| *** naohirot has joined #openstack-ironic | 00:04 | |
| NobodyCam | hey hey krotscheck :) awesome Thank you ... please also feel free to just them to : https://etherpad.openstack.org/p/liberty-ironic-contributor-meetup | 00:05 |
|---|---|---|
| NobodyCam | :) | 00:05 |
| NobodyCam | lol only days left at this point... :-p | 00:05 |
| *** achanda_ has joined #openstack-ironic | 00:06 | |
| *** Marga_ has quit IRC | 00:10 | |
| *** achanda has quit IRC | 00:10 | |
| *** Marga_ has joined #openstack-ironic | 00:10 | |
| *** achanda_ has quit IRC | 00:11 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 00:12 |
| *** derekh has quit IRC | 00:12 | |
| *** chlong_ has joined #openstack-ironic | 00:20 | |
| *** mtanino has quit IRC | 00:27 | |
| *** david-lyle has quit IRC | 00:28 | |
| *** rwsu has quit IRC | 00:54 | |
| rloo | hi NobodyCam, have you been creating the etherpads for the summit? | 01:06 |
| rloo | NobodyCam: was thinking this might be helpful for the driver composition session: http://paste.openstack.org/show/221330/ | 01:07 |
| openstackgerrit | SHIGEMATSU Mitsuhiro proposed openstack/bifrost: Fix documentation in bifrost https://review.openstack.org/182514 | 01:13 |
| TheJulia | woot, doc fix | 01:17 |
| TheJulia | now, If I can just figure out why the CI job does not want to work :( | 01:17 |
| pshige | :) | 01:18 |
| pshige | I will check it too. | 01:18 |
| *** zz_natorious is now known as natorious | 01:31 | |
| *** Sukhdev has joined #openstack-ironic | 01:33 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 01:34 |
| *** rloo has quit IRC | 01:35 | |
| * TheJulia seeks help from infra folk... or a reproduction of a bare-trusty node that can be put under a microscope | 01:45 | |
| *** Sukhdev_ has joined #openstack-ironic | 01:53 | |
| *** Sukhdev has quit IRC | 01:54 | |
| *** chenglch has joined #openstack-ironic | 02:04 | |
| devananda | rloo, nice grid | 02:28 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 02:34 |
| jroll | devananda: more like big scary grid :P | 02:42 |
| jroll | does our ipxe stuff leave httpd setup to the user? | 02:43 |
| jroll | I notice the docs don't say anything about it, except point http stuff to the conductor node | 02:43 |
| jroll | you probably want http and tftp separated at some level of scale, else you'll start dropping udp packets and tftp will get unhappy | 02:44 |
| jroll | something I learned today :) | 02:44 |
| jroll | luca<tab> boo. | 02:45 |
| *** achanda has joined #openstack-ironic | 02:47 | |
| devananda | jroll: yea, up to user | 02:52 |
| devananda | jroll: just like any other service (tftp, nfs, dnsmasq) | 02:52 |
| jroll | devananda: I really want to sprint on our docs | 02:52 |
| jroll | right | 02:52 |
| devananda | jroll: dooo eeet | 02:52 |
| devananda | jroll: oh. you mean "lets all have a doc sprint". | 02:53 |
| jroll | devananda: some of both | 02:53 |
| jroll | turns out I can't just disappear for a week | 02:53 |
| devananda | I'd love to have 4 days of developer meetup, without all the distractions of the summit | 02:53 |
| jroll | the biggest thing that bugs me about our docs is the lack of "what to do when $thing breaks in this way" | 02:53 |
| jroll | yes | 02:53 |
| devananda | then we could have a day to sprint on docs, for example | 02:54 |
| devananda | and a day on logging and notifications | 02:54 |
| jroll | we tried that once and we all sat around silent working on $priorities :( | 02:54 |
| devananda | blah | 02:54 |
| jroll | meh, logging is fine. would love notification support though | 02:54 |
| devananda | logging is terrible | 02:54 |
| devananda | if it was fine, you'd know what to do when $thing broke :p | 02:54 |
| jroll | eh, true | 02:54 |
| jroll | see, I'm biased towards mostly-working things | 02:55 |
| jroll | I need to stand up upstream ironic without any automation at some point | 02:56 |
| devananda | we all do | 02:56 |
| devananda | and then we need to document how its done in various env's | 02:56 |
| jroll | right, ironic is very susceptible to environmental changes | 02:57 |
| jroll | read: networking is hard | 02:57 |
| jroll | also hardware is hard | 02:57 |
| jroll | :P | 02:57 |
| devananda | computers are hard | 02:58 |
| jroll | also that | 02:59 |
| jroll | what are the chances bifrost could be less opinionated and also stand up an environment that works *with* openstack services? | 02:59 |
| devananda | define less opinionated? | 03:00 |
