's tempest code that all drivers support ephemeral disks | 18:56 |
| jroll | which is no longer true | 18:56 |
| Shrews | jroll: i'd modify the scenario test to skip the verification if ephemeral size is 0 | 18:56 |
| jroll | ok | 18:56 |
| jroll | but, doesn't the scenario test set the ephemeral size? :) | 18:56 |
| *** lucas-brb is now known as lucasagomes | 18:56 | |
| lucasagomes | dtantsur, ack | 18:56 |
| lucasagomes | dtantsur, I can rebase on top of that | 18:56 |
| dtantsur | yes please | 18:56 |
| *** wendar_ is now known as wendar | 18:58 | |
| devananda | Shrews: test_baremetal_server_ops looks like it will fail with ephemeral == 0 | 18:58 |
| Shrews | well, i'm tossing out random ideas. been a while since i looked at that code, tbh | 18:59 |
| mrda | Morning Ironic! | 18:59 |
| NobodyCam | morning mrda | 18:59 |
| lucasagomes | mrda, morning | 19:00 |
| jroll | yeah, I'll poke after the meeting | 19:00 |
| dtantsur | meeting time | 19:00 |
| jroll | I can't do this no CI thing any more :( | 19:00 |
| devananda | it currently asserts that it is not "N/A" but a size of 0 will pass the tempest assertion and actually not get created by ironic deploy_utils | 19:01 |
| devananda | because "if ephemeral_mb:" will be faulse | 19:01 |
| *** mikedillion has joined #openstack-ironic | 19:04 | |
| *** openstackgerrit has joined #openstack-ironic | 19:08 | |
| greghaynes | Shrews: Hey, thanks for the comments. I have a Q about https://review.openstack.org/#/c/109688/10/ironic/tests/api/test_acl.py | 19:08 |
| *** athomas has quit IRC | 19:08 | |
| greghaynes | Basically, since im mocking self.dbapi its pretty nasty to do in a decorator | 19:08 |
| greghaynes | seems like ctxt manager might actually better there | 19:08 |
| Shrews | greghaynes: hi. oh, hrm. will look again after the ironic meeting | 19:09 |
| greghaynes | ok | 19:09 |
| * greghaynes joins meeting | 19:09 | |
| *** bmahalakshmi has joined #openstack-ironic | 19:10 | |
| *** jcoufal has joined #openstack-ironic | 19:16 | |
| *** bmahalakshmi has quit IRC | 19:16 | |
| *** bmahalakshmi has joined #openstack-ironic | 19:18 | |
| *** andreykurilin has joined #openstack-ironic | 19:22 | |
| devananda | jroll: reviewed https://review.openstack.org/#/c/119869/ | 19:39 |
| jroll | devananda: I'm just doing what the pxe driver does :) | 19:39 |
| jroll | devananda: that's the test template | 19:39 |
| * jroll will reply in gerrit | 19:39 | |
| devananda | oooh | 19:39 |
| devananda | i should read more carefully | 19:39 |
| devananda | lol | 19:39 |
| jroll | heh | 19:39 |
| jroll | you didn't wonder why I changed two templates? :P | 19:40 |
| mrda | devananda: What can you tell me about ironic/ironic/nova/compute/manager.py ? This ClusteredComputeManager is not in the nova tree. Do we leave it in our tree for now? | 19:41 |
| mrda | (you can leave this until after meeting) | 19:42 |
| *** bmahalakshmi has quit IRC | 19:42 | |
| jroll | mrda: nova hates it and doesn't want it | 19:43 |
| jroll | not sure what to do with it, though, if it should move or whatever | 19:43 |
| mrda | remove it from our tree? | 19:43 |
| jroll | no, we need it | 19:43 |
| jroll | well | 19:43 |
| JayF | heh | 19:43 |
| jroll | it was discussed that Ironic would keep it in tree for now | 19:43 |
| mrda | ok, it shall remain | 19:44 |
| mrda | thanks jroll | 19:44 |
| devananda | what jroll said | 19:44 |
| mrda | and devananda ;) | 19:44 |
| *** penick has quit IRC | 19:52 | |
| openstackgerrit | Michael Davies proposed a change to openstack/ironic: Remove driver from ironic now it's merged in Nova https://review.openstack.org/119892 | 19:53 |
