| *** dwalleck has quit IRC | 00:02 | |
| *** matsuhashi has joined #openstack-ironic | 00:26 | |
| *** derekh has quit IRC | 00:28 | |
| *** openstackgerrit has joined #openstack-ironic | 00:30 | |
| *** eguz has joined #openstack-ironic | 00:39 | |
| *** eghobo has quit IRC | 00:43 | |
| *** matsuhashi has quit IRC | 00:47 | |
| *** matsuhashi has joined #openstack-ironic | 00:48 | |
| *** matsuhashi has quit IRC | 00:52 | |
| *** matsuhashi has joined #openstack-ironic | 00:57 | |
| Shrews | adam_g: So, I just had another thought about your nova unit tests. It isn't testing under both py26 AND py27. | 01:00 |
|---|---|---|
| Shrews | adam_g: Perhaps it could be changed to create it's own temp directory (instead of letting tox create .tox/nova-tests) and changing nova-tests.sh to directly call tox? | 01:02 |
| Shrews | ah shoot, then we'd lose the tox target... | 01:02 |
| Shrews | nevermind | 01:02 |
| adam_g | Shrews, yeah--that would probably work, we'll just need to have some special case wrapper script that gate uses the test this stuff | 01:02 |
| Shrews | yeah, we need a tox target. let me dig through the tox docs a bit. been a while since i've read them. maybe there's something we're missing that could be useful | 01:03 |
| adam_g | Shrews, py26nova + py27nova might work. im still a total tox newbie so yeah--i may be overlooking something too | 01:04 |
| Shrews | technology is hard. maybe i should be a ditch digger | 01:06 |
| Shrews | or a slushie machine repairman | 01:06 |
| rloo | adam_g, Shrews: this is a dumb question. Any idea how long we're planning on keeping the nova-related code in ironic? I think nova will be open for juno soon -- can't we just put the nova code back in nova then? | 01:08 |
| adam_g | rloo, thats a great point | 01:09 |
| adam_g | :) | 01:09 |
| adam_g | i'm not entirely sure what the timeline for that is | 01:09 |
| adam_g | but it might make sense to do that instead of jerry-rigging a strange test runner into ironic / the gate | 01:10 |
| Shrews | hrm, yeah. good question | 01:11 |
| rloo | well, you can only hope :-) | 01:17 |
| *** eguz has quit IRC | 01:20 | |
| *** rloo has quit IRC | 01:23 | |
| *** yongli has joined #openstack-ironic | 01:29 | |
| *** nosnos has joined #openstack-ironic | 01:47 | |
| lifeless | adam_g: strange test runner? | 01:48 |
| lifeless | adam_g: nova uses the same as ironic as far as runners go | 01:48 |
| adam_g | lifeless, i meant running it via some strange set of non-standard scripts instead of thru tox/setup.py | 01:50 |
| lifeless | adam_g: why do you need to do that ? | 01:51 |
| lifeless | adam_g: they should be picked up by ironics .testr.conf and just run | 01:51 |
| adam_g | lifeless, im not sure we need to if the ironic driver is going back into nova in the near future | 01:51 |
| lifeless | adam_g: I mean right now they should already be running | 01:52 |
| adam_g | lifeless, no--we're trying to run nova's tests with ironic's driver inserted into its tree | 01:52 |
| adam_g | since the driver's tests live in ironic currently, they do not get run anywhere | 01:52 |
| lifeless | adam_g: have you identified *any* tests that that will impact? | 01:52 |
| lifeless | adam_g: are you sure? | 01:52 |
| adam_g | lifeless, yes, i am sure | 01:53 |
| lifeless | adam_g: doesn't Ironic run its unit tests? | 01:53 |
| adam_g | lifeless, it does | 01:53 |
| adam_g | the virt drivers unit tests are not part of those | 01:53 |
| lifeless | then the driver tests should be run | 01:53 |
| adam_g | lifeless, context https://review.openstack.org/#/c/84033/ | 01:54 |
| lifeless | adam_g: why do you say the nova drivers are not run? | 01:54 |
| lifeless | bah | 01:54 |
| lifeless | nova driver's tests are not run | 01:54 |
| adam_g | right | 01:55 |
| lifeless | adam_g: *why* do you say that? | 01:58 |
| lifeless | adam_g: why would some tests in the ironic tree not run ? | 01:58 |
| adam_g | lifeless, for several reasons | 02:00 |
| lifeless | adam_g: please go on - as the primary author of much of the test environment, I'm completely bamboolzed here | 02:01 |
| adam_g | 1) the code was copied inf rom nova at some point and they live in ironic/nova/tests. they are not loaded by testr | 02:01 |
| lifeless | they should be found by discover. | 02:01 |
| adam_g | mostly because | 02:01 |
| adam_g | ironic/nova/tests/__init__.py is not there | 02:01 |
| adam_g | but also, that doesnt matter | 02:01 |
| lifeless | thats probably key | 02:01 |
| adam_g | the tests require nova | 02:01 |
| lifeless | add nova as a test-requirement and touch __init__.py and they'll run. | 02:01 |
| adam_g | lifeless, they wont | 02:02 |
| lifeless | why not? | 02:02 |
| adam_g | lifeless, both sets of tests (the main ironic tests, the nova driver tests) cannot be imported by testr at the same time during discovery | 02:03 |
| lifeless | adam_g: details please! | 02:03 |
| adam_g | there are some conflicts around oslo config resources | 02:03 |
| lifeless | ahha, now we're getting to the meat of it | 02:03 |
| adam_g | yup | 02:03 |
| lifeless | oslo.config's global nature | 02:03 |
| lifeless | its a terrible thing | 02:03 |
| adam_g | besides | 02:03 |
| adam_g | i imagine we'd be more interested in testing the driver against nova trunk, not a released tarball from pypi | 02:04 |
| adam_g | either way--getting the driver back into nova would fix all of this | 02:04 |
| lifeless | adam_g: you would test against trunk | 02:04 |
| lifeless | nova doesn't upload to pypi | 02:04 |
| lifeless | in devstack zuul decides what version of projects you test against | 02:04 |
| adam_g | lifeless, not sure devstack is relevant? ie, gate-ironic-python27 | 02:05 |
| lifeless | right, so that checks out code based on zuul | 02:05 |
| lifeless | and requirements like e.g. python-ironicclient are not pulled from pypi, they are git refs | 02:06 |
| adam_g | lifeless, oh, right | 02:06 |
| adam_g | anyway, i gotta run | 02:07 |
| adam_g | cya | 02:08 |
| lifeless | night! | 02:08 |
| lifeless | sorry for the grueling 20 questions | 02:08 |
| lifeless | I really wanted to understand | 02:09 |
| adam_g | np :) | 02:11 |
| openstackgerrit | Tom Fifield proposed a change to openstack/ironic: Fix grammar in error string in pxe driver https://review.openstack.org/84471 | 02:11 |
| *** matsuhashi has quit IRC | 03:21 | |
| *** harlowja is now known as harlowja_away | 03:24 | |
| Haomeng|2 | devananda: got, thanks, let me try to generate these mo files and verify our lazy patch | 03:25 |
| *** nosnos has quit IRC | 03:37 | |
| *** eghobo has joined #openstack-ironic | 03:45 | |
| *** killer_prince is now known as lazy_prince | 03:55 | |
| *** nosnos has joined #openstack-ironic | 04:31 | |
| *** killer_prince has joined #openstack-ironic | 04:35 | |
| *** matsuhashi has joined #openstack-ironic | 04:36 | |
| *** matsuhashi has quit IRC | 04:50 | |
| *** rameshg87 has joined #openstack-ironic | 04:54 | |
| *** radsy has quit IRC | 04:55 | |
| *** matsuhashi has joined #openstack-ironic | 05:00 | |
| *** lazy_prince has quit IRC | 05:46 | |
| *** killer_p- has joined #openstack-ironic | 05:46 | |
| *** sabah has joined #openstack-ironic | 05:48 | |
| openstackgerrit | Jenkins proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/83956 | 06:06 |
| *** mrda is now known as mrda_away | 06:14 | |
| *** raies has joined #openstack-ironic | 06:37 | |
| raies | Hi someone | 06:37 |
| raies | can some one confirm that during ironic installation driver with name "fake" automatically exists ?? | 06:38 |
| *** saju_m has joined #openstack-ironic | 06:39 | |
| *** mdurnosvistov has quit IRC | 06:42 | |
| raies | any body ?? | 06:48 |
| raies | ^^ | 06:48 |
| russell_h | raies: yeah, it does | 06:49 |
| raies | russell_h: can you please locate it in code from github ?? | 06:49 |
| russell_h | raies: https://github.com/openstack/ironic/blob/master/setup.cfg#L33 | 06:49 |
| russell_h | raies: that lists all of the drivers installed by default, and references their actual implementations | 06:50 |
| russell_h | raies: as I understand it, to use some of them you may need to install external dependencies. As of yesterday this was actively being worked on, not sure of the exact status. | 06:50 |
| raies | russell_h: thank you very much | 06:51 |
| russell_h | raies: out of interest, is this causing problems/concerns for you? | 06:51 |
| raies | No No it is not like that | 06:51 |
| raies | Actually I am writing some piece of codes in tempest | 06:52 |
| *** coolsvap has joined #openstack-ironic | 06:52 | |
| russell_h | raies: ahh, ok, gotcha | 06:52 |
| raies | russell_h: So I wanted just to confirm about default drivers in ironic :) | 06:52 |
| russell_h | makes sense | 06:52 |
| raies | russell_h: thanks a lot for sharing information | 06:52 |
| russell_h | for sure | 06:52 |
| * russell_h off for the night | 06:53 | |
| *** subah has joined #openstack-ironic | 06:55 | |
| *** sabah has quit IRC | 06:55 | |
| *** pbrooko has joined #openstack-ironic | 07:16 | |
| GheRivero | morning all | 07:28 |
| *** sabah has joined #openstack-ironic | 07:29 | |
| *** subah has quit IRC | 07:29 | |
| *** ifarkas has joined #openstack-ironic | 07:40 | |
| *** saju_m has quit IRC | 07:42 | |
| *** florentflament has joined #openstack-ironic | 07:43 | |
| *** romcheg has joined #openstack-ironic | 07:49 | |
| *** saju_m has joined #openstack-ironic | 07:52 | |
| *** jistr has joined #openstack-ironic | 07:53 | |
| *** ndipanov has joined #openstack-ironic | 08:01 | |
| *** sabah has quit IRC | 08:08 | |
| *** pradipta_away is now known as pradipta | 08:22 | |
| *** eghobo has quit IRC | 08:23 | |
| *** lucasagomes has joined #openstack-ironic | 08:25 | |
| *** mdurnosvistov has joined #openstack-ironic | 08:25 | |
| *** mdurnosvistov has left #openstack-ironic | 08:28 | |
| *** derekh has joined #openstack-ironic | 08:28 | |
| *** mdurnosvistov has joined #openstack-ironic | 08:28 | |
| *** sabah has joined #openstack-ironic | 08:36 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Supports filtering port by address https://review.openstack.org/82773 | 08:41 |
| *** saju_m has quit IRC | 08:48 | |
| *** saju_m has joined #openstack-ironic | 08:49 | |
| *** Mikhail_D_wk has joined #openstack-ironic | 08:54 | |
| *** mdurnosvistov has left #openstack-ironic | 08:54 | |
