| *** clsacramento has quit IRC | 00:12 | |
| *** bdeschenes has quit IRC | 00:13 | |
| *** dims_ has quit IRC | 00:20 | |
| *** jrist has quit IRC | 00:21 | |
| *** jrist has joined #openstack-ironic | 00:22 | |
| *** dims has joined #openstack-ironic | 00:31 | |
| *** g3ek has quit IRC | 00:44 | |
| *** g3ek has joined #openstack-ironic | 00:54 | |
| *** syed_ has quit IRC | 00:55 | |
| *** hoangcx has joined #openstack-ironic | 00:55 | |
| *** causten_ has quit IRC | 00:56 | |
| *** phuongnh has joined #openstack-ironic | 01:06 | |
| *** MattMan has quit IRC | 01:09 | |
| *** chlong has quit IRC | 01:09 | |
| *** MattMan has joined #openstack-ironic | 01:09 | |
| *** mtanino has quit IRC | 01:17 | |
| *** rloo has quit IRC | 01:19 | |
| *** bfournie1 has left #openstack-ironic | 01:44 | |
| *** bfournie has joined #openstack-ironic | 01:49 | |
| *** mtanino has joined #openstack-ironic | 02:05 | |
| *** zhangjl has joined #openstack-ironic | 02:14 | |
| *** rama_y has joined #openstack-ironic | 02:32 | |
| openstackgerrit | Tuan Luong-Anh proposed openstack/ironic: Fix setting persistent boot device does not work https://review.openstack.org/376322 | 02:39 |
|---|---|---|
| openstackgerrit | Merged openstack/ironic-specs: Changed the home-page of Ironic-specs in setup.cfg https://review.openstack.org/390091 | 02:44 |
| *** yuanying has quit IRC | 02:47 | |
| *** yuanying has joined #openstack-ironic | 02:48 | |
| *** tuanluong has joined #openstack-ironic | 02:51 | |
| *** yuanying has quit IRC | 02:52 | |
| *** gcb has joined #openstack-ironic | 03:01 | |
| *** amotoki has quit IRC | 03:01 | |
| *** mtanino has quit IRC | 03:22 | |
| *** vikrant has joined #openstack-ironic | 03:40 | |
| *** vinsh has quit IRC | 03:43 | |
| *** links has joined #openstack-ironic | 03:52 | |
| *** mtanino has joined #openstack-ironic | 04:12 | |
| *** hshiina has joined #openstack-ironic | 04:16 | |
| *** rama_y has quit IRC | 04:38 | |
| *** rama_y has joined #openstack-ironic | 04:38 | |
| *** yuanying has joined #openstack-ironic | 04:42 | |
| *** sabeen has quit IRC | 04:45 | |
| *** rama_y has quit IRC | 04:55 | |
| *** rama_y has joined #openstack-ironic | 04:56 | |
| *** yuanying has quit IRC | 04:57 | |
| *** yuanying has joined #openstack-ironic | 04:58 | |
| *** yuanying has quit IRC | 05:08 | |
| *** yuanying has joined #openstack-ironic | 05:12 | |
| *** vinsh has joined #openstack-ironic | 05:14 | |
| *** g3ek has quit IRC | 05:15 | |
| *** g3ek has joined #openstack-ironic | 05:18 | |
| *** evgenyzhurko has joined #openstack-ironic | 05:19 | |
| *** vinsh has quit IRC | 05:19 | |
| *** yuanying has quit IRC | 05:28 | |
| *** yuanying has joined #openstack-ironic | 05:32 | |
| *** yuanying has quit IRC | 05:36 | |
| *** yuanying has joined #openstack-ironic | 05:36 | |
| *** anush has joined #openstack-ironic | 05:38 | |
| openstackgerrit | Tuan Luong-Anh proposed openstack/ironic: Fix setting persistent boot device does not work https://review.openstack.org/376322 | 05:40 |
| *** zackf has joined #openstack-ironic | 05:56 | |
| *** mtanino has quit IRC | 05:56 | |
| openstackgerrit | Tuan Luong-Anh proposed openstack/ironic: Fix setting persistent boot device does not work https://review.openstack.org/376322 | 05:57 |
| *** zackf has quit IRC | 06:01 | |
| *** yuanying has quit IRC | 06:11 | |
| *** mtanino has joined #openstack-ironic | 06:15 | |
| *** yuanying has joined #openstack-ironic | 06:19 | |
| *** mtanino has quit IRC | 06:19 | |
| *** rcernin has joined #openstack-ironic | 06:29 | |
| *** hw_wutianwei has joined #openstack-ironic | 06:39 | |
| *** yuanying has quit IRC | 06:40 | |
| *** yuanying has joined #openstack-ironic | 06:57 | |
| *** fxpester has joined #openstack-ironic | 06:58 | |
| *** rbartal has joined #openstack-ironic | 07:00 | |
| *** yuanying has quit IRC | 07:02 | |
| *** fxpester|2 has joined #openstack-ironic | 07:13 | |
| *** fxpester has quit IRC | 07:14 | |
| *** vinsh has joined #openstack-ironic | 07:15 | |
| *** jtomasek_ has joined #openstack-ironic | 07:16 | |
| *** AlexeyAbashkin has quit IRC | 07:20 | |
| *** vinsh has quit IRC | 07:20 | |
| *** tesseract has joined #openstack-ironic | 07:21 | |
| *** tesseract is now known as Guest14381 | 07:21 | |
| *** yuanying has joined #openstack-ironic | 07:29 | |
| *** AlexeyAbashkin has joined #openstack-ironic | 07:33 | |
| *** jtomasek_ has quit IRC | 07:36 | |
| *** zhangyufei has joined #openstack-ironic | 07:40 | |
| *** yuanying has quit IRC | 07:48 | |
| *** yuanying has joined #openstack-ironic | 07:50 | |
| *** zhangyufei has quit IRC | 07:55 | |
| *** yuanying has quit IRC | 07:57 | |
| *** ohamada has joined #openstack-ironic | 07:59 | |
| *** zzzeek has quit IRC | 08:00 | |
| *** yuanying has joined #openstack-ironic | 08:00 | |
| *** ohamada has quit IRC | 08:00 | |
| *** zzzeek has joined #openstack-ironic | 08:01 | |
| *** zhangyufei has joined #openstack-ironic | 08:02 | |
| *** zhangyufei has quit IRC | 08:05 | |
| *** yuanying_ has joined #openstack-ironic | 08:07 | |
| *** yuanying has quit IRC | 08:08 | |
| *** zhangyufei has joined #openstack-ironic | 08:09 | |
| openstackgerrit | Nguyen Hung Phuong proposed openstack/ironic: TrivialFix: Fix typo in config file https://review.openstack.org/377307 | 08:11 |
| openstackgerrit | Nguyen Hung Phuong proposed openstack/ironic: TrivialFix: Fix typo in config file https://review.openstack.org/377307 | 08:16 |
| *** ccamacho has quit IRC | 08:24 | |
| openstackgerrit | yolanda.robla proposed openstack/bifrost: Enable upper requirements on bifrost https://review.openstack.org/391828 | 08:28 |
| *** ohamada has joined #openstack-ironic | 08:28 | |
| *** jtomasek_ has joined #openstack-ironic | 08:34 | |
| *** lucas-afk is now known as lucasagomes | 08:40 | |
| lucasagomes | morning all | 08:40 |
| openstackgerrit | yolanda.robla proposed openstack/bifrost: Enable upper requirements on bifrost https://review.openstack.org/391828 | 08:45 |
| *** moshele has joined #openstack-ironic | 08:47 | |
| tuanluong | morning ironic | 08:48 |
| tuanluong | Hello lucasagomes, please take a look https://review.openstack.org/#/c/391415/ | 08:49 |
| patchbot | patch 391415 - nova - Fix hypervisor-stats show invalid information abou... | 08:49 |
| pas-ha | morning Ironic | 08:50 |
| tuanluong | morning pas-ha | 08:50 |
| openstackgerrit | yolanda.robla proposed openstack/bifrost: Enable upper requirements on bifrost https://review.openstack.org/391828 | 08:54 |
| *** athomas has joined #openstack-ironic | 08:59 | |
| *** ohamada has quit IRC | 08:59 | |
| openstackgerrit | yolanda.robla proposed openstack/bifrost: Enable upper requirements on bifrost https://review.openstack.org/391828 | 09:00 |
| *** mgoddard has joined #openstack-ironic | 09:14 | |
| *** zhangyufei has quit IRC | 09:14 | |
| *** zhangyufei has joined #openstack-ironic | 09:15 | |
| *** jaosorior has joined #openstack-ironic | 09:15 | |
| *** vinsh has joined #openstack-ironic | 09:17 | |
| *** jpich has joined #openstack-ironic | 09:18 | |
| alineb | morning all | 09:20 |
| *** vinsh has quit IRC | 09:22 | |
| *** dtantsur|afk is now known as dtantsur | 09:26 | |
| dtantsur | Morning Ironic | 09:26 |
| dtantsur | hey lucasagomes, how was your trip back? | 09:26 |
| *** milan has joined #openstack-ironic | 09:28 | |
| *** galyna has joined #openstack-ironic | 09:28 | |
| *** portdirect has quit IRC | 09:31 | |
| lucasagomes | dtantsur, it was pretty good, short flight this time :-) | 09:31 |
| lucasagomes | dtantsur, yours ? | 09:31 |
| *** hshiina has quit IRC | 09:31 | |
| *** portdirect_ has joined #openstack-ironic | 09:31 | |