| jroll | and then put a bunch of effort into environment discovery and making things just work (tm) | 03:00 |
| *** kan has joined #openstack-ironic | 03:00 | |
| jroll | like, allow for glance/neutron/nova if the user chooses | 03:00 |
| *** alexpilotti has quit IRC | 03:00 | |
| devananda | so, bifrost is 3 parts | 03:01 |
| devananda | first part actually intalls and configures ironic, mysql, rabbit, etc, and all the deps | 03:01 |
| devananda | but it configures them to run w/o the rest of openstack | 03:01 |
| devananda | so if you have the rest of openstack -- you dont need to run that step | 03:01 |
| jroll | and that's exposed as an option? | 03:02 |
| devananda | second step enrolls hardware. that should work just fine regardless of the presense of nova, neutron, etc -- except we'd need to teach that playbook to use keystone auth | 03:02 |
| jroll | maybe I just need to open source all the automation. | 03:02 |
| devananda | jroll: it's an ansible playbook. just don't run it | 03:02 |
| jroll | sure | 03:02 |
| devananda | third step does the deploy by directly calling ironic. so if you have nova, well, you dont need to do that either | 03:02 |
| devananda | so i really dont understand your question | 03:02 |
| jroll | "exposed" being like "docs say to skip this tag if you're running these things" | 03:03 |
| devananda | if you have already installed openstack and configured ironic ... you dont want to use bifrost | 03:03 |
| jroll | right | 03:03 |
| jroll | well | 03:03 |
| jroll | I guess I'm thinking that providing some automation would be cool | 03:03 |
| devananda | OTOH, I may want to use Ironic + Neutron, without Nova or Keystone or Glance. | 03:03 |
| jroll | whether that's in the openstack-ansible stuff or what | 03:03 |
| jroll | and have options for private/public cloud etc | 03:04 |
| jroll | seems hard, though | 03:04 |
| devananda | because Neutron ML2 extension gives me a convenient abstraction layer for configuring different vendor's switches (at least intheory) | 03:04 |
| devananda | jroll: automation to do what? | 03:04 |
| jroll | stand up ironic in a sane fashion | 03:04 |
| jroll | if you follow devstack or our docs, it may provide a working system, but not necessarily a good or scalable system | 03:05 |
| *** chlong has quit IRC | 03:05 | |
| devananda | oh, TheJulia, we should probably take a look at openstack-ansible and see how close that is to bifrost's install playbook, and whether we should add and/or share things there | 03:05 |
| openstackgerrit | Kan proposed openstack/python-ironicclient: Disable meaningless sort keys in list command https://review.openstack.org/177578 | 03:07 |
| *** Nisha has joined #openstack-ironic | 03:13 | |
| Haomeng|2 | devananda: fyi - db2 ci *ibm-db2-ci-ironic* is ready for ironic master branch | 03:15 |
| zhenguo | hi all, should ironic be integrated with horizon or have a seperate dashboard or tuscar-ui? I find currently devstack doesn't have dashboard enabled? | 03:23 |
| Haomeng|2 | zhenguo: yes, so far, there is no ironic ui in openstack dashboard, for tuscar-ui, it is used for tripleo I think, not focus on ironic ui only | 03:25 |
| *** Haomeng|2 is now known as Haomeng | 03:25 | |
| zhenguo | Haomeng: does tuscar-ui be used without tripleO, as a seperate dashboard for ironic? | 03:26 |
| zhenguo | Haoment: or without tripleO tuscar-ui can't work? | 03:27 |
| Haomeng | zhenguo: not sure, but I know, it is not full ui function to support ironic, such as the port CRUD etc | 03:27 |
| Haomeng | zhenguo: we can play with ironic via command line:) | 03:27 |
| Haomeng | zhenguo: I understand the ironic ui should support all ironic comands/apis | 03:28 |
| Haomeng | zhenguo: and including the ironic discovery functions | 03:28 |
| Haomeng | zhenguo: will discuss on summit with horizon team, hope horizon team can help to support | 03:29 |
| zhenguo | Haomeng: if ironic be used as nova driver, does it full supported to use with horizon to do provision? | 03:29 |
| zhenguo | Haomeng: oops, I can't attend the summit | 03:29 |
| Haomeng | zhenguo: yes, nova boot to kickoff the bm provision, but for node/port loading, have to run ironic command to prepare them ready by manually | 03:29 |
| Haomeng | zhenguo: np, we can see the discusstion result after the meeting | 03:30 |
| Haomeng | zhenguo: I will not attend also:) | 03:30 |
| *** links has joined #openstack-ironic | 03:30 | |
| Haomeng | zhenguo: maybe some horizon ui plugin expert is required from horizon team and work with our ironic team to create ironic ui together | 03:31 |