| *** Mikhail_D_ltp has joined #openstack-ironic | 08:55 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Move duplicated _get_node(s)_mac_addresses() https://review.openstack.org/78787 | 09:00 |
| *** coolsvap has quit IRC | 09:02 | |
| *** romcheg has quit IRC | 09:13 | |
| *** yuriyz has quit IRC | 09:18 | |
| *** yuriyz has joined #openstack-ironic | 09:19 | |
| *** max_lobur has joined #openstack-ironic | 09:21 | |
| *** max_lobur1 has joined #openstack-ironic | 09:22 | |
| *** saju_m has quit IRC | 09:24 | |
| *** saju_m has joined #openstack-ironic | 09:26 | |
| *** max_lobur has quit IRC | 09:26 | |
| *** yuriyz has quit IRC | 09:27 | |
| *** yuriyz has joined #openstack-ironic | 09:27 | |
| *** romcheg has joined #openstack-ironic | 09:29 | |
| *** saju_m has quit IRC | 09:31 | |
| *** subah has joined #openstack-ironic | 09:39 | |
| *** sabah has quit IRC | 09:39 | |
| *** saju_m has joined #openstack-ironic | 09:39 | |
| dtantsur | Hi Ironic | 09:48 |
| Mikhail_D_ltp | dtantsur: hi! And good morning folks! =) | 09:49 |
| romcheg | Morning dtantsur :) | 09:50 |
| dtantsur | lucasagomes, any updates on bug with cleaning up nodes? Maybe I can help? It's actually a blocker for me :( | 09:50 |
| *** matsuhashi has quit IRC | 09:50 | |
| lucasagomes | dtantsur, hey there's a patch up there | 09:51 |
| * lucasagomes lemme find the number | 09:51 | |
| lucasagomes | dtantsur, https://review.openstack.org/#/c/84497/ | 09:51 |
| lucasagomes | dtantsur, lemme know if it works fine for u | 09:52 |
| dtantsur | sure | 09:52 |
| lucasagomes | dtantsur, and morning :) | 09:52 |
| lucasagomes | romcheg, Mikhail_D_ltp morning for u too :D | 09:52 |
| romcheg | Morning lucasagomes | 09:52 |
| Mikhail_D_ltp | lucasagomes: morning :) | 09:53 |
| *** matsuhashi has joined #openstack-ironic | 09:57 | |
| *** max_lobur has joined #openstack-ironic | 09:57 | |
| *** max_lobur1 has quit IRC | 09:59 | |
| *** max_lobur2 has joined #openstack-ironic | 09:59 | |
| *** max_lobur2 has quit IRC | 10:00 | |
| *** max_lobur1 has joined #openstack-ironic | 10:02 | |
| *** max_lobur has quit IRC | 10:02 | |
| *** subah has quit IRC | 10:03 | |
| dtantsur | I think it's a separate issue, but sill: can we somehow enable to delete instance that once failed to delete? Or does it belong to nova? | 10:05 |
| dtantsur | lucasagomes, and I still get "RPC do_node_tear_down not allowed for node e8d9b219-4c7f-44b5-baa5-00d939b35180 in state None" in response to nova boot | 10:07 |
| dtantsur | strange, btw, that it happens instantly.. | 10:07 |
| lucasagomes | hmmm weird | 10:08 |
| lucasagomes | so you issued a nova boot and it's now showing the error as RPC do_node_tear_down ... ? | 10:08 |
| lucasagomes | and now you can't delete the instance from nova? | 10:08 |
| dtantsur | lucasagomes, I've no idea what's going on. But after I changed flavor and nodes, it failes nearly at once | 10:08 |
| lucasagomes | it was before you applied the patch? | 10:08 |
| dtantsur | lucasagomes, before and after | 10:09 |
| lucasagomes | hmm | 10:09 |
| lucasagomes | idk how nova works, but if the instance fail to delete idk if it tries to re-issue the delete | 10:10 |
| lucasagomes | or might have a way to do that | 10:10 |
| dtantsur | lucasagomes, I'm lost :( and nothing interesting in conductor logs - that's what is strange | 10:13 |
| dtantsur | lucasagomes, oh, wait a bit, maybe I didn;t properly apply the patch | 10:15 |
| lucasagomes | dtantsur, oh | 10:15 |
| lucasagomes | dtantsur, you gotta restart the services as well after the patch is applied | 10:15 |
| lucasagomes | n-cpu | 10:15 |
| dtantsur | lucasagomes, one moment. I should have had vim deleted from my machine right after install >_< First time I see console editor hang | 10:17 |
| lucasagomes | dtantsur, ouch | 10:17 |
| lucasagomes | hah ok | 10:17 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Show maintenance field in GET /nodes https://review.openstack.org/84685 | 10:18 |
| dtantsur | ok, I think I'm ready to try again | 10:21 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Better check for available nodes https://review.openstack.org/84687 | 10:22 |
| lucasagomes | dtantsur, ack lemme know | 10:22 |
| dtantsur | lucasagomes, much better, now it's no valid host found (I have 2 suitable hosts) | 10:23 |
| lucasagomes | dtantsur, yeah, no valid host found happens because we have the retry filter enabled | 10:24 |
| lucasagomes | :( | 10:24 |
| dtantsur | lucasagomes, what do you mean? | 10:24 |
| *** sabah has joined #openstack-ironic | 10:24 | |
| lucasagomes | so I tried to see if there's anything I can do to tell nova to not try to rescheduler it | 10:24 |
| lucasagomes | dtantsur, if a deploy fails in the spawn() | 10:24 |
| lucasagomes | nova will try to rescheduler it | 10:24 |
| lucasagomes | 1 sec, better show in code | 10:25 |
| lucasagomes | dtantsur, https://github.com/openstack/nova/blob/master/nova/scheduler/host_manager.py#L47 | 10:26 |
| dtantsur | lucasagomes, now I have some more interesting errors in nova logs: | 10:26 |
| dtantsur | 1. HTTPBadRequest: Couldn't apply patch '[{'path': '/driver_info/pxe_deploy_kernel', 'op': 'remove'}]'. Reason: 'pxe_deploy_kernel' | 10:26 |
| lucasagomes | yeah, the patch is logging what happened in the spawn() | 10:26 |
| dtantsur | 2. KeyError: 'baremetal:deploy_kernel_id' | 10:26 |
| lucasagomes | dtantsur, and u can delete the instance from nova? | 10:27 |
| dtantsur | lucasagomes, yes I can | 10:27 |
| lucasagomes | :) | 10:27 |
| dtantsur | Great progress :) | 10:27 |
| lucasagomes | heh yeah | 10:27 |
| lucasagomes | not perfect tho | 10:27 |
| lucasagomes | but I'm also learning the flow in nova | 10:27 |
| lucasagomes | it solves part of the problem | 10:28 |
| dtantsur | lucasagomes, very good already, +1 on this part | 10:28 |
| lucasagomes | dtantsur, thanks and thanks for testing as well | 10:28 |
| openstackgerrit | A change was merged to openstack/ironic: Improved coverage for ironic API https://review.openstack.org/84376 | 10:32 |
| *** subah has joined #openstack-ironic | 10:34 | |
| *** sabah has quit IRC | 10:34 | |
| dtantsur | lucasagomes, ok, outcome of my debugging: I forgot to properly initialize extras for nova flavor; 2. operation [{'path': '/driver_info/pxe_deploy_kernel', 'op': 'remove'}] needs except case if it happens too early i.e. no pxe_deploy_kernel yet | 10:48 |
| lucasagomes | :) | 10:49 |
| lucasagomes | yeah nova should populate those for u | 10:49 |
| dtantsur | lucasagomes, that's correct, but in my case there was no baremetal:deploy_kernel_id and it tried to delete values that was not set | 10:50 |
| lucasagomes | I see, hmm yeah the way the clean_up works right now in the driver is not bullet proof :/ | 10:51 |
| dtantsur | lucasagomes, ok, now: StdErr: 'Error: You requested a partition from 0.00MiB to 20480MiB (sectors 0..41943039).\nThe closest location we can manage is 0.00MiB to 20480MiB (sectors 1..41943039).\n' | 10:52 |
| dtantsur | this is strange, but that's what I got as parted stderr | 10:53 |
| *** romcheg has quit IRC | 10:55 | |
| dtantsur | lucasagomes, the last comment may help: https://bugzilla.redhat.com/show_bug.cgi?id=602680 | 10:56 |
| lucasagomes | hmm, right, it's a good feedback, lemme see if I can find something | 10:56 |
| *** romcheg has joined #openstack-ironic | 10:56 | |
| lucasagomes | we might need to leave an 1MB at the start | 10:56 |
| dtantsur | lucasagomes, seems like yes, I didn't notice it can't handle sector 0 | 10:58 |
| dtantsur | Could you update your patch? | 10:58 |
| lucasagomes | yeah me too | 10:58 |
| lucasagomes | dtantsur, will do, I have to do some tests as well | 10:58 |
| *** matsuhashi has quit IRC | 11:00 | |
| dtantsur | offset = 1 works for me, left comment on a patch | 11:04 |
| *** nosnos has quit IRC | 11:04 | |
| lucasagomes | dtantsur, cheers for the feedback I will update the patch after finishing one test here | 11:04 |
| dtantsur | I got my instance deployed \o/ | 11:05 |
| *** romcheg has quit IRC | 11:06 | |
| lucasagomes | dtantsur, yay! | 11:06 |
| dtantsur | to shut it down I should use nova delete, right? | 11:07 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Better check for available nodes https://review.openstack.org/84687 | 11:08 |
| *** openstackgerrit has quit IRC | 11:08 | |
| lucasagomes | dtantsur, yeah | 11:08 |
| *** openstackgerrit has joined #openstack-ironic | 11:08 | |
| *** romcheg has joined #openstack-ironic | 11:11 | |
| *** matsuhashi has joined #openstack-ironic | 11:16 | |
| *** lucasagomes is now known as lucas-hungry | 11:23 | |
| dtantsur | Guys, do you think bug https://bugs.launchpad.net/ironic/+bug/1199665 will be appropriate for a newbie? Or maybe some better ideas? I'm a bit stuck on where to begin :) | 11:28 |
| romcheg | dtantsur: You can try working on it. If you have problems, you can always ask here | 11:29 |
| romcheg | dtantsur: We will be glad to provide help | 11:29 |
| *** overlayer has joined #openstack-ironic | 11:29 | |
| *** overlayer has quit IRC | 11:32 | |
| *** killer_p- has quit IRC | 11:34 | |
| rameshg87 | Hello | 11:47 |
| rameshg87 | wanted to discuss regarding one of the bugs: https://bugs.launchpad.net/ironic/+bug/1277571 | 11:47 |
| yuriyz | morning Ironic | 11:50 |
| yuriyz | rameshg87, IMO this is WSME bug and should be fixed in WSME | 11:51 |
| yuriyz | and this WSME fix can broke another projects | 11:52 |
| Mikhail_D_ltp | yuriyz: g afternoon :) | 11:52 |
| yuriyz | hehe :) | 11:53 |
| *** pbrooko has quit IRC | 12:00 | |
| *** subah has quit IRC | 12:09 | |
| *** overlayer has joined #openstack-ironic | 12:11 | |
| rameshg87 | hello yuriyz: | 12:11 |
| rameshg87 | exactly, i was about to tell something like that | 12:12 |
| rameshg87 | ironic doesn't even get to know the non-existent properties passed | 12:12 |
| rameshg87 | wsme takes care of converting them to model objects and i cannot find anything we can do | 12:12 |
| *** linggao has joined #openstack-ironic | 12:13 | |
| *** rameshg87 has left #openstack-ironic | 12:16 | |
| *** killer_prince has quit IRC | 12:16 | |
| *** saju_m has quit IRC | 12:17 | |
| *** jistr is now known as jistr|english | 12:18 | |
| *** jdob has joined #openstack-ironic | 12:28 | |
| *** romcheg has quit IRC | 12:40 | |
| *** romcheg has joined #openstack-ironic | 12:41 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Better check for available nodes https://review.openstack.org/84687 | 12:44 |