| dtantsur | also fine, yeah | 09:31 |
| *** portdirect_ is now known as portdirect | 09:32 | |
| *** derekh has joined #openstack-ironic | 09:35 | |
| *** e0ne has joined #openstack-ironic | 09:38 | |
| tuanluong | Morning ironic | 09:40 |
| *** g3ek has quit IRC | 09:41 | |
| milan | morning Ironic! :) | 09:43 |
| *** g3ek has joined #openstack-ironic | 09:43 | |
| milan | morning aarefiev dtantsur lucasagomes tuanluong ! :) | 09:43 |
| lucasagomes | o/ | 09:43 |
| *** yuanying_ has quit IRC | 09:44 | |
| *** yuanying has joined #openstack-ironic | 09:44 | |
| tuanluong | morming milan | 09:44 |
| *** portdirect has quit IRC | 09:45 | |
| dtantsur | hey milan, you're early this time ;) | 09:47 |
| milan | as always :D | 09:47 |
| milan | how's going? | 09:47 |
| * milan merges conflicts | 09:48 | |
| *** jaosorior has quit IRC | 09:48 | |
| *** jaosorior has joined #openstack-ironic | 09:48 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Docs: Document using operators with root device hints https://review.openstack.org/386714 | 09:57 |
| *** portdirect has joined #openstack-ironic | 09:58 | |
| dtantsur | milan, still somewhat sick from the conference, otherwise ok :) | 09:59 |
| *** dootniz is now known as kragniz | 09:59 | |
| *** tuanluong has quit IRC | 10:00 | |
| milan | I guess plenty laid down with the manflu, sucks... have had mine a week before the summit | 10:02 |
| milan | dtantsur, get better soon! :) | 10:03 |
| lucasagomes | dtantsur, that sucks yeah :-/ | 10:04 |
| *** hoangcx has quit IRC | 10:05 | |
| *** mgould|afk is now known as mgould | 10:05 | |
| mgould | morning lucasagomes pas-ha alineb milan aarefiev dtantsur | 10:05 |
| mgould | dtantsur: get well soon! | 10:05 |
| dtantsur | hey mgould! | 10:05 |
| milan | morning mgould! :) | 10:06 |
| pas-ha | morning mgould :) | 10:06 |
| dtantsur | lucasagomes, fyi this does not seem to work: https://github.com/openstack/python-ironicclient/commit/eeb327b590b489a7f6b0e2176b288f7201de6aa9 :( | 10:06 |
| mgould | milan: clearly you're the one who infected everyone else :-) | 10:06 |
| milan | mgould, Oops O:-) | 10:07 |
| * mgould was also feeling subpar the week before summit, so maybe it was me :-/ | 10:10 | |
| *** sambetts|afk is now known as sambetts | 10:12 | |
| sambetts | Morning al | 10:13 |
| sambetts | all | 10:13 |
| dtantsur | morning sambetts | 10:13 |
| mgould | morning sambetts | 10:17 |
| *** zhangyufei has quit IRC | 10:20 | |
| milan | morning sambetts! :) | 10:21 |
| lucasagomes | dtantsur, checking | 10:25 |
| dtantsur | lucasagomes, see https://bugzilla.redhat.com/show_bug.cgi?id=1346089 for details. in ironicclient we don't use the code you've patched... | 10:26 |
| openstack | bugzilla.redhat.com bug 1346089 in python-keystoneauth1 "Secrets are not filtered when logging requests in the session code" [Low,Assigned] - Assigned to derekh | 10:26 |
| lucasagomes | dtantsur, oh keystone logs it | 10:28 |
| dtantsur | yep | 10:28 |
| lucasagomes | *sigh* | 10:28 |
| mgould | filtering out secrets at the logging stage sounds... fragile... but I guess it's too late to change that architecture now | 10:30 |
| *** phuongnh has quit IRC | 10:31 | |
| dtantsur | having two independent places we do HTTP is also not so robust.. | 10:31 |
| dtantsur | I guess we have to get rid of our implementation eventually... | 10:31 |
| *** zhangjl has quit IRC | 10:34 | |
| *** hshiina has joined #openstack-ironic | 10:35 | |
| dtantsur | lucasagomes, re https://review.openstack.org/#/c/387559: I'm worried to allow infiniband in e.g. network separation code.. does it work there? | 10:39 |
| patchbot | patch 387559 - ironic - Account for Infiniband GID when validating an address | 10:39 |
| dtantsur | (maybe it does, dunno) | 10:40 |
| lucasagomes | dtantsur, I assume yes, but I haven't tested it | 10:40 |
| *** mmedvede has quit IRC | 10:41 | |
| dtantsur | lucasagomes, what if we first only fix the lookup? e.g. I'm not sure it's a great idea to blow up on malformed MACs at all | 10:41 |
| mgould | dtantsur: what assumptions do we make about the mechanism being used to achieve separation? | 10:41 |
| dtantsur | mgould, it's called Neutron and nobody knows how it works :) | 10:41 |
| mgould | looks like Infiniband supports something very similar to VLANs: https://community.mellanox.com/docs/DOC-1463 | 10:41 |
| mgould | aaaaah | 10:42 |
| lucasagomes | dtantsur, like, ignore the malformed ones ? | 10:42 |
| dtantsur | lucasagomes, I suggest just logging and ignoring invalid MACs, yes | 10:42 |
| lucasagomes | that's not a bad idea | 10:42 |
| dtantsur | lucasagomes, we don't filter input to lookup on IPA side too much.. | 10:42 |
| lucasagomes | dtantsur, lemme add a patch for it | 10:42 |
| lucasagomes | dtantsur, yeah, we shouldn't actually fail like that on lookup cause it totally breaks the deployment in the middle | 10:42 |
| dtantsur | lucasagomes, cool! and then we can discuss the bigger infiniband patch, with involving people understanding neutron better | 10:42 |
| lucasagomes | and don't shut down the node or anything | 10:43 |
| lucasagomes | dtantsur, right | 10:43 |
| mgould | hmmm, looks like Neutron only supports Infiniband via a third-party plugin: https://wiki.openstack.org/wiki/Mellanox-Neutron-Liberty-InfiniBand | 10:45 |
| sambetts | mgould: that means that the neutron data structures must support it | 10:48 |
| sambetts | mgould: just it'll only ever get bound if you have enabled that driver and have one of their devices | 10:48 |
| mgould | ah, OK | 10:48 |
| dtantsur | ok, so lucasagomes' patch still makes sense, but it's at the border of a feature, so we need a lighter fix first | 10:50 |
| sambetts | does our DB structure support infiniband right now? | 10:52 |
| sambetts | I'm sure there was a whole bunch of other infinband patches for inspector and ironic | 10:52 |
| dtantsur | dunno. the first thing I care about is the deployment to not fail when an infiniband card is merely present :) | 10:53 |
| sambetts | Yeah we already merged this https://review.openstack.org/#/c/264263/ | 10:53 |
| patchbot | patch 264263 - ironic - Adding InfiniBand Support (MERGED) | 10:53 |
| lucasagomes | dtantsur, yeah working on it | 10:54 |
| lucasagomes | makes me sad that the whole validation goes in a decorator, trying to think of another way | 10:55 |
| dtantsur | lucasagomes, replace the type with list of strings and do validation later? | 10:55 |
| *** yuanying has quit IRC | 10:56 | |
| lucasagomes | dtantsur, that's the plan | 10:56 |
| *** dprince has joined #openstack-ironic | 10:56 | |
| *** chlong has joined #openstack-ironic | 10:57 | |
| *** jkilpatr has quit IRC | 10:57 | |
| sambetts | lucasagomes, dtantsur: Going back to the infiniband spec, we decided we weren't going to store the GID in ironic, https://specs.openstack.org/openstack/ironic-specs/specs/not-implemented/add-infiniband-support.html#proposed-change | 11:00 |
| dtantsur | cool, yeah | 11:01 |
| dtantsur | I think the plan is to eventually provide translation to MACs, but it's vendor specific and is not implemented yet | 11:01 |
| sambetts | ++ | 11:01 |
| lucasagomes | sambetts, right on, will take a look | 11:03 |
| *** evgenyzhurko has quit IRC | 11:11 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-specs: Revise driver composition reform defaults https://review.openstack.org/357262 | 11:13 |
| mat128 | morning Ironic | 11:16 |
| mgould | morning mat128 | 11:16 |
| mgould | dtantsur: does it make sense to translate GIDs to MACs? GIDs are longer, so we can't do it losslessly unless I'm missing something | 11:17 |
| dtantsur | mgould, well, some vendors have procedures for that.. dunno how much sense it makes | 11:18 |