| Shrews | and mrda completes the circle :) | 19:56 |
| mrda | :) | 19:56 |
| openstackgerrit | Michael Davies proposed a change to openstack/ironic: Remove driver from Ironic now it's merged in Nova https://review.openstack.org/119892 | 19:58 |
| mrda | darn capitalisation | 19:58 |
| mrda | :) | 19:58 |
| jroll | rloo (and all): anyone that needs to configure ironic should be capable of running 'pip install' and a shell script to generate the config | 20:00 |
| mrda | Just before people run away, I wanted to let y'all know that I will be going on leave for 4 weeks, as of Thursday. Big family holiday to the USA planned, will be back raring to go in the second week of October (and getting ready for summit!) | 20:00 |
| lucasagomes | yeah, well the more we talk the more benefits I see in removing it | 20:00 |
| jroll | mrda \o/ enjoy your time off | 20:01 |
| Shrews | mrda: 4 weeks? wow! enjoy | 20:01 |
| lucasagomes | mrda, :D enjoy it! | 20:01 |
| NobodyCam | thank you all.. great meeting ... nice to have you back devananda | 20:01 |
| jroll | lucasagomes: yeah, don't think people will have a problem generating it | 20:01 |
| mrda | thanks, 3 years in the planning | 20:01 |
| devananda | NobodyCam: thanks | 20:01 |
| jroll | if you can't generate our config, how are you going to deploy ironic | 20:01 |
| lucasagomes | yeah | 20:01 |
| devananda | and once again, thanks to -EVERYONE- | 20:01 |
| lucasagomes | so I'm +1 for removing it too | 20:01 |
| JayF | I *wish* we had a generated copy somewhere | 20:01 |
| JayF | like in our docs | 20:02 |
| JayF | but that's much less valuable than just not breaking our gate | 20:02 |
| devananda | it's hard to express how awesome it is to come back from ~3weeks away and see all the progress ya'll made w/o me :) | 20:02 |
| jroll | JayF: unfortunately docs are generated from our git repo :P | 20:02 |
| jroll | devananda: :) | 20:02 |
| lucasagomes | devananda, :) yeah ironic community rocks! | 20:02 |
| * jroll thinks many beers may be owed :P | 20:02 | |
| mrda | jroll: I don't like having to have to build a dev environment to get a config file though | 20:02 |
| rloo | hmm, so maybe we should generate the .sample and make it avail with our dev docs. better than nothing. | 20:02 |
| jroll | mrda: what are you going to do with a config file and no ironic environment | 20:02 |
| mrda | devananda: good to have you back! | 20:02 |
| jroll | I don't understand the use case for having a config, but no ironic environment | 20:03 |
| mrda | I've got an "ironic capable environment to run in" just not one that has the build deps on | 20:03 |
| jroll | the build deps are the same as the runtime deps | 20:03 |
| rloo | our dev docs actually refer to the sample (http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#exercising-the-services-locally) | 20:03 |
| mrda | sphynx? | 20:03 |
| jroll | sphynx isn't involved in building configs, is it? | 20:04 |
| dtantsur | g'night folks, c u tomorrow :) | 20:04 |
| devananda | for what do you actually need a sample config file filled with nothing but default values? | 20:04 |
| devananda | that is not satisfied by having documentation _explaining_ the config options? | 20:04 |
| devananda | which, fwiw, we actually dont have .... | 20:04 |
| lucasagomes | rloo, yeah in case we remove we need a step there to generate it first | 20:04 |
| *** dtantsur is now known as dtantsur|afk | 20:04 | |
| rloo | night dtantsur|afk | 20:04 |
| devananda | i think what i'm saying is, if we had _docs_ we wouldn't need a sample config file | 20:04 |
| jroll | agree | 20:04 |
| * mrda is happy with documentation | 20:05 | |
| devananda | but taht's just my momentary opinion :) | 20:05 |