| *** martyntaylor has quit IRC | 12:46 | |
| *** saju_m has joined #openstack-ironic | 12:47 | |
| *** martyntaylor has joined #openstack-ironic | 12:48 | |
| *** lucas-hungry is now known as lucasagomes | 12:49 | |
| *** rloo has joined #openstack-ironic | 12:53 | |
| *** pradipta is now known as pradipta_away | 12:59 | |
| *** matsuhashi has quit IRC | 13:02 | |
| *** matsuhashi has joined #openstack-ironic | 13:02 | |
| *** matsuhashi has quit IRC | 13:07 | |
| *** matsuhashi has joined #openstack-ironic | 13:10 | |
| *** matty_dubs|gone is now known as matty_dubs | 13:18 | |
| *** saju_m has quit IRC | 13:36 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Add DiskPartitioner https://review.openstack.org/83396 | 13:36 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Use DiskPartitioner https://review.openstack.org/83399 | 13:37 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Get rid of the swap partition https://review.openstack.org/83726 | 13:40 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Use GB instead of MB for swap https://review.openstack.org/83788 | 13:42 |
| romcheg | lucasagomes: w00t :) | 13:42 |
| lucasagomes | heh | 13:42 |
| lucasagomes | rebasing and making fixing the problem that dtantsur found when creating partition without leaving some lead space | 13:43 |
| openstackgerrit | Rohan Kanade proposed a change to openstack/ironic: Partially remove hard-coded ID's from testdata https://review.openstack.org/76990 | 13:43 |
| *** jistr|english is now known as jistr | 13:43 | |
| dtantsur | lucasagomes, have question about https://review.openstack.org/#/c/83399/9/ironic/drivers/modules/deploy_utils.py, lines 84-91 | 13:48 |
| dtantsur | what's the point of different ordering of swap and root partitions in two cases (with ephemeral and without)? | 13:48 |
| lucasagomes | dtantsur, oh it's for historic reasons that we have this differentiation | 13:52 |
| lucasagomes | dtantsur, it's being fixed in the next patch of the series | 13:53 |
| lucasagomes | https://review.openstack.org/#/c/83726/5/ironic/drivers/modules/deploy_utils.py | 13:53 |
| lucasagomes | to have a more consistent layout | 13:53 |
| lucasagomes | but I'm trying to make each patch fixing specific problems instead of tackling everything in one patch | 13:53 |
| lucasagomes | to make it easy for reviews | 13:53 |
| lucasagomes | reviewers | 13:53 |
| dtantsur | ah, ok | 13:57 |
| *** jgrimm has joined #openstack-ironic | 13:59 | |
| NobodyCam | good morning Ironic | 14:11 |
| lucasagomes | morning NobodyCam :D | 14:12 |
| NobodyCam | hey hey lucasagomes how goes hump day | 14:12 |
| lucasagomes | NobodyCam, all is well over here, urself? | 14:13 |
| lucasagomes | juno opened yay | 14:13 |
| dtantsur | NobodyCam, morning | 14:14 |
| NobodyCam | morning dtantsur | 14:14 |
| NobodyCam | lucasagomes: doing good here. just waiting for the coffee | 14:14 |
| lucasagomes | NobodyCam, :) | 14:15 |
| lucasagomes | NobodyCam, http://googlefiberblog.blogspot.ie/2014/04/introducing-coffee-to-home.html (/me wish it wasn't an april fools joke) | 14:15 |
| matty_dubs | Haha | 14:16 |
| NobodyCam | morning matty_dubs :) | 14:18 |
| NobodyCam | lucasagomes: lol | 14:20 |
| NobodyCam | the coffee spouts are to low on the wall | 14:20 |
| lucasagomes | heh yeah | 14:21 |
| NobodyCam | i love it | 14:21 |
| lucasagomes | well better than having no coffee :P | 14:21 |
| lucasagomes | NobodyCam, the best one I've seem so far was https://issues.apache.org/jira/browse/INFRA-7524 | 14:21 |
| NobodyCam | this is trey | 14:21 |
| lucasagomes | move the SVN project to use GIT :P | 14:21 |
| lucasagomes | hah | 14:21 |
| lucasagomes | I mean maintain the SVN code in a git repo | 14:21 |
| lucasagomes | hah | 14:22 |
| NobodyCam | lol | 14:22 |
| lucasagomes | mark showed me that one today | 14:22 |
| lucasagomes | the comments fun | 14:22 |
| lucasagomes | r fun* | 14:22 |
| NobodyCam | lol the whole idea is great | 14:23 |
| lucasagomes | yeah | 14:23 |
| NobodyCam | ahh I should have logged in to hot mail yesterday.. http://mashable.com/2014/04/01/clippy-is-back/ | 14:27 |
| lucasagomes | hah I'm surprised hot mail still exist! | 14:28 |
| NobodyCam | :-p | 14:29 |
| *** dwalleck has joined #openstack-ironic | 14:31 | |
| *** Mikhail_D_ltp has quit IRC | 14:37 | |
| *** pbrooko has joined #openstack-ironic | 14:39 | |
| * NobodyCam gets coffee | 14:49 | |
| *** coolsvap has joined #openstack-ironic | 14:49 | |
| *** matsuhashi has quit IRC | 14:52 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/python-ironicclient: node-list to show the maintenance field https://review.openstack.org/84742 | 14:59 |
| *** dwalleck_ has joined #openstack-ironic | 15:01 | |
| NobodyCam | lucasagomes: retesting the parted patches | 15:02 |
| lucasagomes | NobodyCam, :) ack! | 15:02 |
| lucasagomes | NobodyCam, ta much for that | 15:02 |
| NobodyCam | :) do I need to test with large swap? | 15:03 |
| NobodyCam | running now with no swap | 15:03 |
| lucasagomes | NobodyCam, I have tested it with large swap and no swap | 15:03 |
| lucasagomes | both are fixes | 15:03 |
| lucasagomes | dtantsur, ^ adding maintenance to the cli | 15:04 |
| NobodyCam | :) | 15:05 |
| devananda | morning, all | 15:09 |
| NobodyCam | good morning devananda | 15:09 |
| lucasagomes | devananda, morning! | 15:11 |
| lucasagomes | devananda, can u remove some -2s from https://review.openstack.org/#/c/78787/ and https://review.openstack.org/#/c/82773/ now that Juno is open pls :D | 15:12 |
| *** romcheg has quit IRC | 15:12 | |
| NobodyCam | lucasagomes: are those on the list ? | 15:13 |
| lucasagomes | NobodyCam, which list? | 15:13 |
| NobodyCam | https://etherpad.openstack.org/p/IronicWhiteBoard (Patch requested for re-submission now that Juno is Open:) | 15:13 |
| NobodyCam | on the bottom | 15:14 |
| NobodyCam | I don't seee them | 15:14 |
| NobodyCam | humm | 15:14 |
| lucasagomes | I think not | 15:14 |
| lucasagomes | :/ | 15:14 |
| NobodyCam | I missed them | 15:14 |
| devananda | lucasagomes: yep yep | 15:14 |
| lucasagomes | should they be there? | 15:14 |
| lucasagomes | NobodyCam, no worries | 15:15 |
| devananda | ah, NobodyCam, you probably looked at the list of abandoned things only ? | 15:15 |
| devananda | whoa.... the etherpad is in chinese! | 15:15 |
| NobodyCam | devananda: si | 15:16 |
| NobodyCam | :( | 15:16 |
| *** ifarkas has quit IRC | 15:16 | |
| lucasagomes | devananda, hah yeah I saw those chinese letters there :P | 15:16 |
| NobodyCam | ya etherpad needs i18n support | 15:16 |
| lucasagomes | I translated some, no bad words heh | 15:16 |
| devananda | wait. really? | 15:17 |
| devananda | i was about to look at my locale settings | 15:17 |
| lucasagomes | NobodyCam, lol yeah | 15:18 |
| lucasagomes | I just used google translate | 15:18 |
| NobodyCam | :) | 15:18 |
| rloo | cool. ha ha, how did that happen? | 15:19 |
| *** mtaylor is now known as mordred | 15:19 | |
| *** mordred has quit IRC | 15:19 | |
| *** mordred has joined #openstack-ironic | 15:19 | |
| lucasagomes | lol idk someone just added chinese letters there, I suspect it was devananda cause I know he speaks it (jk) | 15:20 |
| NobodyCam | I almost thought it was april 1st joke | 15:20 |
| devananda | heh, i do, poorly, and I was testing localization for zh_CN yesterday | 15:20 |
| devananda | it's a good joke. i wish i could take credit :) | 15:21 |
| *** ifarkas has joined #openstack-ironic | 15:21 | |
| *** openstackgerrit has quit IRC | 15:21 | |
| lucasagomes | heh | 15:21 |
| *** openstackgerrit has joined #openstack-ironic | 15:22 | |
| rloo | lucasagomes: wrt https://review.openstack.org/#/c/83726/ (get rid of swap partition), do you know what is being done in nova baremetal? | 15:22 |
| devananda | I'm going to revert the pad, then copy NobodyCam's changes forward | 15:22 |
| lucasagomes | we didn't have any april fools joke in Ironic, we should have come up with something... idk a driver to turn on/off the boiler of my apartment | 15:22 |
| lucasagomes | rloo, to remove the swap? I suspect none | 15:22 |
| openstackgerrit | A change was merged to openstack/ironic: Clean up IronicNodeStates.update_from_compute_node https://review.openstack.org/83852 | 15:23 |
| NobodyCam | lucasagomes: we did https://review.openstack.org/#/c/84474/ | 15:23 |
| rloo | lucasagomes: ok. | 15:23 |
| lucasagomes | NobodyCam, lol | 15:23 |
| NobodyCam | :-p | 15:23 |
| lucasagomes | rloo, do you think we should port it ? | 15:24 |
| *** pbrooko has quit IRC | 15:24 | |
| rloo | lucasagomes: no, it doesn't break anything, right? just adds a 1mb swap. | 15:24 |
| *** pbrooko has joined #openstack-ironic | 15:24 | |
| lucasagomes | rloo, yeah, pretty much | 15:24 |
| NobodyCam | lucasagomes: no i do not think we should port back unless its a securty issue | 15:25 |
| rloo | lucasagomes: there's some sort of ordering to your reviews... need to find out which to look at first ;) | 15:25 |
| lucasagomes | rloo, the first one is one -1'ed I will fix it, it's just a small nit | 15:25 |
| lucasagomes | rloo, https://review.openstack.org/#/c/83396/ | 15:26 |
| lucasagomes | NobodyCam, +1 | 15:26 |
| rloo | lucasagomes. yeah, am staring at it now. wondering if i should skip your reviews and look at some other one. or maybe take a break ;) | 15:26 |
| lucasagomes | :) | 15:26 |
| NobodyCam | rloo: I'm testng the parted patches atm.. | 15:28 |
| rloo | thx NobodyCam. (I don't usually test when I review something; should I?) | 15:28 |
| NobodyCam | rloo: I do depending on the type of change... ie. I do not for doc string changes | 15:29 |
| devananda | rloo: depending on the change. I often do. | 15:29 |
| NobodyCam | but for something major like this I do! | 15:29 |
| lucasagomes | rloo, depends, some people say yes, some people say no... I try to test it when it does touch some critical part of the code | 15:29 |
| Shrews | i test almost everything i review. it's amazing how many "gotchas" i've found by doing so | 15:30 |
| rloo | ok, thx. I'll keep that in mind. I guess if someone (other than the coder) has tested it, that might be sufficient. | 15:30 |
| NobodyCam | Shrews: ++ there is nothing wrong with that!!!! | 15:30 |
| rloo | Shrews: that is interesting. which means that the unit tests might not be that complete/correct (depending on the change of course) | 15:32 |