| zhenguo | Haomeng: you mean like tuscar-ui? | 03:31 |
| Haomeng | zhenguo: and we can have more functions to cover, such as physical node discovery and management and status monitoring, mac w&b list management etc | 03:32 |
| Haomeng | zhenguo: maybe | 03:32 |
| zhenguo | Haomeng: ok got it, thanks :) | 03:33 |
| *** stendulker has joined #openstack-ironic | 03:33 | |
| Haomeng | zhenguo: then we have total functions supportting on ui for user to management/monitoring physicl server even switch, and use ironic to provision hardware | 03:33 |
| Haomeng | zhenguo: welcome | 03:33 |
| zhenguo | Haomeng: if so, seems ironic should have a seperate dashboard, maybe use horizon lib like tuscar-ui. | 03:35 |
| Haomeng | zhenguo: yes | 03:36 |
| Haomeng | zhenguo: and as we know, ironic supports to run on standalone mode, so it is cool if have ui for standalone, then it will be easy to use to deploy understand cloud without other nova/neutron heavy openstack components | 03:37 |
| zhenguo | Haomeng: I'm not sure whether tuscar-ui can be configed using ironic on standalone mode. need to confirm with tripleO folks. | 03:43 |
| * devananda gives up on trying to make any sense of openwsman code | 03:43 | |
| devananda | lintan_: please make AMT driver work with openwsman 2.4.3. I'd like to use this NUC that I conveniently have ... | 03:46 |
| openstackgerrit | Kan proposed openstack/python-ironicclient: Disable meaningless sort keys in list command https://review.openstack.org/177578 | 03:47 |
| *** david-lyle has joined #openstack-ironic | 03:48 | |
| *** chlong_ has quit IRC | 03:52 | |
| *** Sukhdev_ has quit IRC | 03:53 | |
| devananda | lintan_: welp, I've found your discussion with klaus. Never mind... | 03:54 |
| *** coolsvap|afk is now known as coolsvap | 03:56 | |
| *** natorious is now known as zz_natorious | 03:58 | |
| devananda | lintan_: fwiw, amttool (from the amtterm) package is soooo much simpler than openwsman! | 03:59 |
| *** rameshg87 has joined #openstack-ironic | 04:02 | |
| *** stendulker_ has joined #openstack-ironic | 04:05 | |
| *** stendulker has quit IRC | 04:05 | |
| *** kalpase has joined #openstack-ironic | 04:17 | |
| *** stendulker has joined #openstack-ironic | 04:31 | |
| *** stendulker_ has quit IRC | 04:31 | |
| rameshg87 | good morning ironic | 04:34 |
| *** lazy_prince has joined #openstack-ironic | 04:43 | |
| devananda | victor_lowther: https://bugs.launchpad.net/ironic/+bug/1454492 | 04:54 |
| openstack | Launchpad bug 1454492 in Ironic "openwsman library is not greenthread-aware, leads to DRAC driver lockup" [Medium,Triaged] | 04:54 |
| * devananda goes to bed now | 04:55 | |
| pshige | devananda: night! | 05:12 |
| pshige | rameshg87: morining :) | 05:12 |
| *** achanda_ has joined #openstack-ironic | 05:12 | |
| *** hemna_ has joined #openstack-ironic | 05:13 | |
| *** ir2ivps9 has quit IRC | 05:16 | |
| rameshg87 | morning pshige | 05:16 |
| *** mikal_ has joined #openstack-ironic | 05:17 | |
| *** killer_prince has joined #openstack-ironic | 05:19 | |
| *** yog__ has joined #openstack-ironic | 05:21 | |
| *** lazy_prince has quit IRC | 05:22 | |
| *** coolsvap- has joined #openstack-ironic | 05:23 | |
| *** lazy_prince has joined #openstack-ironic | 05:23 | |
| *** chlong has joined #openstack-ironic | 05:23 | |
| *** killer_prince has quit IRC | 05:24 | |
| *** achanda has quit IRC | 05:26 | |
| *** coolsvap has quit IRC | 05:26 | |
| *** hemna has quit IRC | 05:26 | |
| *** anteaya has quit IRC | 05:26 | |
| *** mikal has quit IRC | 05:27 | |
| *** anteaya has joined #openstack-ironic | 05:28 | |
| *** chlong has quit IRC | 05:34 | |
| *** ir2ivps9 has joined #openstack-ironic | 05:36 | |
| *** chlong has joined #openstack-ironic | 05:48 | |
| *** lazy_prince has quit IRC | 05:55 | |
| *** lazy_prince has joined #openstack-ironic | 05:57 | |
| *** ukalifon has joined #openstack-ironic | 06:01 | |
| *** kalpase has quit IRC | 06:04 | |
| *** ukalifon has quit IRC | 06:06 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: grub2 bootloader support for uefi boot mode https://review.openstack.org/166192 | 06:10 |
| *** jrist has quit IRC | 06:14 | |
| *** dlpartain has joined #openstack-ironic | 06:15 | |
| *** ukalifon has joined #openstack-ironic | 06:15 | |
| *** dlpartain has left #openstack-ironic | 06:23 | |
| *** chlong has quit IRC | 06:24 | |
| *** yog__ has quit IRC | 06:24 | |
| *** jcoufal has joined #openstack-ironic | 06:25 | |