| sambetts | mgould: its part of the infiniband spec for doing PXE over infiniband as I understand | 11:18 |
| sambetts | ethernet over inifiband | 11:18 |
| mgould | ah, OK | 11:18 |
| mgould | tunnel all the things | 11:18 |
| sambetts | ;) | 11:18 |
| mgould | brb meeting | 11:19 |
| *** MattMan has quit IRC | 11:19 | |
| *** vinsh has joined #openstack-ironic | 11:19 | |
| *** MattMan has joined #openstack-ironic | 11:21 | |
| moshele | lucasagomes, sambetts, dtantsur: yes so infinibnad support I have the ironic-inspctor patch https://review.openstack.org/#/c/264257/ I need to address the comments and the IPA patch https://review.openstack.org/#/c/334816/ which I need to remove the vendor code to different Hardware class | 11:22 |
| patchbot | patch 264257 - ironic-inspector - Adding InfiniBand Support | 11:22 |
| patchbot | patch 334816 - ironic-python-agent - InfiniBand Support for IPA | 11:22 |
| *** vinsh has quit IRC | 11:23 | |
| jroll | morning y'all | 11:26 |
| sambetts | o/ jroll | 11:26 |
| dtantsur | morning jroll | 11:26 |
| dtantsur | the snmp driver is still deprecated, right? so we should reject both https://bugs.launchpad.net/ironic/+bug/1635644 and https://review.openstack.org/#/c/389736/ | 11:30 |
| openstack | Launchpad bug 1635644 in Ironic "[RFE] SNMP driver auto-discovery" [Undecided,In progress] - Assigned to Philippe Godin (godp1301) | 11:30 |
| patchbot | patch 389736 - ironic - Adds snmp driver auto discovery for APC and baytech | 11:30 |
| dtantsur | jroll, ^^^ | 11:30 |
| *** ohamada has joined #openstack-ironic | 11:32 | |
| *** mmedvede has joined #openstack-ironic | 11:34 | |
| *** jkilpatr has joined #openstack-ironic | 11:34 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: API: lookup() ignore malformed MAC addresses https://review.openstack.org/392114 | 11:43 |
| lucasagomes | dtantsur, ^ | 11:43 |
| lucasagomes | jroll, hi there | 11:43 |
| dtantsur | lucasagomes, thanks, will look after lunch | 11:43 |
| lucasagomes | yeah no worries | 11:44 |
| *** evgenyzhurko has joined #openstack-ironic | 11:46 | |
| *** amotoki has joined #openstack-ironic | 11:47 | |
| *** rbrud has joined #openstack-ironic | 11:47 | |
| jroll | dtantsur: let's check that bug with mat128, I think they might be depending on that to do the virtualpdu thing | 11:56 |
| * mat128 reads | 11:56 | |
| jroll | ohai :D | 11:56 |
| mat128 | dtantsur, jroll: thats correct, the guys made improvements to the snmp drivers but they are on hold pending the VirtualPDU work landing | 11:57 |
| mat128 | which is also in progress | 11:57 |
| mat128 | s/drivers/driver | 11:57 |
| mat128 | hey :) | 11:57 |
| mat128 | iirc xhku proposed the initial changes for virtualPDU | 11:57 |
| mat128 | yup: https://review.openstack.org/#/c/388154/ | 11:58 |
| patchbot | patch 388154 - ironic - WIP- Add virtualpdu to ironic devstack plugin | 11:58 |
| jroll | ok cool, so that's required to do CI for that driver | 11:58 |
| jroll | chicken/egg, I think let's do the work | 11:58 |
| mat128 | jroll: thats correct | 11:58 |
| mat128 | well | 11:58 |
| mat128 | the auto-discovery and the baytech support are real RFEs | 11:58 |
| mat128 | not required for virtualPDU to work | 11:59 |
| jroll | ahhh | 11:59 |
| mat128 | but adding features requires the driver to be non-deprecated | 11:59 |
| jroll | right | 11:59 |
| mat128 | which is additional pushing for virtualPDU to be prioritized | 11:59 |
| mat128 | if you know what I mean | 11:59 |
| jroll | mat128: yeah, so given we plan to provide CI for snmp driver, we should just put the rfe on hold until that point in time | 12:00 |
| mat128 | jroll: correct | 12:00 |
| mat128 | I agree with this :) | 12:00 |
| jroll | cool :) | 12:00 |
| *** e0ne has quit IRC | 12:00 | |
| * jroll comments in the bug | 12:01 | |
| openstackgerrit | Sam Betts proposed openstack/ironic: Allow to use FLAT network setup https://review.openstack.org/334382 | 12:10 |
| *** portdirect has quit IRC | 12:15 | |
| *** lucasagomes is now known as lucas-hungry | 12:16 | |
| *** portdirect has joined #openstack-ironic | 12:18 | |
| *** vinsh has joined #openstack-ironic | 12:20 | |
| *** jkilpatr_ has joined #openstack-ironic | 12:23 | |
| *** jkilpatr has quit IRC | 12:24 | |
| *** vinsh has quit IRC | 12:25 | |
| *** rbartal has left #openstack-ironic | 12:29 | |
| dtantsur | mat128, glad to hear it! I guess it has to be done by end of ocata to avoid the driver from getting kicked out.. | 12:30 |
| dtantsur | speaking of which, jroll, do we have a decided plan on when and how we can start removing drivers? | 12:30 |
| *** rbartal has joined #openstack-ironic | 12:30 | |
| jroll | dtantsur: 3 months from when the deprecation landed, I guess? | 12:31 |
| jroll | though, that will be close to ocata FF, maybe we just remove in pike | 12:31 |
| dtantsur | jroll, well, we can do it before FF.. I'm not sure I feel like maintaining unsupported drivers for one more cycle | 12:32 |
| dtantsur | we can make exceptions for drivers like SNMP with a CI being worked on, of course | 12:32 |
| jroll | sure, I'm fine with that | 12:32 |
| dtantsur | also, we should finally switch to pxe_ipmitool now... lucas-hungry has a patch for it iirc, not sure about the state | 12:34 |
| jroll | yeah, +1 | 12:34 |
| *** fxpester|2 has quit IRC | 12:36 | |
| *** fxpester has joined #openstack-ironic | 12:36 | |
| *** jcoufal has joined #openstack-ironic | 12:36 | |
| jroll | sorry, still not feeling well today, brain is foggy :/ | 12:37 |
| *** jerrygb_ has quit IRC | 12:38 | |
| *** vikrant has quit IRC | 12:40 | |
| *** moshele has quit IRC | 12:46 | |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: TEST: do not review https://review.openstack.org/392138 | 12:49 |
| *** moshele has joined #openstack-ironic | 12:50 | |
| *** jerrygb has joined #openstack-ironic | 12:52 | |
| *** hw_wutianwei has quit IRC | 12:52 | |
| *** jheroux has joined #openstack-ironic | 12:53 | |
| * jroll steps away for a while | 12:54 | |
| *** amotoki has quit IRC | 12:54 | |
| *** derekh has quit IRC | 12:56 | |
| *** amotoki has joined #openstack-ironic | 12:57 | |
| *** vishwanathj has quit IRC | 13:03 | |
| *** links has quit IRC | 13:14 | |
| *** lucas-hungry is now known as lucasagomes | 13:18 | |
| *** vinsh has joined #openstack-ironic | 13:21 | |
| *** fxpester has quit IRC | 13:21 | |
| *** e0ne has joined #openstack-ironic | 13:23 | |
| *** vinsh has quit IRC | 13:26 | |
| *** baoli has joined #openstack-ironic | 13:30 | |
| *** wajdi has joined #openstack-ironic | 13:31 | |
| *** fxpester has joined #openstack-ironic | 13:31 | |
| lucasagomes | yuriyz, dtantsur re: changing the list_of_mac_addresses type. I don't think it will be consistent with other types to change the validation there. All other types will raise an exception in case a value fails to validate | 13:32 |
| *** fragatina has joined #openstack-ironic | 13:32 | |
| *** yuanying has joined #openstack-ironic | 13:32 | |
| *** fragatina has quit IRC | 13:32 | |
| *** fragatina has joined #openstack-ironic | 13:33 | |
| lucasagomes | I would rather delete that list_of_macs... type if it's not used anymore than making it inconsistent with the rest, because I think it will be misleading if we decide to use it in another places | 13:33 |
| *** amotoki has quit IRC | 13:35 | |
| *** absubram has joined #openstack-ironic | 13:36 | |
| dtantsur | lucasagomes, I'm fine with deleting it too | 13:36 |
| lucasagomes | dtantsur, right, delete in the same patch or another one to facilitate the backport ? | 13:36 |
| * lucasagomes is good with either | 13:36 | |
| dtantsur | lucasagomes, in the same patch IMO, as it's the only place where it's used | 13:37 |