| Shrews | rloo: it's hard to unit test certain interactions | 15:32 |
| NobodyCam | rloo: I would say that our unit test are not 100% | 15:32 |
| Shrews | that goes to more of functional testing | 15:32 |
| Shrews | rloo: but i was mainly refering to other projects where i was core :) | 15:33 |
| rloo | Shrews: oh, so ironic is great as far as you're concerned? :-) | 15:34 |
| Shrews | rloo: ironic is da bomb | 15:34 |
| *** ilives has joined #openstack-ironic | 15:34 | |
| rloo | Shrews: did we say, 'welcome to ironic'! | 15:34 |
| * NobodyCam reads Shrews' comment and grins | 15:34 | |
| Shrews | i'm just glad i get to annoy devananda on a professional basis once again | 15:35 |
| *** pbrooko has quit IRC | 15:35 | |
| rloo | Shrews: yes, it is important that you are happy here :-) | 15:36 |
| * Shrews hugs rloo | 15:36 | |
| *** romcheg has joined #openstack-ironic | 15:37 | |
| *** romcheg has left #openstack-ironic | 15:37 | |
| Shrews | speaking of reviews.... omg, there are soooo many lately | 15:38 |
| *** pbrooko has joined #openstack-ironic | 15:41 | |
| NobodyCam | devananda: just added a second +2 to 83853 but I want to pull it and test befor I +a... will do that in the next 10 or so minutes | 15:43 |
| devananda | Shrews: I'm glad you get to annoy me again too :) | 15:44 |
| *** matty_dubs is now known as matty_dubs|lunch | 15:45 | |
| *** eghobo has joined #openstack-ironic | 15:51 | |
| openstackgerrit | Rohan Kanade proposed a change to openstack/ironic: Partially remove hard-coded ID's from testdata https://review.openstack.org/76990 | 15:53 |
| Shrews | russellb: how often are the openstack-stats pages updated? once a week? | 15:56 |
| russellb | more than once a day | 15:58 |
| russellb | unless it's broken | 15:58 |
| russellb | which it appears to be | 15:58 |
| * russellb kicks it | 15:59 | |
| Shrews | ah, thx | 15:59 |
| lucasagomes | Shrews, you can also check it at http://www.stackalytics.com/report/contribution/ironic/30 | 16:00 |
| Shrews | lucasagomes: ah, i did not know that. gracias | 16:01 |
| lucasagomes | :) | 16:01 |
| Shrews | looking more the oldest active reviews. is that hosted as well? | 16:01 |
| Shrews | more for* | 16:02 |
| *** killer_prince has joined #openstack-ironic | 16:02 | |
| russellb | should be pretty easy to run the scripts manually too btw | 16:02 |
| russellb | http://git.openstack.org/cgit/openstack-infra/reviewstats | 16:02 |
| russellb | install that | 16:02 |
| russellb | then run $ openreviews -p projects/ironic.json | 16:02 |
| russellb | or whatever | 16:02 |
| Shrews | awesomeness | 16:03 |
| lucasagomes | Shrews, you can change that '30' at the end of the URL for the num of days u want | 16:03 |
| NobodyCam | gah /me needs to remember to install tox 1.6 not 1.7 :-p | 16:03 |
| lucasagomes | http://www.stackalytics.com/report/contribution/ironic/180 | 16:03 |
| lucasagomes | or whatever num of days u might want to see | 16:03 |
| NobodyCam | lucasagomes: can I see how many reviews I'll do tomorrow? :-p | 16:04 |
| Shrews | lucasagomes: i wanted the equivalent of this: http://russellbryant.net/openstack-stats/ironic-openreviews.html | 16:04 |
| NobodyCam | http://www.stackalytics.com/report/contribution/ironic/-1 | 16:04 |
| Shrews | but i can get that manually, apparently :) | 16:04 |
| lucasagomes | NobodyCam, sure u can! | 16:04 |
| *** Vic has joined #openstack-ironic | 16:05 | |
| lucasagomes | NobodyCam, check how many reviews u did today, buy one of these: http://upload.wikimedia.org/wikipedia/commons/thumb/b/b7/Delorean_dmc12_rear.jpg/280px-Delorean_dmc12_rear.jpg | 16:06 |
| lucasagomes | goes back in time 1 day | 16:06 |
| lucasagomes | and booom! | 16:06 |
| *** dwalleck_ has quit IRC | 16:06 | |
| devananda | NobodyCam: so I can not un-minus-two an abandoned BP | 16:08 |
| NobodyCam | doh :( | 16:08 |
| devananda | NobodyCam: if you're poking ppl to revive their work, they need to do so before I can un-minus-two it | 16:08 |
| NobodyCam | ack.... I ping'd several yesterday | 16:09 |
| JayF | I'll poke jroll and russell_h when they get in to reopen their stuff in gerrit | 16:09 |
| NobodyCam | thank you JayF :) | 16:10 |
| openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Add worker threads limit to _check_deploy_timeouts task https://review.openstack.org/78651 | 16:10 |
| comstud | devananda: another question | 16:11 |
| devananda | hmm | 16:11 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Add DiskPartitioner https://review.openstack.org/83396 | 16:11 |
| *** coolsvap has quit IRC | 16:11 | |
| * devananda eyes romcheg's -2 on comstud's patch with curiosity | 16:11 | |
| comstud | or anyone that can answer the Node API class stuff.. | 16:11 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Use DiskPartitioner https://review.openstack.org/83399 | 16:11 |
| comstud | https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L327 | 16:11 |
| devananda | comstud: shoot | 16:11 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Get rid of the swap partition https://review.openstack.org/83726 | 16:11 |
| comstud | This happens to add 'id' as a field to Node | 16:11 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Use GB instead of MB for swap https://review.openstack.org/83788 | 16:11 |
| devananda | lucasagomes: that ^ might be a better question for you? | 16:11 |
| NobodyCam | lucasagomes: I just test those...:-p | 16:12 |
| lucasagomes | devananda, checking | 16:12 |
| comstud | it causes a problem for me becaues .as_dict() includes 'id' set to None.. | 16:12 |
| *** ifarkas has quit IRC | 16:12 | |
| NobodyCam | and they passed | 16:12 |
| lucasagomes | NobodyCam, I just changed a small thing in the get_partitions() | 16:12 |
| lucasagomes | NobodyCam, it was -1'd, nothing critical | 16:12 |
| NobodyCam | :) | 16:12 |
| comstud | and I try to set attributes on the objects/node.py:Node object.. and of course, None isn't allowed | 16:12 |
| comstud | trying to figure out if that should actually be in the API Node.. | 16:12 |
| comstud | or if I'll just need to ignore it | 16:12 |
| lucasagomes | comstud, ack I'm a bit lost I think I need a bit more of context | 16:13 |
| comstud | (converting node_create to object Node.create()) | 16:13 |
| lucasagomes | but the id/uuid stuff there | 16:13 |
| lucasagomes | in the api, we always use uuid, node_uuid for e.g | 16:13 |
| comstud | so, what I"m doing is changing the dbapi.node_create() call | 16:13 |
| comstud | to: | 16:13 |
| devananda | comstud: 'id' should never be sent in our API | 16:13 |
| comstud | node = objects.Node(context, api_node_thingy.as_dict()) | 16:13 |
| comstud | node.create() | 16:13 |
| lucasagomes | but internally we use node_id (for performance reasons) | 16:13 |
| comstud | Sure, I got that. | 16:14 |
| lucasagomes | comstud, right | 16:14 |
| comstud | But this is the API object which seems to define the fields allowed for someone to send with POST | 16:14 |
| comstud | and 'id' is in fields | 16:14 |
| lucasagomes | and the api_node_thingy does have a node_id field? | 16:14 |
| *** pbrooko has quit IRC | 16:14 | |
| comstud | lucasagomes: yeah, as 'id' | 16:14 |
| comstud | because it uses objects.Node.keys() to populate fields | 16:14 |
| lucasagomes | comstud, right, so 'id' is not part of the Nodes API object | 16:14 |
| comstud | it is | 16:15 |
| lucasagomes | that's why it's failing | 16:15 |
| comstud | but it shouldn't be ? | 16:15 |
| comstud | :) | 16:15 |
| lucasagomes | it's not /me checks | 16:15 |
| comstud | it is | 16:15 |
| comstud | see paste above | 16:15 |
| comstud | or here: | 16:15 |
| comstud | https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L327 | 16:15 |
| comstud | objects.Node.fields.keys() includes 'id' | 16:15 |
| lucasagomes | comstud, https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L266-L324 | 16:16 |
| comstud | so something can be in self.fields but if there's no class attribute, it's not allowed? | 16:16 |
| NobodyCam | bbt...brb | 16:16 |
| *** coolsvap has joined #openstack-ironic | 16:17 | |
| *** eghobo has quit IRC | 16:17 | |
| comstud | the side effect is: | 16:17 |
| comstud | https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L621 | 16:17 |
| *** eghobo has joined #openstack-ironic | 16:17 | |
| comstud | that node.as_dict() includes {'id': None} when POSTing | 16:17 |
| lucasagomes | ahh, hmmm | 16:18 |
| comstud | because id is in fields | 16:18 |
| lucasagomes | I see the prob, yeah it's messy | 16:18 |
| comstud | so, I have 2 ways to fix this | 16:18 |
| comstud | pop it out before I create the node in the DB | 16:18 |
| comstud | or | 16:18 |
| *** coolsvap has quit IRC | 16:18 | |
| comstud | not include it at all in self.fields | 16:18 |
| comstud | I think the latter is the more correct fix | 16:19 |
| comstud | which I'm about to try | 16:19 |
| lucasagomes | comstud, I agree with the later as well | 16:20 |
| comstud | ok | 16:20 |
| comstud | of course, that causes other tests to fail | 16:20 |
| comstud | hehe | 16:21 |
| comstud | will figure out what's up | 16:21 |
| comstud | This API testing stuff is annoying because I can't get a traceback from where an exception occurs | 16:21 |
| comstud | AppError: Bad response: 500 Internal Server Error (not 200 OK or 3xx redirect for http://localhost/v1/nodes/1be26c0b-03f2-4d2e-ae87-c02d7f33c123) | 16:22 |
| comstud | '{"error_message": "{\\"debuginfo\\": null, \\"faultcode\\": \\"Server\\", \\"faultstring\\": \\"\'Node\' object has no attribute \'id\'\\"}"}' | 16:22 |
| comstud | this is less than useful. | 16:22 |
| lucasagomes | comstud, I think that 'id' might be required when it get's objectified from the db | 16:22 |
| comstud | It is, but that's not the problem here | 16:22 |
| comstud | our db object stubs are fine | 16:23 |
| comstud | This is something trying to set Node.id | 16:23 |
| comstud | ah | 16:23 |
| comstud | maybe I know what you mean | 16:23 |
| comstud | the response forming may be the problem here | 16:23 |
| lucasagomes | comstud, maybe the easiest way would be to pop out the 'id' field from the objects when instantiating the Node object from the api | 16:23 |
| lucasagomes | https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L327 | 16:23 |
| lucasagomes | pop out I mean ignore the id field | 16:24 |
| comstud | That's where I'm popping it | 16:24 |