| rloo | well, I find it useful to start with the sample, and then modify it. | 20:05 |
| jroll | rloo: so generate and modify? | 20:05 |
| * mrda isn't happy with pip -U test-requirements.txt to run ironic | 20:05 | |
| rloo | I'm curious to know what/how people that use nova, put together their nova.conf | 20:05 |
| devananda | rloo: because the sample tells you what options you have to change, right? or is there another reason? | 20:05 |
| jroll | what are you using it with, if you don't have a place to run it with | 20:05 |
| rloo | devananda: yeah, all the options are in one place so you can see them. and the syntax etc is there to minimize dumb typos. | 20:05 |
| devananda | rloo: they probably start with the sample configs their distro or their installer provides | 20:06 |
| openstackgerrit | A change was merged to openstack/ironic: Update ironic.conf.sample https://review.openstack.org/119816 | 20:06 |
| lucasagomes | for me the biggest problem of having the sample is what JayF said, break the gate w/o us doing anything | 20:06 |
| jroll | mrda: what is in test-requirements that we need for building sample configs? | 20:06 |
| rloo | i suppose if nova hasn't heard any operators complain about the .sample disappearing, it won't be a big problem. but i'd still like a sample .sample with our dev docs. | 20:06 |
| devananda | lucasagomes: i agree that that's a huge problem | 20:06 |
| * jroll tests | 20:07 | |
| lucasagomes | rloo, you mean having the config generate as part of our sphinx build? | 20:08 |
| devananda | how about this: post-merge job in Ironic's pipe which a) tests currentness of .conf.sample in our doc tree. IFF not current, it b) proposese a change, with a Doc-impact flag in the commit message | 20:08 |
| JayF | I go back to what I said before: do we really want to increase the number of patches we're pushing through an already-crowded gate? | 20:09 |
| devananda | in principle, if we had doc writer(s), we should use the doc-impact flag any time there's a change to a config option | 20:09 |
| rloo | lucasagomes: was thinking of some url off of http://docs.openstack.org/developer/ironic/, with a sample | 20:09 |
| JayF | I think generating the config as part of the docs build (or a separate job) makes more sense than keeping it in git :) | 20:09 |
| devananda | JayF: right. except our docs are in our git tree today | 20:09 |
| devananda | which, fwiw, is the preferred approach for incubating projects, but not for integrated projects | 20:10 |
| lucasagomes | we have a tox command to generate the config already | 20:10 |
| lucasagomes | maybe just use it as part of the doc | 20:10 |
| rloo | devananda: but we generate docs from 'source' in our git tree. | 20:10 |
| devananda | rloo: those are developer docs | 20:10 |
| devananda | not user docs | 20:10 |
| lucasagomes | and create a link to it in the docs | 20:10 |
| rloo | devananda: does nova have user docs for their configs that is similar to their .sample before? | 20:11 |
| devananda | rloo: these http://docs.openstack.org/developer/ironic/#admin-guide should eventually be moved out of our git tree and into a separate doc tree | 20:11 |
| devananda | rloo: yes. one sec | 20:11 |
| * JayF afk for a while | 20:11 | |
| devananda | rloo: http://docs.openstack.org/icehouse/config-reference/content/ | 20:11 |
| rloo | devananda: yeah i'll be very happy when the admin guide stuff is out ;) | 20:11 |
| devananda | once integrated, we'll need to add those | 20:11 |
| devananda | eg, http://docs.openstack.org/icehouse/config-reference/content/list-of-compute-config-options.html | 20:12 |