| *** chlong has joined #openstack-ironic | 06:27 | |
| *** killer_prince has joined #openstack-ironic | 06:28 | |
| *** lazy_prince has quit IRC | 06:31 | |
| *** killer_prince has quit IRC | 06:33 | |
| *** yog__ has joined #openstack-ironic | 06:37 | |
| *** kalpase has joined #openstack-ironic | 06:39 | |
| *** lazy_prince has joined #openstack-ironic | 06:42 | |
| *** smallbig has left #openstack-ironic | 06:49 | |
| *** takadayuiko has joined #openstack-ironic | 06:49 | |
| *** achanda_ has quit IRC | 06:55 | |
| *** kalpase has quit IRC | 07:00 | |
| openstackgerrit | Kan proposed openstack/python-ironicclient: Disable meaningless sort keys in list command https://review.openstack.org/177578 | 07:06 |
| *** kalpase has joined #openstack-ironic | 07:07 | |
| *** yog__ has quit IRC | 07:11 | |
| *** dlpartain has joined #openstack-ironic | 07:13 | |
| *** dlpartain has left #openstack-ironic | 07:14 | |
| *** sandhya has joined #openstack-ironic | 07:15 | |
| *** kalpase1 has joined #openstack-ironic | 07:17 | |
| *** kalpase has quit IRC | 07:19 | |
| *** sandhya has quit IRC | 07:21 | |
| *** ifarkas has quit IRC | 07:22 | |
| *** ifarkas has joined #openstack-ironic | 07:23 | |
| *** Marga_ has quit IRC | 07:31 | |
| *** kalpase has joined #openstack-ironic | 07:33 | |
| *** kalpase1 has quit IRC | 07:35 | |
| *** romcheg has joined #openstack-ironic | 07:37 | |
| *** romcheg has quit IRC | 07:42 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Secure boot support for pxe_ilo driver https://review.openstack.org/154808 | 07:47 |
| *** jistr has joined #openstack-ironic | 07:49 | |
| *** dguerri is now known as _dguerri | 07:52 | |
| *** _dguerri is now known as dguerri | 07:52 | |
| openstackgerrit | Kan proposed openstack/ironic: Disable meaningless sort keys in list command https://review.openstack.org/182563 | 07:57 |
| *** derekh has joined #openstack-ironic | 08:03 | |
| *** saripurigopi has joined #openstack-ironic | 08:04 | |
| *** athomas has joined #openstack-ironic | 08:06 | |
| *** MattMan has joined #openstack-ironic | 08:13 | |
| *** ppal has joined #openstack-ironic | 08:14 | |
| *** saripurigopi_ has joined #openstack-ironic | 08:19 | |
| *** saripurigopi has quit IRC | 08:19 | |
| *** saripurigopi_ is now known as saripurigopi | 08:19 | |
| *** uggla__ has quit IRC | 08:20 | |
| *** uggla has joined #openstack-ironic | 08:20 | |
| *** Nisha has quit IRC | 08:22 | |
| *** e0ne has joined #openstack-ironic | 08:25 | |
| *** uggla_ has joined #openstack-ironic | 08:27 | |
| *** e0ne has quit IRC | 08:28 | |
| *** uggla has quit IRC | 08:29 | |
| *** lucasagomes has joined #openstack-ironic | 08:29 | |
| *** e0ne has joined #openstack-ironic | 08:30 | |
| *** e0ne has quit IRC | 08:31 | |
| *** uggla_ has quit IRC | 08:37 | |
| *** kalpase has quit IRC | 08:39 | |
| *** uggla has joined #openstack-ironic | 08:39 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Fix chainloading iPXE (undionly.kpxe) https://review.openstack.org/182387 | 08:44 |
| *** kalpase has joined #openstack-ironic | 08:45 | |
| *** edcragg has joined #openstack-ironic | 08:49 | |
| rameshg87 | lucasagomes: hi | 08:51 |
| lucasagomes | rameshg87, hi there! | 08:51 |
| rameshg87 | lucasagomes: the # on L858 was it intentional - https://review.openstack.org/#/c/182387/2/doc/source/deploy/install-guide.rst | 08:51 |
| rameshg87 | ?? | 08:51 |
| rameshg87 | # dhcp-match=set:ipxe,175 | 08:52 |
| lucasagomes | rameshg87, yeah, both of them create the same tag | 08:52 |
| lucasagomes | so you gotta use one or another | 08:52 |
| rameshg87 | lucasagomes: okay so both works, right ? | 08:52 |
| lucasagomes | either you create the ipxe tag by looking at dhcp option 175 or at the userclass | 08:52 |
| lucasagomes | yup | 08:52 |
| rameshg87 | lucasagomes: so is it advised to keep both ? | 08:53 |
| rameshg87 | because in the bug seems like "dhcp-match:set:ipxe,175" didn't work for you. is it ? | 08:54 |
| lucasagomes | rameshg87, yeah, I've never seem that before. Only on that machine | 08:54 |
| lucasagomes | using the dhcp opt 175 wasn't matching | 08:54 |
| rameshg87 | okay, so should we ask folks to keep both (so that it works on all hardware) ? | 08:54 |
| lucasagomes | now I'm not sure what it can be, either it's something in the PXE image flashed in the nic | 08:55 |
| lucasagomes | that was setting that dhcp option and therefor not matching or something like that | 08:55 |