| lucasagomes | ack | 13:38 |
| rbartal | HI all, I have HP 380/360 Gen9 computer and while downloading inspector.ipxe it get connection reset. The problem is it ignore the kernel --timeout option we added in the ipxe scrip | 13:39 |
| dtantsur | rbartal, wait, does it reset on downloading the kernel or inspector.ipxe? in the latter case there is no --timeout | 13:39 |
| rbartal | dtantsur, it reset on inspector.ipxe | 13:42 |
| dtantsur | ok, so timeout is irrelevant here.. | 13:42 |
| dtantsur | rbartal, two questions then: 1. what's the size of your inspector.ipxe, 2. what does tcpdump show for port 8088 (or whichever we use)? | 13:43 |
| *** hshiina has quit IRC | 13:44 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: API: lookup() ignore malformed MAC addresses https://review.openstack.org/392114 | 13:44 |
| rbartal | dtantsur, OK, so --timeout will work for agent and kernel download. I will ask for the file side and some tcpdump files | 13:45 |
| mariojv | morning ironic o/ | 13:45 |
| dtantsur | morning mariojv | 13:46 |
| *** rloo has joined #openstack-ironic | 13:46 | |
| *** fragatina has quit IRC | 13:48 | |
| *** causten has joined #openstack-ironic | 13:50 | |
| *** g3ek has quit IRC | 13:51 | |
| *** amotoki has joined #openstack-ironic | 13:51 | |
| *** fxpester|2 has joined #openstack-ironic | 13:53 | |
| *** cdearborn has joined #openstack-ironic | 13:55 | |
| *** fxpester has quit IRC | 13:56 | |
| *** g3ek has joined #openstack-ironic | 13:56 | |
| *** derekh has joined #openstack-ironic | 13:57 | |
| *** jerrygb_ has joined #openstack-ironic | 14:00 | |
| *** yuanying has quit IRC | 14:01 | |
| *** amotoki has quit IRC | 14:02 | |
| *** jerrygb has quit IRC | 14:02 | |
| *** hamzy has quit IRC | 14:09 | |
| *** mtanino has joined #openstack-ironic | 14:13 | |
| *** vinsh has joined #openstack-ironic | 14:19 | |
| *** MattMan has quit IRC | 14:20 | |
| *** rpioso has joined #openstack-ironic | 14:26 | |
| *** yufei has joined #openstack-ironic | 14:29 | |
| *** awiddersheim has quit IRC | 14:30 | |
| *** e0ne has quit IRC | 14:30 | |
| *** jkilpatr_ has quit IRC | 14:31 | |
| *** jerrygb has joined #openstack-ironic | 14:31 | |
| NobodyCam | Good Morning Ironic'ers :) | 14:31 |
| *** awiddersheim has joined #openstack-ironic | 14:33 | |
| dtantsur | morning NobodyCam | 14:33 |
| *** jerrygb_ has quit IRC | 14:33 | |
| NobodyCam | hey hey dtantsur :) did you have a good Halloween? | 14:34 |
| dtantsur | well, this time we didn't have energy or mood for Halloween.. | 14:34 |
| dtantsur | and you? | 14:34 |
| NobodyCam | We are in the RV because of some house remodeling going on and as a result no kids at all :p ended up having pizza with some friends :) | 14:35 |
| dtantsur | :) | 14:36 |
| NobodyCam | :) | 14:37 |
| lucasagomes | NobodyCam, morning | 14:38 |
| NobodyCam | Good Morning lucasagomes :) | 14:39 |
| NobodyCam | How are you today :) | 14:39 |
| *** jkilpatr has joined #openstack-ironic | 14:40 | |
| *** e0ne has joined #openstack-ironic | 14:42 | |
| lucasagomes | good good, loads of emails and non-fun things to sort out today tho | 14:43 |
| NobodyCam | :( | 14:44 |
| *** MattMan has joined #openstack-ironic | 14:46 | |
| lucasagomes | NobodyCam, it's all good :-) | 14:47 |
| * dtantsur still has to do expense reports when he gets to the office... | 14:48 | |
| *** vishwanathj has joined #openstack-ironic | 14:51 | |
| *** ipukha has quit IRC | 14:53 | |
| NobodyCam | hehehe :) | 14:54 |
| *** yufei__ has joined #openstack-ironic | 14:57 | |
| dtantsur | jroll, oslo libs seem to add reno support (looking at futurist patch now). should we do the same with ironic-lib now? | 14:58 |
| *** agordeev has quit IRC | 14:59 | |
| *** amotoki has joined #openstack-ironic | 14:59 | |
| *** yufei has quit IRC | 14:59 | |
| *** yufei__ is now known as yufei | 14:59 | |
| *** jerrygb_ has joined #openstack-ironic | 15:00 | |
| lucasagomes | dtantsur, did the "it's not a user facing library" motto changed ? | 15:02 |
| dtantsur | apparently? | 15:02 |
| *** jerrygb has quit IRC | 15:03 | |
| lucasagomes | so it did change to ... ? | 15:03 |
| * lucasagomes looks | 15:03 | |
| dtantsur | dunno, I was not in their sessions too much | 15:05 |
| *** sabeen has joined #openstack-ironic | 15:09 | |
| *** syed_ has joined #openstack-ironic | 15:14 | |
| *** vdefeo has joined #openstack-ironic | 15:16 | |
| mgould | morning NobodyCam | 15:18 |
| *** fxpester|2 has quit IRC | 15:18 | |
| NobodyCam | Morning mgould. :) | 15:18 |
| *** wajdi has quit IRC | 15:22 | |
| *** athomas has quit IRC | 15:22 | |
| *** hj-hpe has joined #openstack-ironic | 15:24 | |
| *** hj-hpe has quit IRC | 15:24 | |
| *** hj-hpe has joined #openstack-ironic | 15:25 | |
| *** milan has quit IRC | 15:28 | |
| *** wajdi has joined #openstack-ironic | 15:28 | |
| *** athomas has joined #openstack-ironic | 15:29 | |
| *** absubram has quit IRC | 15:31 | |
| *** vdefeo has quit IRC | 15:33 | |
| *** hpe-hj has joined #openstack-ironic | 15:35 | |
| *** absubram has joined #openstack-ironic | 15:36 | |
| *** hj-hpe has quit IRC | 15:37 | |
| *** pcaruana has joined #openstack-ironic | 15:44 | |
| *** Guest14381 has quit IRC | 15:51 | |
| *** rbrud has quit IRC | 15:59 | |
| openstackgerrit | Galyna Zholtkevych proposed openstack/ironic: ETAG supporting to enhance API evolution https://review.openstack.org/392213 | 16:00 |
| *** rama_y has quit IRC | 16:03 | |
| *** rama_y has joined #openstack-ironic | 16:03 | |
| *** hamzy has joined #openstack-ironic | 16:04 | |
| *** rcernin has quit IRC | 16:06 | |
| *** milan has joined #openstack-ironic | 16:07 | |
| *** pcaruana has quit IRC | 16:07 | |
| *** mat128_ has quit IRC | 16:08 | |
| *** tdurakov has quit IRC | 16:09 | |
| *** mat128_ has joined #openstack-ironic | 16:10 | |
| *** tdurakov has joined #openstack-ironic | 16:11 | |
| mgould | huh, that was quick | 16:11 |
| mgould | nice work, gzholtkevych | 16:12 |
| mgould | ah, still WIP :-) | 16:12 |
| galyna | Yeap, this is just a first patch :) nothing special yet | 16:16 |
| *** ipukha has joined #openstack-ironic | 16:16 | |
| *** jaosorior has quit IRC | 16:17 | |
| *** yufei has quit IRC | 16:21 | |
| *** rpioso has quit IRC | 16:23 | |
| *** zackf has joined #openstack-ironic | 16:25 | |
| *** sabeen has quit IRC | 16:26 | |
| jroll | dtantsur: dunno, I don't think it's useful until we want non-ironic projects to use it | 16:26 |
| dtantsur | makes sense | 16:27 |
| rloo | dtantsur: I +2'd https://review.openstack.org/#/c/357262. Can you +2 too, even though you revised it? | 16:31 |
| patchbot | patch 357262 - ironic-specs - Revise driver composition reform defaults | 16:31 |
| JayF | I'm reviewing that right now as well | 16:32 |
| dtantsur | rloo, me? well, I can +1 it, I've rewritten it too much :) | 16:32 |
| rloo | JayF: ok, will wait for your approval :) | 16:32 |
| rloo | dtantsur: oh, didn't realize you did two revisions on it. | 16:33 |
| dtantsur | yep | 16:33 |
| JayF | rloo: land it with my +2? | 16:33 |
| rloo | JayF: yes please! | 16:33 |
| rloo | JayF: or you can +2, fix that nit i found, then +A | 16:34 |
| JayF | patch 357262 cleared for landing, runway 33R | 16:34 |
| patchbot | https://review.openstack.org/#/c/357262/ - ironic-specs - Revise driver composition reform defaults | 16:34 |
| dtantsur | \o/ | 16:34 |
| JayF | rloo: I mean, it's one character? Meh about us spending time to fix it :D | 16:34 |
| rloo | JayF: exactly. i wouldn't want a followup to that, but edit'ing is easy. | 16:34 |
| JayF | rloo: I'm sure someone will come around and trivialfix it :P | 16:34 |
| rloo | JayF: NOOOOOO | 16:35 |