| rloo | devananda: this is ringing a bell. the docs folks generate those configs somehow. | 20:12 |
| devananda | yep :) | 20:12 |
| devananda | looks auto generated | 20:12 |
| devananda | right from the oslo.config stuff | 20:12 |
| rloo | i have notes somewhere... looking... | 20:13 |
| lucasagomes | :) | 20:13 |
| lucasagomes | alright so it's late here | 20:13 |
| rloo | ok, my notes only say that Anne Gentle said they have a script that generates it ;) | 20:13 |
| * devananda notes the nice samples here http://docs.openstack.org/icehouse/config-reference/content/section_compute-config-samples.html | 20:13 | |
| devananda | rloo: heh :) | 20:14 |
| devananda | lucasagomes: ciao! | 20:14 |
| lucasagomes | have a good night everyone! | 20:14 |
| lucasagomes | devananda, night | 20:14 |
| rloo | night lucasagomes | 20:14 |
| lucasagomes | I will read the backlog later to, but now my votes are: +1 to add to the docs | 20:14 |
| *** lucasagomes is now known as lucas-dinner | 20:14 | |
| rloo | so that docn doesn't show the entire list of configs in one page. it breaks it up into sections. which I guess is fine. but if we could, it might be useful to keep a .sample in our dev docs. | 20:15 |
| jroll | mrda (and all), reporting back, one can generate sample config with only requirements.txt installed, no test-requirements.txt needed | 20:15 |
| rloo | that'd help me when reviewing changes to conf, to see how they look, cuz I think the docs folks use those descriptions for their generated config doc. | 20:16 |
| *** Shrews has quit IRC | 20:17 | |
| *** Shrews has joined #openstack-ironic | 20:17 | |
| *** penick has joined #openstack-ironic | 20:19 | |
| mrda | jroll: great news | 20:19 |
| jroll | :) | 20:19 |
| mrda | my fears relieved! I'm not unhappy now. | 20:19 |
| NobodyCam | night lucas-dinner | 20:20 |
| Shrews | greghaynes: i see now that the decorator doesn't work, though i'm clear why | 20:22 |
| Shrews | greghaynes: _not_ clear why | 20:22 |
| adam_g | devananda, a few things added to the graduation pad re QA, not much to add tho LGTM | 20:26 |
| *** r-daneel has quit IRC | 20:27 | |
| *** r-daneel has joined #openstack-ironic | 20:28 | |
| NobodyCam | humm wounder who runs ask.openstack | 20:31 |
| greghaynes | Shrews: Because the decorator is evaluated when the module is imported, rather than at run time | 20:31 |
| lifeless | NobodyCam: foundation I believe | 20:32 |
| greghaynes | Shrews: basically I think to make it work youd have to mock.patch the result of dbapi.get_instance() but that seems like a layering violation | 20:32 |
| NobodyCam | :) Ty lifeless | 20:32 |
| *** sbadia has quit IRC | 20:32 | |
| devananda | adam_g: ty | 20:34 |
| openstackgerrit | Michael Davies proposed a change to openstack/ironic: Remove driver from Ironic now it's merged in Nova https://review.openstack.org/119892 | 20:34 |
| *** chuckC has joined #openstack-ironic | 20:35 | |
| *** ChuckC_ has quit IRC | 20:35 | |
| NobodyCam | lol https://bugs.launchpad.net/openstack-community/+bug/1366975 | 20:35 |
| jroll | heh | 20:36 |
| devananda | NobodyCam: nice one | 20:36 |
| *** sbadia has joined #openstack-ironic | 20:37 | |
| NobodyCam | :-p | 20:37 |
| * jroll steps away for a bit | 20:37 | |
| * devananda steps away for the afternoon to take care of some errands | 20:39 | |
| *** HenryG has quit IRC | 20:48 | |
| *** eguz has joined #openstack-ironic | 20:56 | |
| *** eghobo has quit IRC | 20:59 | |
| *** HenryG has joined #openstack-ironic | 21:00 | |
| *** killer_prince has joined #openstack-ironic | 21:03 | |
| *** killer_prince is now known as lazy_prince | 21:03 | |
| *** pelix has quit IRC | 21:06 | |
| *** linggao has quit IRC | 21:15 | |