| comstud | I mean, ignoring it | 16:24 |
| lucasagomes | yeah | 16:24 |
| lucasagomes | that might be a better fix, cause that ironic/object/* is more related the the db objects | 16:24 |
| comstud | That's what I was suggesting in the first place | 16:25 |
| lucasagomes | and the problem is that the API object doesn't have the id field | 16:25 |
| comstud | the problem is, the tests seem to be trying to set Node.id | 16:25 |
| comstud | and now it no longer exists | 16:25 |
| comstud | ah, you're suggesting I still add it to the API Node object? | 16:25 |
| lucasagomes | comstud, heh no, was suggesting you to just ignore it when instantiating the API Node object | 16:26 |
| comstud | I am | 16:26 |
| comstud | but something is trying to set it, it seems | 16:26 |
| comstud | maybe not | 16:26 |
| comstud | i have to find the real traceback that the API tests are covering up. | 16:27 |
| comstud | :-/ | 16:27 |
| lucasagomes | :/ | 16:27 |
| comstud | I keep having to put traceback.print_stack() in places | 16:27 |
| lucasagomes | lemme see if I can find something if I start ignoring that field | 16:27 |
| comstud | because the API tests cover up the real traceback | 16:27 |
| comstud | and for some reason I don't have any LOG output in tox | 16:27 |
| lucasagomes | oh I see :( | 16:30 |
| lucasagomes | hmm | 16:30 |
| comstud | lol, something weird is going on, and i'm not sure what yet | 16:30 |
| *** derekh has quit IRC | 16:32 | |
| *** jistr has quit IRC | 16:33 | |
| *** stevehuang has joined #openstack-ironic | 16:35 | |
| lucasagomes | comstud, ok found it | 16:37 |
| comstud | 654 # Update only the fields that have changed | 16:37 |
| comstud | 655 for field in objects.Node.fields: | 16:37 |
| comstud | 656 if rpc_node[field] != getattr(node, field): | 16:37 |
| comstud | 657 rpc_node[field] = getattr(node, field) | 16:37 |
| lucasagomes | comstud, hacky http://paste.openstack.org/show/74876/ | 16:37 |
| lucasagomes | yeah | 16:37 |
| lucasagomes | exactly | 16:37 |
| comstud | yeah | 16:37 |
| comstud | your patch looks almost exactly like mine | 16:37 |
| comstud | well | 16:38 |
| lucasagomes | heh | 16:38 |
| comstud | the first part | 16:38 |
| comstud | hm | 16:38 |
| lucasagomes | problem found, now needs to make it look better than that patch :/ | 16:39 |
| comstud | 656 if field not in node.fields: | 16:39 |
| comstud | 657 continue | 16:39 |
| comstud | that's my fix to the 2nd part | 16:39 |
| lucasagomes | I know it's a bit messy having 2 diff types of objects | 16:39 |
| lucasagomes | comstud, looks better than mine | 16:39 |
| comstud | I don't think it's terribly messy | 16:40 |
| lucasagomes | yeah not terrible | 16:40 |
| comstud | but there's some assumptions that they have the same fields | 16:40 |
| lucasagomes | but confusing | 16:40 |
| comstud | which shouldn't be in here | 16:40 |
| comstud | it gets confusing to talk about | 16:40 |
| lucasagomes | comstud, +2! | 16:40 |
| comstud | because I can't just say 'Node object' | 16:40 |
| comstud | I have to clarify which one | 16:40 |
| comstud | :) | 16:40 |
| lucasagomes | yes | 16:40 |
| lucasagomes | exactly! | 16:40 |
| comstud | anyway, cool, that fixed my problem | 16:40 |
| comstud | now I just have a general question so I can understand this better | 16:41 |
| *** martyntaylor has left #openstack-ironic | 16:41 | |
| comstud | so this *API* Node object... | 16:41 |
| lucasagomes | sure | 16:41 |
| comstud | so we set self.fields | 16:41 |
| comstud | types are declared as class attributes | 16:41 |
| *** Vic has quit IRC | 16:41 | |
| comstud | does this mean... | 16:41 |
| *** blamar_ has joined #openstack-ironic | 16:41 | |
| comstud | If you have something in self.fields but NOT have a class attribute set for it.. | 16:41 |
| comstud | It won't be allowed in the API request? | 16:42 |
| *** blamar has quit IRC | 16:42 | |
| *** blamar_ is now known as blamar | 16:42 | |
| comstud | I'm trying to figure out if this accidentally allowed you to pass a 'id' in the API | 16:42 |
| lucasagomes | comstud, yeah, well wsme will ignore it if it's not in the body of the object class | 16:42 |
| *** ilives has quit IRC | 16:42 | |
| lucasagomes | comstud, it won't even be passed to us | 16:42 |
| comstud | what if it WERE in the body | 16:42 |
| comstud | ? | 16:43 |
| comstud | what would happen? would it set it, because it's in self.fields ? | 16:43 |
| lucasagomes | comstud, then wsme will pass it to us and we probably would have problems with it heh | 16:43 |
| comstud | or would it not, because there's no type declared for it | 16:43 |
| comstud | yeah okay :) so this really is a bug | 16:43 |
| *** ndipanov has quit IRC | 16:43 | |
| comstud | and not just some internal messiness | 16:44 |
| lucasagomes | comstud, yeah, well id is not in the body class so if you POST it to the API | 16:44 |
| lucasagomes | it will be ignored | 16:44 |
| comstud | ok, i misunderstood | 16:44 |
| comstud | what you meant by 'body' | 16:44 |
| comstud | gotcha | 16:44 |
| lucasagomes | yeah body I mean the attributes of the Node class in the API | 16:44 |
| comstud | yep | 16:45 |
| comstud | ok | 16:45 |
| comstud | that's exactly what I was attempting to ask | 16:45 |
| comstud | thanks | 16:45 |
| lucasagomes | np | 16:45 |
| *** martyntaylor has joined #openstack-ironic | 16:46 | |
| comstud | i think i'm going to fix this slightly differently to avoid list scan | 16:47 |
| comstud | with a try/except | 16:47 |
| comstud | but anyway | 16:47 |
| lucasagomes | comstud, :) add me in the patch as reviewer pls | 16:48 |
| comstud | will do | 16:49 |
| russell_h | if I somehow wound up reordering some imports, could that cause the new sample config to be totally empty? | 16:50 |
| NobodyCam | devananda: 83853 gives me a create failed on deploying... looking in to logs to see why | 16:50 |
| devananda | russell_h: import order is enforced by pep ... | 16:51 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding a reference driver for the agent https://review.openstack.org/84795 | 16:52 |
| NobodyCam | devananda: http://paste.openstack.org/show/Rnl9imi9Hj5qJjSDOo2P/ | 16:55 |
| NobodyCam | note lines 10-12 | 16:56 |
| devananda | NobodyCam: sure. extra_specs fails | 16:57 |
| russell_h | devananda: I mean more in terms of changing the dependency graph | 17:02 |
| russell_h | devananda: specifically, I imported ironic.dbapi from ironic.common.hash_ring | 17:02 |
| russell_h | which I suspect is breaking things | 17:03 |
| devananda | russell_h: oh. yes. you could get a circular import | 17:03 |
| russell_h | in this case the code actually runs fine, its just that when pep8 tries to verify the sample config it discovers that it should now be entirely empty | 17:03 |
| russell_h | I'll sort it out | 17:04 |
| lucasagomes | right I'm done for today | 17:05 |
| lucasagomes | have a good night everybody :) | 17:05 |
| NobodyCam | have a good night lucasagomes :) | 17:05 |
| *** lucasagomes is now known as lucas-afk | 17:06 | |
| NobodyCam | devananda: 83853 requires extra_specs changes? | 17:06 |
| devananda | NobodyCam: interesting. it shouldnt | 17:09 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Refactor nova.virt.ironic.driver get_host_stats https://review.openstack.org/83853 | 17:09 |
| *** eguz has joined #openstack-ironic | 17:09 | |
| NobodyCam | humm my testing is "devtest.sh --trash-my-machine" | 17:10 |
| NobodyCam | bbaifm ... post bbt walkies | 17:11 |
| *** eghobo has quit IRC | 17:14 | |
| devananda | JayF, jroll - question on the ipa-core reviewer group. I think it makes sense to include ironic-core. any objections? | 17:15 |
| devananda | as a subproject, ipa could have separate core reviewers, but anyone with core on ironic automaticaly has core status on ipa as well. | 17:16 |
| *** killer_prince has quit IRC | 17:16 | |
| JayF | I thought of it as somewhat of a positive that different groups of people had to approve stuff for the agent and agent driver | 17:17 |
| JayF | sort of a dual-layer of acceptance | 17:17 |
| JayF | but I don't have any real opposition to that | 17:17 |
| JayF | russell_h: ^ ? | 17:17 |
| NobodyCam | and back | 17:17 |
| devananda | JoshNang: russel_h: can you pls run, in some project, "git remote -v | grep gerrit" and tell me the user name? | 17:20 |
| JayF | devananda: I already checked with russell_h and JoshNang | 17:20 |
| JoshNang | devananda: JoshNang | 17:20 |
| devananda | nvm :) | 17:20 |
| devananda | cool | 17:20 |
| russell_h | devananda: should we just merge the review teams? | 17:20 |
| russell_h | you can kick me off | 17:20 |
| russell_h | or not, but jroll JoshNang and JayF are going to be more active than I am | 17:22 |
| devananda | actually, probably a good idea | 17:24 |
| devananda | will need to vet it with teh current core team | 17:24 |
| devananda | but i doubt anyone will object to having more review bandwidth :) | 17:24 |
| russell_h | awesome | 17:25 |
| devananda | NobodyCam: ooh, i think I see the problem | 17:27 |
| devananda | hypervisor_type/version are missing now | 17:27 |
| jroll | late to the party but +1 on merging | 17:28 |
| devananda | NobodyCam: nope, nvm. they're there | 17:29 |
| NobodyCam | devananda: also jsonutils.dumps is removed from supported_instances | 17:29 |
| *** matty_dubs|lunch is now known as matty_dubs | 17:29 | |
| NobodyCam | ya they are there | 17:29 |
| *** harlowja_away is now known as harlowja | 17:35 | |
| *** dividebin has joined #openstack-ironic | 17:37 | |
| *** dividehex has quit IRC | 17:38 | |
| *** dividebin is now known as dividehex | 17:38 | |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding a reference driver for the agent https://review.openstack.org/84795 | 17:41 |
| NobodyCam | oh... | 17:42 |
| NobodyCam | nice | 17:42 |
| NobodyCam | agent only works with real ipmi? | 17:43 |
| jroll | NobodyCam: ? | 17:43 |
| NobodyCam | https://review.openstack.org/#/c/84795/2/setup.cfg | 17:43 |
| JoshNang | Last time I tested, native wouldn't connect to our hardware. But I includeed hooks for it | 17:43 |
| NobodyCam | no agent_ssh | 17:44 |
| jroll | mmm | 17:44 |
| jroll | I haven't actually looked at the ssh driver | 17:44 |
| JoshNang | hmm neither have i | 17:44 |
| jroll | what does it do? | 17:44 |
| jroll | does the ssh driver just issue reboot commands? | 17:44 |
| * NobodyCam notes he can only test with ssh as he only has two laptops | 17:45 | |