| *** zackf has quit IRC | 16:35 | |
| dtantsur | LOL | 16:35 |
| *** jrcloud has joined #openstack-ironic | 16:35 | |
| openstackgerrit | Merged openstack/ironic-specs: Revise driver composition reform defaults https://review.openstack.org/357262 | 16:36 |
| mgould | yay! | 16:36 |
| *** jrcloud is now known as zackf | 16:37 | |
| *** rpioso has joined #openstack-ironic | 16:38 | |
| rloo | dtantsur, JayF: question for you. why would the home page be ironic-specs? I'd think it would be developer/ironic. what does 'home page' mean? https://review.openstack.org/#/c/390091/ | 16:39 |
| patchbot | patch 390091 - ironic-specs - Changed the home-page of Ironic-specs in setup.cfg (MERGED) | 16:39 |
| JayF | rloo: it's what would be listed in pypi if we published a package there. It's pointed to our devdocs for Ironic. | 16:39 |
| *** aNuposic has joined #openstack-ironic | 16:39 | |
| JayF | rloo: IMO it doesn't matter since we don't publish that to pypi, and the link to the generated specs is 10x more accurate than just a pointer to openstack.org | 16:39 |
| rloo | JayF: so if i were looking for info about a py package, i'd want the home page to be where i could find out info about that package, eg how to file bugs, who to contact, ...?? | 16:40 |
| rloo | JayF: ironic-specs just has our specs... | 16:40 |
| JayF | uh | 16:40 |
| JayF | our pypi page for ironic looks screwed up | 16:41 |
| mgould | rloo: how to use the package, most importantly | 16:41 |
| JayF | https://pypi.python.org/pypi/ironic/0.0 | 16:41 |
| * dtantsur does not care much about home page link for specs | 16:41 | |
| *** jerrygb has joined #openstack-ironic | 16:41 | |
| dtantsur | JayF, yeah, it's been like that for loooong time | 16:41 |
| sambetts | JayF: openstack services don't publish to pypi | 16:41 |
| JayF | Has anyone mentioned it to jroll or the releases team? | 16:41 |
| dtantsur | ... except for ironic-inspector and a few other ;) | 16:41 |
| JayF | sambetts: seriously/ Why not/ | 16:41 |
| sambetts | JayF: no idea | 16:42 |
| * rloo sorry she asked | 16:42 | |
| rloo | JayF: maybe open a bug then about the ironic one | 16:42 |
| JayF | we even publish IPA to pypi | 16:42 |
| sambetts | i thought it was weird too, but none of neutron/nova/glance etc publish on there for some reason | 16:42 |
| JayF | rloo: https://pypi.python.org/pypi/ironic-inspector/4.2.0 the "Home Page:" there under the files is what that would be used for | 16:42 |
| *** jerrygb_ has quit IRC | 16:43 | |
| rloo | JayF: that looks nice, good job dtantsur :) | 16:43 |
| *** jerrygb_ has joined #openstack-ironic | 16:44 | |
| *** moshele has quit IRC | 16:45 | |
| *** jerrygb has quit IRC | 16:46 | |
| JayF | It's completely acceptable to land the small typo fixes with only one +2, correct? | 16:46 |
| dtantsur | IMO yes | 16:47 |
| rloo | JayF: yes. we have that documented somewhere | 16:47 |
| rloo | JayF: https://wiki.openstack.org/wiki/Ironic/CoreTeam#Other_notes | 16:48 |
| *** sabeen has joined #openstack-ironic | 16:48 | |
| JayF | That's what I thought, but haven't done it yet since being core | 16:48 |
| JayF | I also commented that we prefer more substantive changes to typo-only fixes | 16:49 |
| rloo | JayF: i just ignore those patches unless i recognize that person as having contributed something more 'useful' | 16:51 |
| jkilpatr | dtantsur, thoughts on exposing introspection delay as a command line flag and adding a retires flag? That would satisfy the reliability RFE I think. I'm willing to implement. | 16:52 |
| JayF | If it conflicts with nothing, it's hard for me to not land it because it is /better/... if I have it open in my review anyway | 16:52 |
| dtantsur | jkilpatr, which command are you referring to right now, normal or bulk? :) | 16:52 |
| jkilpatr | dtantsur, normal, talked to the tripleo guys and they don't seem interested in doing the business logic on their end. | 16:54 |
| *** e0ne has quit IRC | 16:54 | |
| dtantsur | jkilpatr, "delay" is present in ironic-inspector and makes no sense for our single-node introspection command | 16:54 |
| jkilpatr | one the other hand, delay makes no sense as a single command option. | 16:54 |
| dtantsur | jkilpatr, maybe the biggest win is to make the delay bigger for tripleo and cover all drivers? | 16:54 |
| dtantsur | jkilpatr, i.e. expose this options in puppet (if they're not already) https://github.com/openstack/ironic-inspector/blob/master/example.conf#L58-L63 and change them in instack-undercloud? | 16:55 |
| jkilpatr | dtantsur, that would do it, even a few seconds would be fine. You just need to prevent 50+ nodes all deciding to pxe at exactly the same second. | 16:55 |
| dtantsur | actually, I think https://github.com/openstack/ironic-inspector/blob/master/example.conf#L63 should be changed to .* as we're moving away from pxe_ssh drivers | 16:55 |
| dtantsur | jkilpatr, maybe you can try doing ^^^ locally, and propose it upstream if it fixes your problem? | 16:56 |
| openstackgerrit | Prudhvi Rao Shedimbi proposed openstack/ironic: Making IronicException to subclass from ironic_lib https://review.openstack.org/387703 | 16:57 |
| jkilpatr | dtantsur, that's the plan. Just wanted to know I was hacking at the right level, there are a lot of different levels to potentially drop this in. | 16:57 |
| *** e0ne has joined #openstack-ironic | 16:59 | |
| *** mgoddard_ has joined #openstack-ironic | 16:59 | |
| dtantsur | jkilpatr, if changing the default value for this option works for you, I'm inclined to just do it | 16:59 |
| jkilpatr | dtantsur, while a default delay will help (probably a good bit) it's not going to solve everything, previous iterations of my own introspection script demonstrated that, which is why I still want to add a retry option. | 17:00 |
| dtantsur | jkilpatr, I'm not sure we're going to implement retries in ironic-inspector itself... needs broader discussion, probably mailing list | 17:01 |
| *** jpich has quit IRC | 17:01 | |
| *** mgoddard has quit IRC | 17:02 | |
| jkilpatr | dtantsur, I figured as much. I guess I'll start collecting some hard data to present my case. What scale do you guys normally test at? It works perfectly at ~10 nodes and ok up to about ~20 once you get much beyond that you have to start writing scripts to organize it and do retires. | 17:03 |
| *** e0ne has quit IRC | 17:03 | |
| dtantsur | jkilpatr, I guess answer "1-2 nodes" will disappoint you then :) | 17:04 |
| jkilpatr | dtantsur, I have 24 for personal testing, serious tests get 200+, right now getting a couple hundred node deployment takes a lot of fighting with ironic/tripleo which isn't any good to actually get customers to use, they frequently have or want deployments that big. | 17:07 |
| dtantsur | I wish our team had something like that... right now we (at least the RH ironic team) can only rely on your feedback | 17:08 |
| JayF | jkilpatr: I'm curious what actually is the bottleneck | 17:08 |
| JayF | jkilpatr: just getting them pxe booted? or actually getting the inspection completed? | 17:08 |
| *** e0ne has joined #openstack-ironic | 17:09 | |
| jkilpatr | dtantsur, the lab we are using takes schedule requests from anyone, if your team wants to set some time aside and schedule a slot they could have 200 nodes for testing for a week or so. | 17:09 |
| jkilpatr | JayF, getting them pxe booted, once they boot it tends to go well, although I've seen nodes that boot successfully and never return, it's hard to keep track because the issues are infrequent but always showstoppers when they happen. | 17:10 |
| JayF | Makes sense. | 17:10 |
| JayF | I'm not that familiar with how inspection works, to be honest | 17:10 |