| mrda | NobodyCam: btw, thanks for the last recheck on that final driver patch. Saved us a few hours :) | 21:35 |
| *** SpamapS_ is now known as SpamapS | 21:35 | |
| *** SpamapS has quit IRC | 21:35 | |
| *** SpamapS has joined #openstack-ironic | 21:35 | |
| *** jcoufal has quit IRC | 21:35 | |
| devananda | looks like my devstack patch to swap drivers failed at: http://logs.openstack.org/44/119844/2/check/check-tempest-dsvm-virtual-ironic-nv/ce443f8/logs/screen-n-cpu.txt.gz#_2014-09-08_20_34_23_010 | 21:36 |
| lifeless | devananda: its all landed? we should stop having tripleo patch out the driver? | 21:36 |
| devananda | lifeless: as of today, yes. all landed | 21:37 |
| devananda | lifeless: however, see ^ -- my attempt to switch the config in devstack hit something unexpected | 21:37 |
| devananda | mrda: ^ | 21:37 |
| mrda | hmmm | 21:39 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Fix compute_driver in install-guide https://review.openstack.org/119937 | 21:39 |
| jroll | why do I feel like I've seen this? | 21:40 |
| mrda | extra_specs? | 21:40 |
| Shrews | is a flavor update needed? | 21:40 |
| NobodyCam | oh is there a patch for this? | 21:40 |
| adam_g | devananda, https://review.openstack.org/#/c/119707/ wonder if this is relevant? | 21:42 |
| jroll | that actually doesn't do anything with the flavor objects | 21:43 |
| devananda | i think i found the cause | 21:43 |
| devananda | waiting for git ... | 21:43 |
| devananda | nope, nvm | 21:43 |
| * adam_g spawns a local devstack | 21:44 | |
| devananda | yep | 21:46 |
| *** mikedillion has quit IRC | 21:46 | |
| devananda | the driver code in nova != driver code in ironic | 21:46 |
| jroll | gah | 21:47 |
| jroll | what's different? | 21:47 |
| devananda | some of this is trivial, some is ... not | 21:47 |
| devananda | mrda: http://paste.openstack.org/show/auIHAS93ScQZsL4RXCQ3/ | 21:47 |
| devananda | L92 is the problem, I think | 21:48 |
| jroll | devananda: that passed ironic/tempest in the nova review, afaik | 21:48 |
| * jroll looks | 21:48 | |
| devananda | jroll: the code in ironic passed everything. | 21:48 |
| mrda | ok, so this change is in ironic. jroll had a patch | 21:48 |
| jroll | hrm | 21:49 |
| jroll | yeah | 21:49 |
| devananda | jroll: *nothing* in the nova reviews was actually tested by tempest | 21:49 |
| jroll | uh oh | 21:49 |
| jroll | hm | 21:49 |
| jroll | devananda: https://review.openstack.org/#/c/119165/ | 21:49 |
| jroll | ah shit | 21:49 |
| devananda | right | 21:49 |
| jroll | same error | 21:49 |
| * jroll sighs | 21:50 | |
| devananda | wait, huh? | 21:50 |
| jroll | yeah, that shouldn't have been ported back to nova until it passed tests over here | 21:50 |
| jroll | though I think that was the 24-hour check queue day | 21:50 |
| mrda | sigh | 21:50 |
| jroll | devananda: http://logs.openstack.org/65/119165/3/check/check-tempest-dsvm-virtual-ironic/c161a89/logs/screen-n-cpu.txt.gz#_2014-09-08_08_41_06_821 | 21:50 |
| devananda | gotcha | 21:50 |
| mrda | that was the last non-nit change that Nova needed | 21:51 |
| jroll | that was probably my bad in porting it back | 21:51 |
| jroll | right | 21:51 |
| mrda | and it corresponded to the worst gate day | 21:51 |
| devananda | so the feedback on rev 22 there failed tempest when applied to ironic, but was included in the nova driver patch anyway | 21:51 |
| devananda | :( | 21:51 |
| devananda | oh well | 21:52 |
| jroll | yeah :/ | 21:52 |
| mrda | so close | 21:52 |
| jroll | I'm not sure I understand why that doesn't work :/ | 21:52 |
| rloo | well, here's our first bug in nova. (the only bright side I guess) | 21:52 |