| lucasagomes | rameshg87, not sure how we keep both, unless we set different tags | 08:55 |
| lucasagomes | like 1 tag for each | 08:55 |
| lucasagomes | but, still that can cause more problems | 08:55 |
| lucasagomes | because then we would need to match on both tags and it won't work on that machine if we do | 08:56 |
| rameshg87 | oh okay. so only 1 tag works at a time ? either iPXE or 175 ? | 08:56 |
| lucasagomes | I not sure there's any automagically thing we can do here, we need operators to be able to configure their dhcp server to work with their machines | 08:57 |
| lucasagomes | rameshg87, yeah only 1 works | 08:57 |
| rameshg87 | okay | 08:57 |
| lucasagomes | and looking at the userclass worked for all the machines I tested so | 08:57 |
| lucasagomes | I left that by default | 08:57 |
| lucasagomes | (uncommented) | 08:57 |
| rameshg87 | okay got it | 08:58 |
| *** kalpase has left #openstack-ironic | 09:05 | |
| *** jrist has joined #openstack-ironic | 09:16 | |
| *** jrist has quit IRC | 09:16 | |
| *** jrist has joined #openstack-ironic | 09:16 | |
| *** e0ne has joined #openstack-ironic | 09:22 | |
| *** uggla_ has joined #openstack-ironic | 09:25 | |
| *** uggla has quit IRC | 09:25 | |
| *** uggla__ has joined #openstack-ironic | 09:25 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Ironic doesn't use cacert while talking to Swift https://review.openstack.org/182393 | 09:26 |
| *** uggla_ has quit IRC | 09:26 | |
| *** Nisha has joined #openstack-ironic | 09:30 | |
| *** killer_prince has joined #openstack-ironic | 09:33 | |
| *** athomas has quit IRC | 09:34 | |
| *** lazy_prince has quit IRC | 09:36 | |
| *** e0ne is now known as e0ne_ | 09:36 | |
| *** lazy_prince has joined #openstack-ironic | 09:37 | |
| *** e0ne_ is now known as e0ne | 09:38 | |
| *** killer_prince has quit IRC | 09:40 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic-specs: exposing driver capabilities https://review.openstack.org/182601 | 09:50 |
| *** yog__ has joined #openstack-ironic | 10:06 | |
| *** jamielennox is now known as jamielennox|away | 10:08 | |
| *** saripurigopi has quit IRC | 10:08 | |
| *** pas-ha has joined #openstack-ironic | 10:09 | |
| *** saripurigopi has joined #openstack-ironic | 10:15 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic-specs: exposing driver capabilities https://review.openstack.org/182601 | 10:19 |
| *** saripurigopi has quit IRC | 10:23 | |
| *** chenglch has quit IRC | 10:25 | |
| *** e0ne is now known as e0ne_ | 10:25 | |
| *** wendar has quit IRC | 10:25 | |
| *** wendar_ has joined #openstack-ironic | 10:25 | |
| *** Nisha has quit IRC | 10:26 | |
| *** e0ne_ is now known as e0ne | 10:26 | |
| *** early has quit IRC | 10:26 | |
| *** vdrok has quit IRC | 10:27 | |
| *** mikal_ has quit IRC | 10:28 | |
| *** early has joined #openstack-ironic | 10:28 | |
| *** mikal has joined #openstack-ironic | 10:28 | |
| *** EmilienM has quit IRC | 10:33 | |
| *** ekarlso has quit IRC | 10:33 | |
| *** degorenko has quit IRC | 10:33 | |
| *** degorenko has joined #openstack-ironic | 10:34 | |
| *** EmilienM has joined #openstack-ironic | 10:35 | |
| *** e0ne is now known as e0ne_ | 10:39 | |
| *** e0ne_ is now known as e0ne | 10:40 | |
| *** ekarlso has joined #openstack-ironic | 10:41 | |
| *** ekarlso has quit IRC | 10:45 | |
| *** ekarlso has joined #openstack-ironic | 10:45 | |
| *** lucasagomes is now known as lucas-afk | 10:57 | |
| *** yuanying has quit IRC | 11:16 | |
| *** e0ne is now known as e0ne_ | 11:26 | |
| *** mjturek1 has joined #openstack-ironic | 11:34 | |
| openstackgerrit | Brad P. Crochet proposed openstack/python-ironicclient: Implementation of baremetal power state commands https://review.openstack.org/172517 | 11:35 |
| openstackgerrit | Brad P. Crochet proposed openstack/python-ironicclient: Introduce openstackclient plugin https://review.openstack.org/171672 | 11:35 |
| openstackgerrit | Brad P. Crochet proposed openstack/python-ironicclient: Implementation of baremetal port create https://review.openstack.org/172461 | 11:35 |
| *** thrash|g0ne is now known as thrash | 11:35 | |
| *** stendulker has quit IRC | 11:38 | |
| * TheJulia yawns | 11:38 | |
| *** stendulker has joined #openstack-ironic | 11:39 | |
| openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/181204 | 11:41 |
| *** coolsvap- is now known as coolsvap|afk | 11:41 | |
| *** rameshg87 has left #openstack-ironic | 11:52 | |
| *** dprince has joined #openstack-ironic | 11:54 | |
| *** lazy_prince has quit IRC | 11:58 | |