| jkilpatr | http://scalelab.redhat.com/ | 17:10 |
| JayF | i.e. if inspector handles the pxe booting or if Ironic does | 17:10 |
| sambetts | JayF: just like cleaning | 17:10 |
| JayF | perfect | 17:10 |
| jkilpatr | JayF, neither am I, but it's a major problem for my team, so I'm going to learn. | 17:10 |
| sambetts | JayF: although we have our own DHCP server | 17:11 |
| JayF | so then have you tried with the options enabled to use ipxe pulling directly from swift? | 17:11 |
| JayF | IME pxe booting bottlenecks are rarely the dhcp server, more often TFTP or image downloads | 17:11 |
| *** harlowja has joined #openstack-ironic | 17:11 | |
| * jkilpatr goes looking for the swift performance metrics he saw on the perf mailing list a few days ago | 17:11 | |
| JayF | if you're not using ipxe, try enabling it -- if you're already using ipxe, try enabling the settings for pulling the images directly from swift over http | 17:11 |
| sambetts | dtantsur: did we get iPXE working for inspection yet? | 17:12 |
| dtantsur | sambetts, for looooong time | 17:12 |
| dtantsur | it's the default in tripleo | 17:12 |
| sambetts | oh awesome :D | 17:12 |
| jkilpatr | ok so I'm already using that good. | 17:12 |
| dtantsur | sambetts, JayF, not that even iPXE uses TFTP to bootstrap.. and DHCP can be a bottleneck too | 17:13 |
| *** rbartal has quit IRC | 17:13 | |
| sambetts | jkilpatr: do you have native iPXE on your hardware or are chain loading it | 17:13 |
| JayF | dtantsur: IME with OnMetal, we rarely saw DHCP itself be the bottleneck. We supported hundreds of nodes pxe booting simultaneously with one isc dhcp server | 17:13 |
| JayF | dtantsur: we /did/ see slow TFTP and slow HTTP as a culprit | 17:13 |
| jkilpatr | sambetts, how would I tell? | 17:13 |
| JayF | dtantsur: particularly at one point when we had our http and tftp servers co-located on the same box. The http sessions would choke out tftp at high scale | 17:14 |
| dtantsur | oh interesting | 17:14 |
| dtantsur | JayF, we do have everything co-located on one box in tripleo | 17:14 |
| sambetts | good question :-P dtantsur do you know how to tell if your using native iPXE or not? | 17:14 |
| JayF | dtantsur: bingo | 17:14 |
| JayF | dtantsur: I got $20 that says if you move http and tftp to different boxes (tftp can colocate with dhcp, but not http), you scale up 10x higher without issue | 17:15 |
| dtantsur | sambetts, jkilpatr, either by looking at xinet.d logs or by looking at the screen | 17:15 |
| dtantsur | jkilpatr, see JayF comment ^^^ | 17:15 |
| jkilpatr | JayF, that's useful, but undercloud. | 17:15 |
| dtantsur | jkilpatr, well, we can start splitting undercloud........ | 17:15 |
| JayF | dtantsur: obviously all this knowledge is from our static setup with dhcp_provider=None, but the issues match ours related to performance | 17:16 |
| * dtantsur imagines tripleo folks' reaction | 17:16 | |
| jkilpatr | be careful that's how you start a riot. | 17:16 |
| JayF | jkilpatr: I don't use or really like tripleo tbh, I'm just giving general systems+ironic advice :D | 17:16 |
| jkilpatr | JayF, thanks for the info, it's very valuable. Those of us stuck with tripleo just need to figure this issue out. | 17:16 |
| dtantsur | jkilpatr, you may want to talk to dsneddon about his routed networks work.. maybe he'll think about separating ironic nodes per network or something | 17:17 |
| *** baoli has quit IRC | 17:17 | |
| JayF | Seriously though, colocating TFTP and HTTP is a scaling nightmare | 17:18 |
| jkilpatr | dtantsur, I guess by the time we get to dozens or hundreds of nodes more than one node for an undercloud isn't such a hard sell, but at 20-30 nodes even blowing 1 on an undercloud can be difficult to stomach. | 17:18 |
| JayF | because TCP connections from the HTTP server choke out TFTP and break PXE booting | 17:18 |
| JayF | a single overloaded HTTP server just makes all of them equally slow to boot, but they get there | 17:18 |
| JayF | and overloaded server running TFTP leads to full timeouts and failures | 17:18 |
| *** derekh has quit IRC | 17:18 | |
| jkilpatr | right now I'll settle for just making it possible to rate limit these things and get the job done. | 17:19 |
| jkilpatr | which I guess is an option already, so I'm off to test that | 17:19 |
| jkilpatr | talk to you all a few hundred introspections from now. | 17:19 |
| jkilpatr | thanks a lot for the help. | 17:19 |
| dtantsur | good luck :) | 17:20 |
| JayF | even the numbers match fwiw as well | 17:20 |
| JayF | we saw cleaning max out at about 10-20 machines at a time until we split tftp and http servers | 17:20 |
| *** evgenyzhurko has quit IRC | 17:21 | |
| *** intr1nsic has joined #openstack-ironic | 17:21 | |
| dtantsur | oh, interesting! | 17:22 |
| dtantsur | that's why it's so awesome to have real operators on the team :) | 17:23 |
| *** priteau has joined #openstack-ironic | 17:23 | |
| sambetts | \o/ | 17:23 |
| JayF | I think it was jroll who figured this one out | 17:23 |
| JayF | way back, many months >1y ago | 17:23 |
| JayF | jroll is more pr0 at operating software than he lets on, don't let him fool you | 17:23 |
| JayF | lol | 17:23 |
| dtantsur | :D | 17:23 |
| *** sabeen has quit IRC | 17:24 | |
| *** jerrygb has joined #openstack-ironic | 17:25 | |
| *** intr1nsic is now known as matt_welch | 17:27 | |
| *** jerrygb_ has quit IRC | 17:28 | |
| *** ralonsoh has joined #openstack-ironic | 17:30 | |
| *** rpioso has quit IRC | 17:33 | |
| *** e0ne has quit IRC | 17:34 | |
| * TheJulia walks in and grumbles about BCN StackPlague | 17:36 | |
| NobodyCam | :( | 17:36 |
| *** Goneri has joined #openstack-ironic | 17:36 | |
| JayF | TheJulia: I got the flu without even having to go to the summit | 17:36 |
| JayF | TheJulia: it's transferred through irc, apparently | 17:36 |
| TheJulia | JayF: \o/ | 17:36 |
| *** rpioso has joined #openstack-ironic | 17:39 | |
| mgould | TheJulia: o/ | 17:39 |
| TheJulia | We like need... a wave but somehow also indicating a hazmat suit is in use or something | 17:42 |
| JayF | [o]/ | 17:42 |
| TheJulia | perfect! | 17:43 |
| mgould | [o]/ | 17:43 |
| NobodyCam | lol | 17:43 |
| mariojv | [⚠ ]/ | 17:43 |
| *** zackf has quit IRC | 17:44 | |
| NobodyCam | [+][+][+] | 17:44 |
| TheJulia | Now we just need to convert pixie boots to ascii art! | 17:45 |
| mgould | #pixiesay I scorn your limited 7-bit character sets | 17:50 |
| mgould | dammit | 17:51 |
| *** hpe-hj has quit IRC | 17:51 | |
| NobodyCam | TheJulia: http://paste.openstack.org/show/HpFcZEWMj5fHnblcTXBs/ | 17:51 |
| mgould | did pixiebot's host go away again? | 17:51 |
| *** krtaylor has quit IRC | 17:52 | |
| openstackgerrit | Merged openstack/ironic: TrivialFix: Fix typo in config file https://review.openstack.org/377307 | 17:52 |
| *** jtomasek_ has quit IRC | 17:53 | |
| *** zackf has joined #openstack-ironic | 17:54 | |
| lucasagomes | mgould, it did :-( It was in an internal vm before | 17:56 |
| lucasagomes | which is now dead | 17:56 |
| *** vishwanathj has quit IRC | 17:56 | |
| lucasagomes | we should find a new home for it | 17:56 |
| *** mgoddard_ has quit IRC | 17:57 | |
| mgould | this whole "cloud computing" thing is far too unreliable, it'll never catch on | 17:58 |
| sambetts | haha you need to turn pixebot into cattle ;) | 17:58 |
| lucasagomes | ++ | 17:58 |
| sambetts | its not a pet | 17:59 |
| sambetts | :-P | 17:59 |
| lucasagomes | for users it's happy land but for people working on the cloud it's a nightmare | 17:59 |
| mgould | highly-available IRC bots as a service | 17:59 |
| mgould | with service discovery and failover and leader-election | 17:59 |