| jroll | that's a bright side? | 21:53 |
| rloo | jroll: the first of many? ugh. | 21:53 |
| mrda | rloo: hope not | 21:53 |
| jroll | nope | 21:54 |
| jroll | bug free | 21:54 |
| jroll | (after this one) | 21:54 |
| rloo | not bug free. we know there are bugs already. | 21:54 |
| *** eguz has quit IRC | 21:54 | |
| rloo | (or enhancements in some cases) | 21:54 |
| mrda | so Nova isn't happy with the extra db call, so we need to know why get_flavor() isn't working as expected | 21:54 |
| *** eghobo has joined #openstack-ironic | 21:55 | |
| jroll | well | 21:55 |
| jroll | I wonder if something is cached | 21:55 |
| jroll | and not hitting the db makes this present | 21:55 |
| devananda | time to ask in -nova :) | 21:55 |
| mrda | as in the flavor in nova doesn't match what's in the db? | 21:55 |
| devananda | explanatory comment posted just now on https://review.openstack.org/#/c/111429/ | 21:56 |
| jroll | mrda: something like that, idk | 21:56 |
| *** penick has quit IRC | 21:58 | |
| jroll | mrda, devananda, I think extra_specs needs to be in here: https://github.com/openstack/nova/blob/master/nova/compute/flavors.py#L73-84 | 21:58 |
| jroll | more system_metadata hackiness, I think | 21:59 |
| devananda | jroll: ... for our use case | 21:59 |
| jroll | comstud will love it. | 21:59 |
| jroll | devananda: right | 21:59 |
| comstud | hah | 22:00 |
| devananda | jroll: yea. seems like a thing to point out to the original proposers of that change (alaski and dansmith) | 22:00 |
| jroll | devananda: indeed | 22:00 |
| devananda | jroll: also worth sorting out asap, IMO | 22:00 |
| jroll | pinging now | 22:00 |
| *** rloo is now known as rloo_afk | 22:00 | |
| devananda | we shouldn't be landing any changes to ironic/nova/* or to nova/virt/ironic/* right now | 22:01 |
| comstud | so | 22:01 |
| devananda | but I can't stop Nova from doing that,a nd folks are already proposing changes there | 22:01 |
| devananda | comstud: maybe you can help with ^ | 22:01 |
| devananda | comstud: there's a string of changes from garyk: https://review.openstack.org/#/c/119707/4 | 22:01 |
| mrda | comstud: if you could that would be appreciated | 22:01 |
| comstud | yeah, i saw them last night | 22:01 |
| comstud | so | 22:01 |
| comstud | instance.get_flavor() is not going to return extra specs | 22:02 |
| *** penick has joined #openstack-ironic | 22:02 | |
| comstud | you'd have to use Flavor.get* | 22:02 |
| * comstud looks at the series again | 22:02 | |
| jroll | comstud: right, we broke it here: https://review.openstack.org/#/c/119165/3/ironic/nova/virt/ironic/driver.py | 22:02 |
| devananda | comstud: it was originally using flavor_obj.Flavor.get_by_id | 22:03 |
| devananda | which dansmith -1'd | 22:03 |
| mrda | ...he didn't like the extra db call | 22:03 |
| jroll | devananda: dansmith and alaski are fine with switching it back for cases where we need extra_specs | 22:04 |
| comstud | yeah, so yeah, probably didn't catch that we need extar_specs there | 22:04 |
| comstud | assuming that we do | 22:04 |
| JoshNang | ntpdate ntp.ubuntu.com pool.ntp.org | 22:04 |
| jroll | devananda: you wanna propose that change | 22:04 |
| JoshNang | blah, wrong window. sorry guys | 22:04 |
| jroll | JoshNang: -1 | 22:04 |
| devananda | jroll: sure | 22:05 |
| jroll | thanks | 22:05 |
| mrda | damn it | 22:06 |
| comstud | that first patch from gary is fine | 22:07 |
| comstud | i didn't catch that we were not using objects.<Name> in our driver yet | 22:07 |
| comstud | but ya, let's wait until devstack is testing | 22:07 |
| mrda | sorry ironic for missing this :( | 22:09 |
| comstud | (that one is just a cleanup) | 22:10 |