| NobodyCam | jroll: ??? it provides on/off/reboot | 17:45 |
| * jroll looks | 17:45 | |
| jroll | curious how you ssh in to power on a box | 17:45 |
| JoshNang | using vbox it looks like | 17:46 |
| NobodyCam | jroll: vm's only | 17:46 |
| JoshNang | or vmware | 17:46 |
| jroll | ohhh | 17:46 |
| jroll | I see | 17:46 |
| jroll | yeah | 17:46 |
| NobodyCam | I use virsh | 17:46 |
| jroll | I mean, it should be straightforward to make it work, yeah? | 17:46 |
| NobodyCam | ya | 17:46 |
| jroll | put it on the review :) | 17:46 |
| JoshNang | i'll work on adding it righ tnow | 17:46 |
| NobodyCam | its just powercontrol for vm's | 17:46 |
| jroll | right | 17:47 |
| NobodyCam | sweet :) | 17:47 |
| NobodyCam | Ty jroll | 17:47 |
| JoshNang | np. don't want to impede testing. | 17:47 |
| jroll | hm, that might be nice for local testing | 17:47 |
| * jroll will investigate this today | 17:47 | |
| NobodyCam | :) it is required for ci testing as that is all vm based | 17:47 |
| jroll | right right | 17:48 |
| jroll | we were talking the other day about making a full agent environment through vagrant | 17:48 |
| NobodyCam | heheh use tripleO | 17:48 |
| NobodyCam | its the way to go :-p | 17:48 |
| jroll | :P | 17:48 |
| JoshNang | :) | 17:49 |
| devananda | JoshNang: russell_h: you guys are showing up as "independent" in review stats | 17:49 |
| devananda | http://git.openstack.org/cgit/openstack/governance/tree/reference/programs.yaml#n137 | 17:49 |
| jroll | NobodyCam: we have a lab, but I want a virtual lab in my desktop | 17:49 |
| NobodyCam | :) jroll as I work out of the RV my lab is a hp laptop | 17:50 |
| jroll | JoshNang, russell_h: do something like this: https://review.openstack.org/#/c/83239/ | 17:50 |
| jroll | NobodyCam: :D | 17:50 |
| NobodyCam | which I have to stand upside down or it overheats | 17:50 |
| jroll | :| | 17:51 |
| NobodyCam | :-p | 17:51 |
| * jroll tells hp to buy NobodyCam a proper lab | 17:51 | |
| JoshNang | i've got mine in waiting review | 17:52 |
| NobodyCam | lol .. I will say it is a nice brick *cough* i mean laptop... | 17:52 |
| jroll | heh | 17:52 |
| JayF | devananda: me and jroll updated our info in stackalytics, and I know JoshNang was working on that as well ... is there somewhere else we need to put it in? | 17:53 |
| openstackgerrit | Russell Haering proposed a change to openstack/ironic: Factor hash ring management out of the conductor https://review.openstack.org/84814 | 17:53 |
| JayF | It seems like there are three places for everything :) | 17:53 |
| jroll | JayF: no, that's it | 17:53 |
| JayF | russell_h: pquerna and JoshNang I think still need to update the stackalytics json file | 17:53 |
| *** max_lobur1 has quit IRC | 17:54 | |
| comstud | roman around? | 17:54 |
| comstud | dunno his nick | 17:54 |
| devananda | romcheg | 17:55 |
| comstud | so not in here | 17:56 |
| comstud | devananda: https://review.openstack.org/#/c/84573/ | 17:56 |
| comstud | -2 seems a little overkill here | 17:56 |
| devananda | yea | 17:56 |
| comstud | i'm certainly not waiting until next Monday to discuss at a meeting | 17:56 |
| devananda | right | 17:56 |
| comstud | hehe | 17:56 |
| comstud | so, I could break out where we're tesitng with dicts instead of objects | 17:57 |
| comstud | but I think I'll just be editing the same sections again anyway | 17:57 |
| comstud | maybe not, actually.. | 17:58 |
| comstud | i'll see what it looks like, but this patch is small | 17:58 |
| comstud | it just looks bad because of a lot of files touched | 17:58 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding a reference driver for the agent https://review.openstack.org/84795 | 17:59 |
| devananda | comstud: so i want to review this a bit ... the first file i looked in, you're adding a db query? | 17:59 |
| devananda | https://review.openstack.org/#/c/84573/5/ironic/api/controllers/v1/port.py | 17:59 |
| JoshNang | NobodyCam: I think that should cover SSH for power | 18:00 |
| comstud | devananda: It's moving a query | 18:00 |
| comstud | sec | 18:00 |
| comstud | which one, to make sure we're talking about the same thing | 18:00 |
| NobodyCam | :) JoshNang will look in a bit | 18:00 |
| comstud | right | 18:00 |
| comstud | devananda: get_ports_by_node() was doing the query before. This was unnecessary in a different case where we already had 'node' | 18:01 |
| comstud | so I moved it up a layer so you can just pass the node in | 18:01 |
| comstud | if you have it | 18:01 |
| russell_h | JoshNang: IMO we should merge a lot of these modules into one file | 18:01 |
| devananda | comstud: ah, gotcha | 18:01 |
| devananda | comstud: yea, the check for node-not-fond | 18:02 |
| comstud | the reason I changed it to begin with was... | 18:02 |
| NobodyCam | oh poop... wrong command | 18:02 |
| comstud | get_ports_by_node internally calling "get_node" which I removed in flavor of specific uuid/id versions | 18:02 |
| comstud | but then I found this happened to actually *remove* a DB query as well, so win/win imo | 18:03 |
| comstud | (this is #6 in my commit msg) | 18:04 |
| comstud | i can break that one out, but i'll just be re-touching it | 18:05 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Fix self.fields on API Node object https://review.openstack.org/84822 | 18:06 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Add create() and destroy() to Node https://review.openstack.org/84823 | 18:06 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Clean up calls to get_node() https://review.openstack.org/84573 | 18:06 |
| devananda | comstud: so I find this a bit awkward: ports = self.dbapi.get_ports_by_node(node) | 18:06 |
| devananda | passing a node object to the dbapi | 18:06 |
| comstud | it won't be to the dbapi for long | 18:06 |
| devananda | comstud: what about ports = objects.Port.get_by_node(node) | 18:06 |
| comstud | right | 18:07 |
| devananda | cool | 18:07 |
| comstud | that's the goal | 18:07 |
| devananda | great | 18:07 |
| comstud | I didn't want to mix ports and nodes | 18:07 |
| comstud | (changes) | 18:07 |
| devananda | sure | 18:07 |
| devananda | so i added a comment to rev5 | 18:07 |
| comstud | ok | 18:07 |
| devananda | just after you posted 6 | 18:07 |
| comstud | yeah, not much changed in 6 | 18:08 |
| devananda | I dont have time to review it all now, but I'll comment about the -2 | 18:08 |
| comstud | i changed the new test module I added slightly | 18:08 |
| comstud | and rebased to switch dependency order of patches | 18:08 |
| comstud | Created 'Fix self.fields on API Node object' and put it at top of my tree. | 18:09 |
| *** zul has quit IRC | 18:09 | |
| comstud | devananda: I saw that, also, re: your comment :) | 18:09 |
| comstud | I was going to do that in a separate patch | 18:09 |
| comstud | but can mix it in | 18:09 |
| devananda | comstud: separate is fine, too | 18:10 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding a reference driver for the agent https://review.openstack.org/84795 | 18:10 |
| comstud | i was like, 'why isn't provision_state' just in the list of columns | 18:10 |
| devananda | yep | 18:10 |
| comstud | anyway, cools | 18:10 |
| openstackgerrit | Jim Rollenhagen proposed a change to openstack/ironic: Add Node.instance_info field https://review.openstack.org/79466 | 18:10 |
| *** romcheg has joined #openstack-ironic | 18:11 | |
| *** zul has joined #openstack-ironic | 18:12 | |
| romcheg | devananda: are you around? | 18:13 |
| JoshNang | russell_h: I'm not opposed to merging those files (and their tests) into single files. | 18:14 |
| devananda | romcheg: hi! yes | 18:14 |
| romcheg | Do you think we can review and merge 84573 without splitting it into a series of patches? | 18:15 |
| *** Mikhail_D_ltp has joined #openstack-ironic | 18:15 | |
| devananda | romcheg: possibly. it's a lot of small changes in many files, but afaict they're all related changes | 18:16 |
| *** dkehn__ has joined #openstack-ironic | 18:19 | |
| romcheg | I will review it then and replace my -2 with an appropriate score | 18:19 |
| romcheg | Sorry for -2 | 18:19 |
| devananda | romcheg: even so, I wouldn't have -2'd for that | 18:19 |
| devananda | -1, yes :) | 18:20 |
| romcheg | I might have got the purpose of -2 wrong | 18:21 |
| devananda | romcheg: -2 is an un-overridable way of saying "dont merge this" | 18:22 |
| devananda | romcheg: either because of something ilke feature freeze (where it's a temporary -2) | 18:22 |
| devananda | romcheg: or the feature is being rejected permananently from the project | 18:22 |
| romcheg | So it's like "We don't want to see this change in our code"? | 18:23 |
| devananda | right | 18:23 |
| romcheg | Ah, now I understand | 18:23 |
| devananda | -1 with some feedback gives the author the ability to change the patch and resubmit | 18:23 |
| romcheg | I will replace it with -1 then | 18:23 |
| *** dkehn_ has quit IRC | 18:23 | |
| devananda | but -2 will not go away until you go remove it, no matter how many new patch sets the author uploads | 18:23 |
| romcheg | *after making more review | 18:24 |
| devananda | thanks! | 18:24 |
| romcheg | Sorry again | 18:24 |
| devananda | no worries. I think that was your first -2 :) | 18:24 |
| romcheg | yup | 18:25 |
| devananda | bbiab | 18:33 |
| NobodyCam | :) | 18:34 |
| NobodyCam | devananda: quick ? | 18:34 |
| NobodyCam | mind if I push ontop of 83853? | 18:34 |
| devananda | NobodyCam: nope | 18:35 |
| NobodyCam | :) | 18:35 |
| adam_g | Shrews, were you hitting some RAM/CPU exhaustion when running the ironic test suite recently? | 18:35 |
| Shrews | adam_g: indeed | 18:35 |
| Shrews | adam_g: 1 sec... | 18:36 |
| Shrews | adam_g: https://bugs.launchpad.net/ironic/+bug/1300420 | 18:36 |
| adam_g | Shrews, ever figure it out? i uploaded the RC to ubuntu last night and the test suite runs for 4 hours on the builder before its killed :) | 18:36 |
| adam_g | hhmm wonder if its related | 18:37 |
| Shrews | adam_g: i haven't taken time to explore it much yet | 18:37 |
| Shrews | adam_g: but i'm suspecting a race condition between tests | 18:37 |
| Shrews | which is... fun | 18:38 |
| NobodyCam | humm missed bagel time... brb | 18:40 |
| *** pradipta_away has quit IRC | 18:40 | |
| *** pradipta_away has joined #openstack-ironic | 18:42 | |
| *** max_lobur has joined #openstack-ironic | 18:46 | |