| lucasagomes | users should learn how to play a guitar, talk to friends, socialize in general when they can't access their email | 17:59 |
| lucasagomes | so ppl like us could live happier | 17:59 |
| lucasagomes | :-) | 17:59 |
| mgould | and on that note, /me -> home :-) | 18:00 |
| mgould | good night, everyone! | 18:00 |
| *** mgould is now known as mgould|afk | 18:00 | |
| NobodyCam | night mgould | 18:00 |
| lucasagomes | mgould, g'night | 18:00 |
| *** [1]rpioso has joined #openstack-ironic | 18:00 | |
| *** ralonsoh has quit IRC | 18:00 | |
| *** rpioso has quit IRC | 18:01 | |
| *** athomas has quit IRC | 18:02 | |
| TheJulia | IRC bots + leader election could get challenging, what if multiple bots decide they are the leaders? | 18:02 |
| dtantsur | sounds like fun, let's do it! | 18:03 |
| TheJulia | And if they are the leaders! are they more than puppets? | 18:03 |
| TheJulia | (since, most irc bots are scripted puppets... of sorts) | 18:04 |
| TheJulia | NobodyCam: \o/ for ascii pixie botos | 18:04 |
| NobodyCam | hehehehe :p | 18:04 |
| *** dougs1 has quit IRC | 18:06 | |
| *** jjohnson2 has joined #openstack-ironic | 18:07 | |
| JayF | TheJulia: leader election via irc is easy | 18:07 |
| JayF | TheJulia: whoever has the nickname wins | 18:07 |
| TheJulia | yes, but what if there are multiple splits... | 18:08 |
| JayF | TheJulia: and they can join a dummy channel while waiting. if the nickname goes away, whoever wins the race is the new bot and joins all the channels | 18:08 |
| TheJulia | and desynced clocks on irc servers | 18:08 |
| JayF | TheJulia: there's no state. it doesn't matter. You'd have maximum bot service on both sides of the split | 18:08 |
| JayF | when the split resolves and one bot is forced to change name, it's not the leader anymore and parts all channels except the coordination one | 18:08 |
| TheJulia | eh, true | 18:08 |
| *** lucasagomes is now known as lucas-afk | 18:09 | |
| TheJulia | well, I thought on desync conditions or returning from a split, the youngest session to the ircd gets kicked | 18:09 |
| TheJulia | to resolve the conflict | 18:09 |
| JayF | I mean, how it's resolved doesn't matter, as long as the IRC server end state is "exactly one bot with the nickname" | 18:10 |
| TheJulia | IRC server operators would hate us :) This sounds like a grand idea! | 18:11 |
| JayF | lol "it looks like you're running a botnet" | 18:11 |
| JayF | well, I am, but you know, the good kind! | 18:11 |
| TheJulia | lol | 18:12 |
| dtantsur | :D | 18:13 |
| dtantsur | ok, I'm calling it a day. g'night! | 18:13 |
| *** dtantsur is now known as dtantsur|afk | 18:13 | |
| NobodyCam | night dtantsur|afk | 18:22 |
| *** baoli has joined #openstack-ironic | 18:23 | |
| *** baoli_ has joined #openstack-ironic | 18:24 | |
| *** baoli has quit IRC | 18:29 | |
| *** baoli_ has quit IRC | 18:29 | |
| *** zackf has quit IRC | 18:31 | |
| *** baoli has joined #openstack-ironic | 18:35 | |
| * sambetts -> home | 18:35 | |
| *** sambetts is now known as sambetts|afk | 18:35 | |
| NobodyCam | night sambetts :) | 18:35 |
| sambetts|afk | o/ NobodyCam | 18:35 |
| *** zackf has joined #openstack-ironic | 18:37 | |
| *** rama_y has quit IRC | 18:41 | |
| openstackgerrit | Merged openstack/ironic-specs: Boot from Volume - Reference Drivers https://review.openstack.org/294995 | 18:43 |
| JayF | TheJulia: ^ congrats | 18:44 |
| NobodyCam | nice :) | 18:45 |
| *** krtaylor has joined #openstack-ironic | 18:49 | |
| *** rbrud has joined #openstack-ironic | 18:55 | |
| *** aNuposic has quit IRC | 18:56 | |
| *** madorn has joined #openstack-ironic | 18:57 | |
| *** sabeen has joined #openstack-ironic | 18:57 | |
| openstackgerrit | Merged openstack/ironic: Use kvm for ironic VMs when possible https://review.openstack.org/389109 | 18:59 |
| *** rama_y has joined #openstack-ironic | 19:00 | |
| *** Goneri has quit IRC | 19:06 | |
| *** strigazi is now known as strigazi_AFK | 19:09 | |
| *** Administrator__ has joined #openstack-ironic | 19:11 | |
| *** zhugaoxiao has quit IRC | 19:15 | |
| *** MattMan has quit IRC | 19:19 | |
| *** portdirect is now known as harbor_log | 19:24 | |
| *** bnemec has joined #openstack-ironic | 19:28 | |
| *** zackf has quit IRC | 19:31 | |
| *** clsacramento has joined #openstack-ironic | 19:33 | |
| *** ohamada has quit IRC | 19:33 | |
| *** zackf has joined #openstack-ironic | 19:34 | |
| *** causten has quit IRC | 19:34 | |
| rloo | hey, is the IPA gate still down? | 19:34 |
| *** rama_y has quit IRC | 19:35 | |
| *** matt_welch has quit IRC | 19:37 | |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic-specs: Add a 'preview' tox target https://review.openstack.org/392285 | 19:39 |
| *** causten has joined #openstack-ironic | 19:40 | |
| *** mtreinish has quit IRC | 19:40 | |
| *** rcernin has joined #openstack-ironic | 19:41 | |
| *** mtreinish has joined #openstack-ironic | 19:43 | |
| *** moshele has joined #openstack-ironic | 19:44 | |
| openstackgerrit | Milan Kováčik proposed openstack/ironic-inspector: Add API for listing all introspection statuses https://review.openstack.org/344921 | 19:44 |
| *** dougs has joined #openstack-ironic | 19:45 | |
| rloo | JayF, jroll: ^^ IronicWhiteBoard shows IPA gate as down. Is that true? | 19:47 |
| *** dougs1 has joined #openstack-ironic | 19:47 | |
| *** baoli has quit IRC | 19:47 | |
| *** baoli_ has joined #openstack-ironic | 19:47 | |
| jroll | rloo: no, this fixed it, sorry https://github.com/openstack-infra/project-config/commit/a044716224a00e174bf7f4faede7d6e57011f68a | 19:48 |
| rloo | jroll: thx | 19:48 |
| jroll | rloo: updated whiteboard | 19:49 |
| jroll | sorry, signed off before verifying yesterday | 19:49 |
| rloo | jroll: no worries | 19:49 |
| *** dougs1 has quit IRC | 19:49 | |
| *** dougs has quit IRC | 19:50 | |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic-specs: Amend dynamic iPXE configuration spec https://review.openstack.org/392290 | 19:50 |
| openstackgerrit | Milan Kováčik proposed openstack/ironic-inspector: Add API for listing all introspection statuses https://review.openstack.org/344921 | 19:50 |
| *** dougs has joined #openstack-ironic | 19:50 | |
| *** dougs1 has joined #openstack-ironic | 19:51 | |
| *** baoli has joined #openstack-ironic | 19:51 | |
| *** baoli_ has quit IRC | 19:52 | |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic-specs: Add a 'preview' tox target https://review.openstack.org/392285 | 19:52 |
| *** dougs has quit IRC | 19:55 | |
| rloo | JayF, jroll: an easy one. has been +A'd several times | 19:56 |
| *** Goneri has joined #openstack-ironic | 19:57 | |
| milan | night Ironic! :) | 19:57 |
| jroll | rloo: which one? | 19:58 |
| rloo | jroll: oops, https://review.openstack.org/338117 | 19:58 |
| patchbot | patch 338117 - ironic - Add IRONIC_TEST_NETWORK_NAME devstack plugin var | 19:58 |
| *** mmedvede has quit IRC | 19:58 | |
| * rloo still trying to get back into the swing of things... | 19:58 | |
| jroll | so many conflicts on that :/ | 19:59 |
| * jroll lands it anyway | 19:59 | |
| jroll | oh, that depends on https://review.openstack.org/#/c/334382/ | 20:00 |
| patchbot | patch 334382 - ironic - Allow to use FLAT network setup | 20:00 |
| * jroll not sure if up to review that right now | 20:01 | |
| rloo | jroll: ugh. missed that. and yeah, not sure i feel like reviewing that today | 20:01 |
| *** baoli has quit IRC | 20:04 | |
| *** baoli has joined #openstack-ironic | 20:05 | |
| *** zackf has quit IRC | 20:10 | |
| openstackgerrit | Mario Villaplana proposed openstack/ironic: Add ability to time out on individual clean steps https://review.openstack.org/391554 | 20:11 |