| comstud | the flavor thing... hard to blame anyone | 22:10 |
| comstud | it's just one of those "fucked up nova things" | 22:10 |
| jroll | mrda: pretty sure I'm the bad guy that said we should just put it in | 22:11 |
| comstud | we hackily store flavor info in sys_meta... but we don't include extra_specs | 22:11 |
| openstackgerrit | Syed Ismail Faizan Barmawer proposed a change to openstack/ironic: Add uefi boot mode support in IloVirtualMediaIscsiDeploy https://review.openstack.org/116561 | 22:16 |
| *** penick has quit IRC | 22:20 | |
| *** romcheg1 has quit IRC | 22:21 | |
| devananda | huh, wtf ... pip install -r requirements.txt isn't working for nova now? | 22:25 |
| jroll | wat | 22:26 |
| devananda | No package 'libvirt' found | 22:26 |
| devananda | ah, test-requirements has libvirt-python | 22:27 |
| devananda | but wtf? | 22:27 |
| mrda | devananda: there's been some fun in the libvirt space over the last few weeks, IIRC | 22:27 |
| devananda | no, it's failing to install libvirt itself | 22:27 |
| mrda | ^maybe month | 22:27 |
| devananda | right. not going to spend the time right now | 22:28 |
| devananda | already waay behind on some errands i need to do | 22:28 |
| mrda | devananda: do you want me to propose a new ironic patch that reverts get_flavor() in spawn() and see that pass? | 22:30 |
| mrda | and then do the same for the driver in nova? | 22:30 |
| devananda | https://review.openstack.org/#/c/119944/ | 22:34 |
| devananda | mrda: the offending change never landed in ironic in the first place :) | 22:34 |
| devananda | it failed our tempest tests | 22:34 |
| jroll | well | 22:34 |
| mrda | right, so perhaps, to be careful, get that to land with your update? | 22:34 |
| jroll | we should fix my ironic patch to just revert the one we changed in spawn() | 22:34 |
| jroll | and see if it passes | 22:34 |
| jroll | we should use get_flavor() if we aren't using extra_specs there | 22:35 |
| devananda | jroll: feel free to clobber my patch. I just reverted everything in your patch | 22:35 |
| devananda | just to be on the safe side | 22:35 |
| jroll | devananda: heh, that's fine with me | 22:35 |
| jroll | working > perfect | 22:35 |
| jroll | +1 | 22:35 |
| devananda | i'm more concerned with being able to test the code as it is in Nova than with optimizing db access at this point | 22:35 |
| *** penick has joined #openstack-ironic | 22:36 | |
| mrda | thanks devananda | 22:36 |
| mrda | again, sorry | 22:36 |
| jroll | indeed | 22:36 |
| devananda | np | 22:36 |
| mrda | I think 2 or the 3 of the db accessess will be revoked, but let's start with a clean baseline first :) | 22:37 |
| mrda | s/or/of/ | 22:37 |
| devananda | mrda: fair enough. like i said, feel free to clobber / take over that patch if you want to improve it | 22:38 |
| *** penick has quit IRC | 22:38 | |
| devananda | i really do need to step away now and attend to some legal matters. might be back tonight | 22:38 |
| mrda | thanks, I'll take a look | 22:39 |
| openstackgerrit | A change was merged to openstack/ironic: Fix race conditions running pxe_utils tests in parallel https://review.openstack.org/119746 | 22:52 |
| *** lucas-dinner has quit IRC | 23:01 | |
| *** pcrews has quit IRC | 23:03 | |
| *** pcrews has joined #openstack-ironic | 23:05 | |
| * NobodyCam brb | 23:14 | |
| openstackgerrit | Syed Ismail Faizan Barmawer proposed a change to openstack/ironic: Support for setting boot mode in pxe_ilo driver https://review.openstack.org/118578 | 23:26 |
| *** chuckC has quit IRC | 23:56 | |
| *** chuckC has joined #openstack-ironic | 23:57 | |
| *** chuckC has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!