| *** igordcard_ has quit IRC | 11:59 | |
| *** igordcard_ has joined #openstack-ironic | 11:59 | |
| *** stendulker has quit IRC | 12:02 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 12:05 |
| *** e0ne_ is now known as e0ne | 12:08 | |
| *** dlpartain has joined #openstack-ironic | 12:08 | |
| *** derekh_ has joined #openstack-ironic | 12:08 | |
| *** dlpartain has left #openstack-ironic | 12:10 | |
| *** derekh has quit IRC | 12:12 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 12:16 |
| *** vdrok has joined #openstack-ironic | 12:23 | |
| *** alexpilotti has joined #openstack-ironic | 12:32 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 12:33 |
| *** trown|outttypeww is now known as trown | 12:37 | |
| *** uggla__ has quit IRC | 12:38 | |
| *** uggla has joined #openstack-ironic | 12:46 | |
| *** jjohnson2 has joined #openstack-ironic | 12:49 | |
| *** dlpartain has joined #openstack-ironic | 12:53 | |
| *** dlpartain has left #openstack-ironic | 12:54 | |
| *** sambetts has quit IRC | 12:54 | |
| *** sambetts has joined #openstack-ironic | 12:55 | |
| TheJulia | wow it is quiet in ehre today | 12:59 |
| TheJulia | in here | 12:59 |
| *** lucas-afk is now known as lucasagomes | 13:01 | |
| lucasagomes | and I was set as afk still heh | 13:01 |
| lucasagomes | TheJulia, morning | 13:01 |
| TheJulia | good morning | 13:13 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Wake-On-Lan Power interface https://review.openstack.org/179078 | 13:13 |
| *** rameshg87 has joined #openstack-ironic | 13:15 | |
| NobodyCam | good morningn Ironicers | 13:16 |
| lucasagomes | NobodyCam, good morning | 13:16 |
| NobodyCam | hey hey lucasagomes :) how goes the day? | 13:17 |
| *** kkoski has joined #openstack-ironic | 13:17 | |
| lucasagomes | NobodyCam, hey there, good... gearing up for vancouver | 13:18 |
| rameshg87 | morning NobodyCam | 13:18 |
| NobodyCam | oh ya! | 13:18 |
| NobodyCam | morning rameshg87 :) | 13:18 |
| *** yog__ has quit IRC | 13:20 | |
| *** rloo has joined #openstack-ironic | 13:24 | |
| rameshg87 | lucasagomes: some questions on https://review.openstack.org/#/c/182387/2/ironic/common/pxe_utils.py | 13:24 |
| rameshg87 | oh 2 mins | 13:24 |
| rameshg87 | brb | 13:24 |
| *** rameshg871 has joined #openstack-ironic | 13:26 | |
| *** rameshg87 has quit IRC | 13:26 | |
| lucasagomes | rameshg871, yup what about it? | 13:27 |
| *** kkoski has joined #openstack-ironic | 13:30 | |
| *** rameshg871 is now known as rameshg87 | 13:31 | |
| rameshg87 | lucasagomes: sorry | 13:31 |
| rameshg87 | lucasagomes: had to leave suddenly, back now | 13:32 |
| rameshg87 | lucasagomes: so people who just upgrade to this code have to change neutron dnsmasq.conf ? | 13:32 |
| rameshg87 | lucasagomes: because neutron dnsmasq won't know about ipxe tag, right ? | 13:32 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 13:34 |
| *** kkoski has quit IRC | 13:35 | |
| lucasagomes | rameshg87, they already had to do it even without that code | 13:35 |
| lucasagomes | rameshg87, the tag was used when chainloading the undionly.kpxe | 13:35 |
| *** cdearborn has joined #openstack-ironic | 13:35 | |
| lucasagomes | opt_name': 'tag:!ipxe,bootfile-name | 13:36 |
| lucasagomes | was already there | 13:36 |
| rameshg87 | lucasagomes: yeah, so now they will add the ipxe tag, right ? | 13:36 |
| lucasagomes | rameshg87, yeah, but not now I mean this tag have been used before that code too | 13:37 |
| lucasagomes | (when using neutron) | 13:37 |
| lucasagomes | see L261 | 13:37 |
| lucasagomes | it's already using the tag | 13:37 |
| rameshg87 | lucasagomes: oh okay | 13:38 |
| rameshg87 | lucasagomes: so it was just that it was not documented earlier ? | 13:38 |
| * TheJulia /win 7 | 13:38 | |
| TheJulia | doh | 13:38 |
| lucasagomes | rameshg87, yup it's in our install-docs | 13:38 |
| rameshg87 | lucasagomes: okay | 13:38 |
| lucasagomes | rameshg87, http://docs.openstack.org/developer/ironic/deploy/install-guide.html#neutron-configuration | 13:38 |
| rameshg87 | got it | 13:39 |
| rameshg87 | lucasagomes: and since we are removing use of 175 for neutron | 13:39 |
| lucasagomes | rameshg87, I have a patch in neutron to create this tag by default without the need of a custom dnsmasq conf | 13:39 |
| lucasagomes | but there's some discussions going on there | 13:39 |
| rameshg87 | lucasagomes: so the alternative mentioned in L857 in https://review.openstack.org/#/c/182387/2/doc/source/deploy/install-guide.rst is not valid, right ? | 13:39 |