| Shrews | adam_g: if you are seeing the same thing, please comment on the bug. though i would think that if it was eating all memory, it would surely do it faster than 4 hours | 18:48 |
| adam_g | Shrews, yeah--im not sure, the ubuntu build systems don't really give you much info. the log just showed testr starting and hanging forever. i've seen that in other projects. im retrying the build, will maybe limit concurrency to 1 if it still fails | 18:49 |
| adam_g | devananda, is there a plan for merging the nova code back into nova? | 18:49 |
| comstud | lunch | 18:59 |
| * comstud & bbs | 18:59 | |
| *** rwsu has quit IRC | 19:01 | |
| NobodyCam | ya ... brb | 19:02 |
| *** overlayer has quit IRC | 19:02 | |
| matty_dubs | Say, are you folks hitting this? https://bitbucket.org/hpk42/tox/issue/150/posargs-configerror ? I learned of the 'tox -r' syntax in the recent meeting, as an alternative to rm -fr .venv and rebuilding. But it doesn't appear to work. | 19:04 |
| adam_g | matty_dubs, try tox==1.6.1 | 19:04 |
| *** adam_g has left #openstack-ironic | 19:05 | |
| *** adam_g has joined #openstack-ironic | 19:05 | |
| *** mrda_away is now known as mrda | 19:07 | |
| *** rwsu has joined #openstack-ironic | 19:07 | |
| matty_dubs | adam_g: Thanks, that seems to work. | 19:08 |
| * matty_dubs doesn't love downgrading, but it gets me unjammed | 19:08 | |
| devananda | adam_g: taht depends. is "merge it" a plan? :) | 19:09 |
| adam_g | devananda, i'm not sure? did the ironic code ever actually exist in nova? this all pre-dates me and i'm trying to piece it together | 19:09 |
| devananda | ah | 19:10 |
| devananda | so | 19:10 |
| devananda | the nova.virt.baremetal code is in nova trunk | 19:10 |
| adam_g | yup | 19:10 |
| devananda | functionally, nova.virt.ironic will replace it | 19:10 |
| devananda | but the actuall process for that is a bit complex | 19:10 |
| devananda | adam_g: does that answer your question? shall i sketch the process? | 19:11 |
| adam_g | devananda, is there an etherpad or something? i assume this is a goal for J? | 19:11 |
| devananda | https://etherpad.openstack.org/p/BaremetalDeprecationPlan | 19:12 |
| devananda | russellb: do you know of any notes on nova's side for the actual process of deprecating baremetal // merging ironic? | 19:12 |
| adam_g | devananda, we were discussing https://review.openstack.org/#/c/84033/ yesterday, and i began wondering if it made more sense to get the driver into nova now that Juno is open. rather than jerry-rigging infra to get the tests to run in ironic. but if its still a ways out, maybe it doesn't | 19:13 |
| russellb | devananda: nope | 19:14 |
| devananda | adam_g: probably worth discussing in -nova as we don't have any control over it | 19:14 |
| devananda | russellb: ack. I'll jot down what's in my head | 19:14 |
| devananda | russellb: while ironic is incubated, is it considered a third-party driver in nova? will that change when ironic graduates? | 19:19 |
| russellb | i think we can merge the driver with it incubated | 19:20 |
| russellb | IMO anyway | 19:20 |
| devananda | actually. lets move to -infra | 19:21 |
| devananda | as i need to ask about voting/gating specifically | 19:21 |
| adam_g | sure | 19:21 |
| *** overlayer has joined #openstack-ironic | 19:22 | |
| NobodyCam | devananda: has 83853 worked for you? | 19:22 |
| *** dwalleck has quit IRC | 19:41 | |
| *** dividehex has quit IRC | 19:44 | |
| *** dividehex has joined #openstack-ironic | 19:51 | |
| Shrews | NobodyCam: what isn't working for you? | 19:52 |
| NobodyCam | Shrews: I'm getting a extra_specs error from nova schudler | 19:52 |
| Shrews | yuriyz: 84142 works for me. +1'd | 19:53 |
| Shrews | NobodyCam: hrm. i can run it through a devstack test... and remove my "have not tested" caveat in my +1 | 19:54 |
| Shrews | :) | 19:54 |
| NobodyCam | testsing :) | 19:54 |
| NobodyCam | here too | 19:54 |
| devananda | NobodyCam: not sure if I tested it. might not have ... | 19:55 |
| Shrews | devananda: i'm on it. keep doing core-type stuff | 19:55 |
| devananda | thanks | 19:58 |
| Shrews | NobodyCam: waiting for devstack to do it's thing. say you see the error in the n-sch logs? | 19:59 |
| Shrews | anything you're doing to trigger it? | 19:59 |
| NobodyCam | http://paste.openstack.org/show/Rnl9imi9Hj5qJjSDOo2P/ | 20:00 |
| Shrews | k | 20:01 |
| * Shrews makes tea waiting for his slow-as-mud vm to do stuff | 20:02 | |
| linggao | Hi matty_dubs, | 20:10 |
| linggao | matty_dubs, what is the value for ipmi_terminal_port? | 20:10 |
| matty_dubs | linggao: Let me check what I put. I _think_ it's 4200, the default port for shellinaboxd. | 20:10 |
| matty_dubs | Yes, I have u'ipmi_terminal_port': u'4200' | 20:13 |
| linggao | ok, thanks | 20:13 |
| Shrews | NobodyCam: i got the same error. I -1'd it, but I'll see if I can fix it for devananda | 20:17 |
| russell_h | devananda: so this is basically the same change as I submitted last time, but it still instantiates a new HashRingManager for every API request: https://review.openstack.org/#/c/84814/ | 20:17 |
| russell_h | devananda: ie, it shouldn't change any functionality at all, just a refactoring to start | 20:17 |
| NobodyCam | :) | 20:18 |
| Shrews | devananda: if you don't mind me jumping in, that is | 20:18 |
| devananda | Shrews: not a bit | 20:18 |
| devananda | just for the record | 20:18 |
| devananda | nearly any time i put up a patch, ya'll are welcome to clobber it | 20:18 |
| devananda | in fact, I appreciate it when you do :) | 20:19 |
| comstud | noted | 20:19 |
| devananda | also, http://summit.openstack.org/cfp/details/215 | 20:20 |
| JayF | russell_h: I'm not 100% sure, but I think that non-voting tempest failure, after looking at the logs, might be a real issue in your hashring patch | 20:20 |
| russell_h | oh damn | 20:20 |
| devananda | yea, so it's pretty good to look at those virtual-ironic tempest jobs | 20:21 |
| devananda | even though they fail. they sometimes fail *differently* | 20:21 |
| JayF | Yeah I just saw the error was a 500 from ironic about a server not being available | 20:21 |
| JayF | and given the patch deals with finding a server for a request aiui, it seems relevant. | 20:22 |
| devananda | adam_g: since your question started the discussion, also for you http://summit.openstack.org/cfp/details/215 | 20:23 |
| adam_g | devananda, yes, sorry to push into a rabbit hole :) | 20:23 |
| adam_g | devananda, i think the short answer to my question is that we're still a ways off and it would be good to have as much testing running while we are still out-of-tree | 20:24 |
| devananda | adam_g: np. it needed to be discussed | 20:24 |
| devananda | yes | 20:24 |
| devananda | we'll have to moev that testing at some point | 20:24 |
| adam_g | devananda, in terms of functionality, are there still gaps between ironic and bm? | 20:24 |
| devananda | but it's not soon | 20:24 |
| devananda | adam_g: yes, a few | 20:24 |
| adam_g | if all of that testing is happening in upstream -infra, it should be easys enough to adjust | 20:24 |
| devananda | adam_g: mostly i think it is serial console | 20:24 |
| devananda | adam_g: yea,all the testing that matters for this discussion is upstream infra. ironic will need our own third-party CI (from vendors) but that won't change | 20:25 |
| adam_g | Shrews, | 20:25 |
| adam_g | MemTotal: 16371064 kB | 20:25 |
| adam_g | MemFree: 184412 kB | 20:25 |
| adam_g | Shrews, test suite is DOS'ing me too | 20:25 |
| devananda | adam_g: nova.virt.baremetal also has some entries in the general doc repos, which we do not | 20:25 |
| Shrews | adam_g: yay! | 20:26 |
| adam_g | devananda, ack | 20:26 |
| devananda | adam_g: that is just from tox? | 20:26 |
| adam_g | devananda, im not sure | 20:26 |
| adam_g | devananda, actually, no tox. running testr directly | 20:26 |
| Shrews | adam_g: what does ps output report as the offender? | 20:27 |
| adam_g | Shrews, that said, it ran fine during a rebuild on the ubuntu build system. | 20:27 |
| * NobodyCam need to do quick afternoon walkies..brb | 20:27 | |
| adam_g | Shrews, didn't poke at it quickly, had to kill it | 20:27 |
| Shrews | adam_g: yeah, it's random | 20:27 |
| adam_g | s/quickly/closely | 20:27 |
| adam_g | devananda, so ill contineu to investigate improvements to https://review.openstack.org/#/c/84033/ and we can hopefully ahve that running somewhere if for noone else but ironic devs | 20:28 |
| russell_h | JayF: which failure are you looking at | 20:28 |
| *** dkehn__ is now known as dkehn_ | 20:29 | |
| *** blamar_ has joined #openstack-ironic | 20:34 | |
| linggao | matty_dubs, in the instruction curl -X PUT -H 'Content-Type: application/json' -H 'Accept: application/json' \ | 20:35 |
| linggao | -H "X-Auth-Token: $TOKEN" -d '{"enabled": "true"}' \ | 20:35 |
| linggao | http://127.0.0.1:6385/v1/nodes/$NODE/states/console | 20:35 |
| *** blamar has quit IRC | 20:36 | |
| *** blamar_ is now known as blamar | 20:36 | |
| linggao | can it be done by ironic node-set-console-mode command? | 20:36 |
| NobodyCam | back | 20:36 |
| matty_dubs | linggao: That command didn't exist when I was doing this. I think that command may be exactly what that API call does. | 20:38 |
| linggao | ok, I just got an error when using this command. Remote error: UnsupportedRpcVersion Specified RPC version, 1.13, not supported by this endpoint. | 20:39 |
| devananda | linggao: that error comes from inside ironic, between API and conductor tier | 20:39 |
| linggao | devananda, is this version mismatch between client and server? | 20:40 |
| devananda | linggao: check ironic/conductor/manager and ironic/conductor/rpcapi. it is there for me | 20:40 |
| devananda | linggao: no. RPC error has nothing to do with client | 20:40 |
| linggao | let me check... | 20:40 |
| *** Mikhail_D_ltp has quit IRC | 20:42 | |
| devananda | adam_g: awesome. i'll give that a shot in a bit | 20:42 |
| devananda | need to move cafe's now | 20:42 |
| devananda | bbiab | 20:42 |
| NobodyCam | :) | 20:42 |
| *** blamar has quit IRC | 20:44 | |
| linggao | devananda, set_console_mode are in both places. | 20:44 |
| linggao | are-> is | 20:44 |
| *** blamar has joined #openstack-ironic | 20:44 | |
| *** blamar has quit IRC | 20:47 | |
| *** blamar has joined #openstack-ironic | 20:50 | |
| linggao | devananda, I found out the error is from the ipmitool call. The error message on the command line is misleading. | 20:54 |