| JayF | mariojv: I am almost certain you can't use actual signals for that | 20:12 |
| *** dprince has quit IRC | 20:12 | |
| JayF | mariojv: because a single given conductor could be handling multiple nodes simultaneously | 20:12 |
| JayF | right? | 20:12 |
| mariojv | JayF: the signal's only used for sync clean steps | 20:12 |
| JayF | .o(there are sync clean steps?) | 20:13 |
| JayF | even still, doesn't my comment still apply? the conductor could be running a step on multiple nodes, so a single unix signal would effect them all, right/ | 20:13 |
| *** rama_y has joined #openstack-ironic | 20:14 | |
| mariojv | my understanding was that a sync clean step was a blocking operation | 20:14 |
| *** zackf has joined #openstack-ironic | 20:14 | |
| JayF | but the conductor uses worker threads for stuff like that | 20:14 |
| JayF | but I think signals are process-wide, not thread specific | 20:15 |
| mariojv | ah, yeah, and the signals are per-... yup | 20:15 |
| JayF | also what would happen if a user just did a kill -ALRM `pidof conductor` | 20:15 |
| * mariojv pokes around looking at docs | 20:15 | |
| JayF | I really dislike use of actual unix signals for this for that reason if none other; that someone could send the signal outside of the process and cause that to happen | 20:16 |
| jroll | even with syncronous clean steps, the greenthreads could yield, those aren't completely blocking | 20:16 |
| rloo | mariojv: I haven't looked -- but we shouldn't have long-running sync clean steps? | 20:20 |
| *** baoli has quit IRC | 20:20 | |
| *** baoli has joined #openstack-ironic | 20:21 | |
| mariojv | i'm not sure if they're actually used, but there is logic in the conductor manager to handle them | 20:21 |
| mariojv | i'm having a hard time coming up with a way to time those out other than messing with the internals of the task manager though | 20:23 |
| *** dougs1 has quit IRC | 20:23 | |
| mariojv | i'm not sure that would be possible without overloading the worker pool either, 1 greenthread per clean step might not be a good idea | 20:29 |
| JayF | change the goalposts, at least for the initial patch, and only support async steps? | 20:29 |
| *** baoli has quit IRC | 20:30 | |
| *** baoli_ has joined #openstack-ironic | 20:30 | |
| mariojv | i can do that. one downside to that is it makes it difficult for creators of clean steps to see that sync steps won't support it, other than documenting well | 20:30 |
| mariojv | the only difference codified in code between sync vs. async that i can tell is that async returns CLEANWAIT | 20:31 |
| mariojv | meaning you can't error out if someone tries to make a sync clean step with a timeout; it'll just not work | 20:31 |
| mariojv | (not work, meaning the timeout won't work) | 20:32 |
| JayF | I'm suspicious | 20:33 |
| JayF | that asyncc vs sync | 20:33 |
| JayF | is more in-band vs oob | 20:33 |
| JayF | because for in band cleaning steps I can't imagine why we'd ever put the node in cleanwait | 20:33 |
| *** priteau has quit IRC | 20:34 | |
| *** priteau has joined #openstack-ironic | 20:35 | |
| mariojv | you mean for OOB, right? | 20:35 |
| JayF | yeah | 20:36 |
| mariojv | would there be anything wrong with using this? http://eventlet.net/doc/modules/timeout.html i think that'd work since the exception would be raised in the current greenthread | 20:38 |
| *** baoli has joined #openstack-ironic | 20:39 | |
| *** priteau has quit IRC | 20:39 | |
| JayF | I'm not familiar enough with eventlet to have an educated opinion on that | 20:40 |
| *** baoli_ has quit IRC | 20:40 | |
| *** jkilpatr has quit IRC | 20:41 | |
| JayF | mariojv: re-reading it, it seems sane because it'll happen in the same greenthread | 20:42 |
| *** baoli has quit IRC | 20:43 | |
| *** zackf has quit IRC | 20:51 | |
| *** aNuposic has joined #openstack-ironic | 20:52 | |
| *** baoli has joined #openstack-ironic | 20:53 | |
| *** zackf has joined #openstack-ironic | 20:53 | |
| *** baoli_ has joined #openstack-ironic | 20:54 | |
| *** e0ne has joined #openstack-ironic | 20:55 | |
| *** baoli has quit IRC | 20:58 | |
| *** priteau has joined #openstack-ironic | 20:59 | |
| *** mmedvede has joined #openstack-ironic | 21:03 | |
| *** absubram has quit IRC | 21:12 | |
| *** Goneri has quit IRC | 21:14 | |
| TheJulia | JayF: Awesome! *happy dance* | 21:19 |
| * TheJulia goes and finds more sudafed | 21:20 | |
| *** e0ne_ has joined #openstack-ironic | 21:25 | |
| *** e0ne has quit IRC | 21:27 | |
| *** zackf has quit IRC | 21:34 | |
| *** e0ne_ has quit IRC | 21:34 | |
| *** jheroux has quit IRC | 21:35 | |
| *** zackf has joined #openstack-ironic | 21:35 | |
| *** krtaylor has quit IRC | 21:37 | |
| *** sabeen has quit IRC | 21:37 | |
| *** maratoid_ has joined #openstack-ironic | 21:38 | |
| *** maratoid_ has quit IRC | 21:40 | |
| openstackgerrit | Merged openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/388351 | 21:40 |
| *** maratoid_ has joined #openstack-ironic | 21:41 | |
| *** sabeen has joined #openstack-ironic | 21:46 | |
| *** jerrygb has quit IRC | 21:46 | |
| *** jcoufal has quit IRC | 21:52 | |
| openstackgerrit | Mario Villaplana proposed openstack/ironic: Add ability to time out on individual clean steps https://review.openstack.org/391554 | 21:53 |
| mariojv | JayF: switched that to use the eventlet method ^ | 21:53 |
| JayF | cool :D | 21:54 |
| *** zackf has quit IRC | 21:54 | |
| *** jkilpatr has joined #openstack-ironic | 21:55 | |
| *** jerrygb has joined #openstack-ironic | 22:04 | |
| *** sabeen has quit IRC | 22:05 | |
| *** galyna has quit IRC | 22:06 | |
| *** jerrygb has quit IRC | 22:06 | |
| *** priteau has quit IRC | 22:06 | |
| *** zackf has joined #openstack-ironic | 22:10 | |
| JayF | jroll: https://review.openstack.org/#/c/392114/2 do you think smoething like this needs a microversion bump? | 22:11 |
| patchbot | patch 392114 - ironic - API: lookup() ignore malformed MAC addresses | 22:11 |
| *** zackf has quit IRC | 22:13 | |
| *** zackf has joined #openstack-ironic | 22:18 | |
| openstackgerrit | Merged openstack/ironic-python-agent: Remove duplicated logging in configdrive creation https://review.openstack.org/390158 | 22:18 |
| *** zackf has quit IRC | 22:19 | |
| *** baoli_ has quit IRC | 22:24 | |
| openstackgerrit | Merged openstack/ironic: Correct log the node UUID on failure https://review.openstack.org/391031 | 22:25 |
| *** baoli has joined #openstack-ironic | 22:25 | |
| *** rbrud has quit IRC | 22:25 | |
| *** g3ek has quit IRC | 22:30 | |
| *** rcernin has quit IRC | 22:31 | |
| *** baoli has quit IRC | 22:33 | |
| *** baoli has joined #openstack-ironic | 22:33 | |
| *** g3ek has joined #openstack-ironic | 22:40 | |
| *** jkilpatr has quit IRC | 22:40 | |
| *** jkilpatr has joined #openstack-ironic | 22:40 | |
| *** wajdi has quit IRC | 22:47 | |
| *** [1]cdearborn has joined #openstack-ironic | 22:50 | |
| *** baoli_ has joined #openstack-ironic | 22:53 | |
| *** baoli has quit IRC | 22:56 | |
| *** shobhar_ has joined #openstack-ironic | 23:00 | |
| *** causten has quit IRC | 23:00 | |
| *** causten has joined #openstack-ironic | 23:00 | |
| *** ElCoyote_ has quit IRC | 23:00 | |
| *** baoli has joined #openstack-ironic | 23:01 | |
| *** baoli_ has quit IRC | 23:01 | |
| *** cdearborn has quit IRC | 23:03 | |
| *** causten has quit IRC | 23:05 | |
| *** baoli has quit IRC | 23:07 | |
| *** zackf has joined #openstack-ironic | 23:16 | |
| *** zackf has quit IRC | 23:19 | |
| *** yuanying has joined #openstack-ironic | 23:37 | |
| *** aNuposic has quit IRC | 23:45 | |
| *** rama_y has quit IRC | 23:48 | |
| *** [1]rpioso has quit IRC | 23:52 | |
| *** moshele has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!