| lucasagomes | rameshg87, it's still valid. I just got problem with 1 machine that for some reason seems to be sending 175 even for non ipxe | 13:40 |
| lucasagomes | but since matching the user class worked for all machines I tested | 13:40 |
| lucasagomes | I left that as default | 13:40 |
| rameshg87 | lucasagomes: but how would neutron dnsmasq know what to do when it gets 175 ? | 13:40 |
| lucasagomes | rameshg87, it does nothing by default | 13:41 |
| lucasagomes | maybe I didn't get the question right | 13:41 |
| rameshg87 | ah let me put it this way | 13:42 |
| lucasagomes | rameshg87, why we are removing the use of 175? | 13:42 |
| lucasagomes | I mean it's documented how to create the ipxe tag in neutron in 2 ways | 13:42 |
| rameshg87 | yeah | 13:42 |
| lucasagomes | either use the userclass to determine the request is coming from iPXE or use the DHCP option 175 | 13:42 |
| lucasagomes | that is set by gPXE/iPXE | 13:42 |
| rameshg87 | yeah so if no action is to be taken for neutron for 175 | 13:43 |
| rameshg87 | why do we still need 175 to be mention in L858 - https://review.openstack.org/#/c/182387/2/doc/source/deploy/install-guide.rst | 13:44 |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: Update iLO documentation for UEFI secure boot https://review.openstack.org/178582 | 13:44 |
| rameshg87 | that part i still didn't get | 13:44 |
| lucasagomes | oh you mean why I left it commented? | 13:44 |
| rameshg87 | ah no. even if someone wanted to use that as alternative | 13:44 |
| rameshg87 | we don't care anymore about 175 for neutron dnsmasq, right ? | 13:45 |
| lucasagomes | if they want to use that, they uncomment that line and comment the other one | 13:45 |
| lucasagomes | rameshg87, we don't know about 175 | 13:45 |
| lucasagomes | we care if there's a tag called "ipxe" in dnsmasq | 13:45 |
| lucasagomes | if it's created either by looking at the dhcp option 175 | 13:46 |
| lucasagomes | or userclass | 13:46 |
| lucasagomes | ironic don't care about it | 13:46 |
| rameshg87 | oh may be i am missing this part | 13:46 |
| lucasagomes | it's up to the deployer | 13:46 |
| rameshg87 | dhcp_opts.append({'opt_name': 'bootfile-name', 'opt_value': ipxe_script_url}) | 13:46 |
| *** BadCub-Hiding is now known as BadCub | 13:46 | |
| BadCub | mornin folks | 13:46 |
| rameshg87 | does it refer to 175 ? | 13:46 |
| rameshg87 | morning BadCub | 13:46 |
| lucasagomes | rameshg87, it's just to keep as it was before for non-neutorn dhcp_provider | 13:46 |
| BadCub | heya rameshg87 :) | 13:47 |
| * rameshg87 still thinks | 13:47 | |
| *** edcragg has quit IRC | 13:47 | |
| rameshg87 | lucasagomes: let me go through ipxe documentation once more. may be i am still missing somethings there. | 13:48 |
| lucasagomes | rameshg87, ack | 13:48 |
| *** links has quit IRC | 13:48 | |
| *** lazy_prince has joined #openstack-ironic | 13:51 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 13:54 |
| NobodyCam | morning BadCub | 14:01 |
| BadCub | Morning NobodyCam :-) | 14:01 |
| rloo | morning NobodyCam, BadCub, TheJulia. Hi lucasagomes, rameshg87 | 14:02 |
| * BadCub has to reboot for updates of evil | 14:02 | |
| lucasagomes | rloo, morning | 14:02 |
| pshige | NobodyCam, BadCub: Morning | 14:03 |
| lucasagomes | rameshg87, the 175 you're talking about is the one at L268? | 14:06 |
| *** kkoski has joined #openstack-ironic | 14:06 | |
| jlvillal | rloo, Looks like they abandoned https://review.openstack.org/#/c/182413/ | 14:07 |
| *** mtanino has joined #openstack-ironic | 14:07 | |
| *** zz_natorious is now known as natorious | 14:07 | |
| rloo | jlvillal: sigh. guess i need to 'undepend' from that. | 14:07 |
| jlvillal | rloo, Yeah. Sorry :( | 14:08 |
| rloo | jlvillal: thanks for noticing! | 14:08 |
| rloo | jlvillal: no worries. I'm not (worried) :) | 14:08 |
| jlvillal | rloo, No problem! | 14:08 |
| NobodyCam | morning rloo, pshige & jlvillal | 14:08 |
| NobodyCam | :) | 14:08 |
| *** kkoski has quit IRC | 14:08 | |
| jlvillal | NobodyCam, Good morning | 14:08 |
| NobodyCam | :) | 14:09 |
| jlvillal | NobodyCam: mrda and I started working on the liaison work. https://wiki.openstack.org/wiki/Nova-Ironic and https://wiki.openstack.org/wiki/Nova-Ironic-Bugs | 14:09 |
| rameshg87 | morning rloo | 14:09 |
| jlvillal | NobodyCam, If you have any feedback, that would be great. | 14:10 |
| nagyz | lucasagomes, which workaround? but yeah now with pxe_ipmitool it works | 14:10 |