| linggao | I'll open a bug report and fix it. | 20:55 |
| matty_dubs | Weird -- I never hit that. | 20:56 |
| NobodyCam | arrrgggg the delete bug | 20:57 |
| linggao | matty_dubs, in fact the last_error attribute of the node reports the correct error message though. | 20:58 |
| *** jdob has quit IRC | 21:00 | |
| matty_dubs | Oh, interesting. | 21:00 |
| *** blamar has quit IRC | 21:02 | |
| comstud | ok | 21:03 |
| comstud | spam coming. | 21:03 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Sync base object code with Nova's https://review.openstack.org/84859 | 21:03 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Make tests use Node object instead of dict https://review.openstack.org/84860 | 21:03 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Remove a DB query for get_ports_by_node() https://review.openstack.org/84861 | 21:03 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Remove DB query from _sync_power_states https://review.openstack.org/84862 | 21:03 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Add create() and destroy() to Node https://review.openstack.org/84823 | 21:03 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Clean up calls to get_node() https://review.openstack.org/84573 | 21:03 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Add create() and destroy() to Node https://review.openstack.org/84823 | 21:05 |
| openstackgerrit | Chris Behrens proposed a change to openstack/ironic: Clean up calls to get_node() https://review.openstack.org/84573 | 21:05 |
| NobodyCam | ieek | 21:06 |
| NobodyCam | brb | 21:11 |
| NobodyCam | and back :-p | 21:12 |
| devananda | back, sorta | 21:15 |
| NobodyCam | WB devananda (sorta) | 21:16 |
| *** eguz has quit IRC | 21:25 | |
| *** eghobo has joined #openstack-ironic | 21:25 | |
| *** matty_dubs is now known as matty_dubs|gone | 21:29 | |
| NobodyCam | didn't we land the delete node when spawn fails fix? | 21:31 |
| *** romcheg has quit IRC | 21:33 | |
| Shrews | NobodyCam: devananda: ah ha... we aren't setting cpu_arch anymore | 21:35 |
| * Shrews fixes | 21:35 | |
| NobodyCam | Shrews: line 574 | 21:35 |
| NobodyCam | and 568 | 21:36 |
| Shrews | ugh, yeah. but nova isn't seeing it for some reason | 21:36 |
| *** eguz has joined #openstack-ironic | 21:38 | |
| NobodyCam | hummm | 21:41 |
| *** eghobo has quit IRC | 21:41 | |
| Shrews | ugh. annoying | 21:45 |
| adam_g | gah almost there http://logs.openstack.org/59/84859/1/check/check-tempest-dsvm-virtual-ironic/245d2bd/logs/testr_results.html.gz | 21:46 |
| Shrews | NobodyCam: ooooh, think i see the problem | 21:49 |
| rloo | Shrews. I think that line 576 adds 'cpu_arch'. | 21:49 |
| Shrews | devananda: any reason why you no longer set the 'stats' attribute in the dict? | 21:51 |
| Shrews | i think the scheduler is using that to match cpu arch | 21:51 |
| NobodyCam | Shrews: I'd bet your right | 21:51 |
| Shrews | rloo: yep | 21:51 |
| *** blamar has joined #openstack-ironic | 21:51 | |
| rloo | Shrews. good catch. I assumed deva knew that stats wasn't being used :-( | 21:52 |
| rloo | it kind of seems lame, not to know what info is needed... | 21:53 |
| Shrews | yup, that's what it's trying to do | 21:54 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding a reference driver for the agent https://review.openstack.org/84795 | 21:54 |
| russell_h | JoshNang: is all that localization stuff supposed to be in there? | 21:56 |
| NobodyCam | Shrews: w00t.. are you pushing up a patch? for that? | 21:57 |
| JoshNang | heh fixing that up right now | 21:57 |
| Shrews | NobodyCam: yeah. just need to test it first | 21:57 |
| JoshNang | my git skills still aren't perfect | 21:57 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding a reference driver for the agent https://review.openstack.org/84795 | 21:58 |
| Shrews | NobodyCam: I'm also concerned that _node_resources() used to set the extra_specs values, but now that's moved to get_host_stats(), even though _node_resources is called elsewhere | 22:00 |
| Shrews | NobodyCam: i think i need to chat with deva first | 22:00 |
| devananda | Shrews: i may have overlooked that. changing functionality wasn't my intent | 22:03 |
| NobodyCam | Shrews: ack... I had the same concern | 22:03 |
| Shrews | devananda: oh hi! i just left a comment on the review about it. | 22:03 |
| Shrews | get_available_resources() also uses _node_resources(), so that's likely breaking something else somewhere | 22:04 |
| devananda | doh | 22:04 |
| Shrews | devananda: would you like to take another pass at it? or want me to do it? i won't get to it until tomorrow, but don't mind | 22:05 |
| devananda | Shrews: all yours | 22:05 |
| *** max_lobur has quit IRC | 22:06 | |
| Shrews | dang it | 22:06 |
| Shrews | i mean... great! | 22:06 |
| Shrews | :) | 22:06 |
| NobodyCam | lol | 22:06 |
| Shrews | time for dinner. see you folks tomorrow | 22:06 |
| *** overlayer has quit IRC | 22:06 | |
| NobodyCam | have a good night Shrews :) | 22:06 |
| *** max_lobur has joined #openstack-ironic | 22:08 | |
| JoshNang | does anyone know why the SQLAlchemy default is getting changed in my ironic.conf.sample? Nothing in this commit should do that. https://review.openstack.org/#/c/84795/6/etc/ironic/ironic.conf.sample | 22:10 |
| *** max_lobur has quit IRC | 22:10 | |
| *** max_lobur has joined #openstack-ironic | 22:10 | |
| rloo | JoshNang: that's odd. If i recall correctly, the default connection should be sqlite:.... Did you rebase? | 22:12 |
| JoshNang | rloo: yup rebased against master. and yeah it should be #connection=<None> | 22:13 |
| rloo | JoshNang: I mean, there was a change that caused default connection to be <None>, but then it was fixed to sqlite:... | 22:13 |
| rloo | JoshNang: why should it be <None>? | 22:13 |
| NobodyCam | JoshNang: you have sqlight installed? | 22:13 |
| NobodyCam | https://github.com/openstack/ironic/blob/master/ironic/db/sqlalchemy/api.py#L49 | 22:14 |
| rloo | JoshNang. Did you generate that .sample, or did you get it from rebasing? | 22:14 |
| NobodyCam | the patch that added that is about 21 days old | 22:15 |
| JoshNang | pretty sure i don't have sqlite installed | 22:15 |
| JoshNang | i regenerated it when i added new config options | 22:15 |
| rloo | gotta go. but just looked and i see connection=<None>. which is odd cuz I thought it had been changed. maybe my memory is going... | 22:17 |
| JoshNang | no worries. thanks rloo! | 22:18 |
| rloo | or maybe it was that the .sample shows the wrong value for the default. can't remember now. | 22:18 |
| NobodyCam | JoshNang: do you have a ironic.sqlite file n your path? | 22:29 |
| JoshNang | appears so: | 22:33 |
| JoshNang | /data/dev/ironic-teeth-driver/.tox/py27/src/ironic/ironic/ironic.sqlite | 22:33 |
| JoshNang | /data/dev/ironic-teeth-driver/src/ironic/ironic/openstack/common/db/ironic.sqlite | 22:33 |
| JoshNang | /data/dev/openstack-ironic/ironic/openstack/common/db/ironic.sqlite | 22:33 |
| JoshNang | well, 1 of those is relevant | 22:33 |
| NobodyCam | root@ubuntu:/opt/stack/ironic# find / -name ironic.sqlite | 22:33 |
| NobodyCam | root@ubuntu:/opt/stack/ironic# | 22:33 |
| JoshNang | gotcha. weird. | 22:34 |
| *** harlowja has quit IRC | 22:35 | |
| *** harlowja has joined #openstack-ironic | 22:36 | |
| *** max_lobur has quit IRC | 22:50 | |
| NobodyCam | humm looks like we may want to land https://review.openstack.org/#/c/83471/3 based on: http://lists.openstack.org/pipermail/openstack/2014-March/006238.html | 22:54 |
| *** harlowja has quit IRC | 22:54 | |
| *** harlowja has joined #openstack-ironic | 22:55 | |
| NobodyCam | devananda: we may even want to backport this one ^^^^ | 22:55 |
| openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding a reference driver for the agent https://review.openstack.org/84795 | 23:00 |
| devananda | NobodyCam: hmm, checking | 23:04 |
| NobodyCam | I also found this: https://bugzilla.redhat.com/show_bug.cgi?id=1082165 | 23:07 |
| openstackgerrit | Jenkins proposed a change to openstack/ironic: Updated from global requirements https://review.openstack.org/83471 | 23:13 |
| NobodyCam | speak of the ... | 23:14 |
| *** lucas-afk has quit IRC | 23:15 | |
| linggao | I got the following error when running a ironic command. NodeLocked: Node 482efded-0d63-46a8-96f7-bc9a3c5af3c7 is locked by host test-host, please retry after the current operation is completed. | 23:20 |
| linggao | restarting ironic-api and ironic-conducor does not help. | 23:20 |
| linggao | Is the lock in the db? | 23:20 |
| NobodyCam | linggao: and you did retry the command | 23:20 |
| NobodyCam | it possiable that you hit at the same time peredioc task | 23:21 |
| linggao | NobodyCam, yes I did. | 23:21 |
| NobodyCam | anything in the logs? | 23:21 |
| linggao | where is the url for the paste? | 23:23 |
| NobodyCam | I use http://paste.openstack.org | 23:24 |
| linggao | http://paste.openstack.org/show/74915/ | 23:25 |
| linggao | that's the log for ironic-api | 23:25 |
| NobodyCam | linggao: can I get a node show? | 23:26 |
| openstackgerrit | Russell Haering proposed a change to openstack/ironic: Drivers may expose a top-level passthru API https://review.openstack.org/81919 | 23:26 |
| linggao | NobodyCam http://paste.openstack.org/show/74916/ | 23:28 |
| NobodyCam | thats strange | 23:30 |
| openstackgerrit | Michael Davies proposed a change to openstack/ironic: Handling validation in conductor consistently https://review.openstack.org/84374 | 23:30 |
| NobodyCam | linggao: have you looked at the db to see why ironic/ironic/db/sqlalchemy/api.py", line 194 is triggering | 23:32 |
| NobodyCam | ahh | 23:33 |
| NobodyCam | reservation | test-host | 23:33 |
| NobodyCam | it is loched | 23:33 |
| NobodyCam | locked even | 23:33 |
| devananda | NobodyCam: no need to back port that requierments change | 23:34 |
| NobodyCam | ok.. | 23:36 |
| NobodyCam | I'll land it in a few | 23:36 |
| linggao | NobodyCam, you are right. The node is reserved by test_host. I'll clear it manually in the db. It may be caused I killed the ironic deamons while they were doing something. | 23:36 |
| NobodyCam | linggao: :) that would do it yep | 23:37 |
| *** radsy has joined #openstack-ironic | 23:37 | |
| linggao | NobodCam, it is working again. thanks. | 23:40 |
| NobodyCam | :) | 23:40 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