| rameshg87 | lucasagomes: ah no. | 14:10 |
| nagyz | ah I saw your review | 14:10 |
| lucasagomes | nagyz, right, but no ipxe right? | 14:10 |
| rameshg87 | lucasagomes: at line 268, it's about non-neutron dhcp, right | 14:11 |
| lucasagomes | rameshg87, yup | 14:11 |
| BadCub | mornin rloo :) | 14:11 |
| rameshg87 | lucasagomes: i am still wondering about 175 for neutron dhcp (why is it in documentation when we no longer care for it) | 14:11 |
| lucasagomes | rameshg87, right, but I don't get why we don't care about it | 14:11 |
| lucasagomes | it's just to inform that there's more than 1 way of creating the "ipxe" tag in neutron | 14:12 |
| nagyz | lucasagomes, no ipxe | 14:12 |
| lucasagomes | and deployers can use the one that works for them | 14:12 |
| NobodyCam | jlvillal: awesome I will take a look in a bit... have a call in 15 I attempting to be ready for :) | 14:12 |
| jlvillal | NobodyCam, Thanks | 14:12 |
| rameshg87 | lucasagomes: but how will dhcp know to hand over ipxe_script_url | 14:12 |
| rameshg87 | lucasagomes: i mean how will neutron dhcp know to hand over ipxe_script_url if some machine still requests 175 | 14:13 |
| lucasagomes | nagyz, you want to enable it? | 14:13 |
| lucasagomes | rameshg87, right say the tag was creating based on the userclass | 14:13 |
| lucasagomes | rameshg87, machine request 175 but userclass is not iPXE | 14:14 |
| lucasagomes | then neutron will ACK it with undionly.kpxe and the machine will chainload it | 14:14 |
| nagyz | lucasagomes, it works now tbh:) | 14:14 |
| nagyz | I'll play with other parts and then we can come back to making it work with ipxe | 14:14 |
| lucasagomes | rameshg87, now the machine will request with userclass iPXE | 14:14 |
| lucasagomes | and DHCP will give it the boot script | 14:14 |
| rameshg87 | same 175 again second time ? | 14:14 |
| lucasagomes | rameshg87, 175 is set for both gPXE and iPXE, both are capable of downloading via HTTP | 14:15 |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Do not preserve partitions after ephemeral https://review.openstack.org/182688 | 14:15 |
| lucasagomes | rameshg87, if the image ipxe image it chainloaded is not returning userclass iPXE there's something wrong with the image | 14:15 |
| lucasagomes | it will keep chainloading it in loop | 14:15 |
| lucasagomes | nagyz, ack | 14:16 |
| lucasagomes | rameshg87, I don't get why we care about how the ipxe tag is set | 14:17 |
| rloo | jlvillal: nice wiki. I am thinking (not sure) that maybe we should put nova-ironic-bugs as a 'subteam' report in the weekly meetings. | 14:17 |
| *** kan has quit IRC | 14:17 | |
| lucasagomes | we just need to document how it can be set, it's dhcp configuration it's out of ironic's hand | 14:18 |
| jlvillal | rloo, That seems like a valid idea to me. | 14:18 |
| rameshg87 | lucasagomes: let me reiterate if i got it right | 14:18 |
| NobodyCam | rloo: ++ | 14:18 |
| rameshg87 | lucasagomes: machine boots up. it starts requesting over pxe with code 175 (but not with tag ipxe) | 14:18 |
| rloo | jlvillal: so py 3.4 is enabled now. (it shows merged). yay. | 14:18 |
| rameshg87 | lucasagomes: neutron dhcp hands over undionly.kpxe | 14:18 |
| rameshg87 | lucasagomes: right so far ? | 14:18 |
| rloo | jlvillal: maybe worth sending email about that (py 3.4)? | 14:19 |
| NobodyCam | also a wiki may be overkill for the -Bugs | 14:19 |
| NobodyCam | but I love the info on it :) | 14:19 |
| lucasagomes | rameshg87, if the tag was created with dhcp-match=set:ipxe,175 no. The dhcp request will now ACK with the boot script | 14:19 |
| jlvillal | rloo, Okay. To openstack-dev ? | 14:19 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 14:20 |
| rloo | NobodyCam, jlvillal: yeah, i was thinking maybe etherpad. But I'm fine with either or none. the problem being 'how to keep it current'. but if jlvillal and mrda want to do it that way, I'm fine :) | 14:20 |
| rloo | jlvillal: yeah, openstack-dev [Ironic]. So developers know. | 14:20 |
| lucasagomes | rameshg87, because "!ipxe" is a negation "!" means if the request doesn't match that tag | 14:20 |
| jlvillal | rloo, Okay. | 14:20 |
| lucasagomes | rameshg87, if the tag is set using the userclass, and the request only had 175 set. Yes we will hand over the undionly.kpxe | 14:20 |
| jlvillal | NobodyCam, rloo We went with the wiki to keep track of the bugs. mrda and I will see how it works. We can always switch to Etherpad if desired. | 14:21 |
| jlvillal | NobodyCam, rloo: mrda have scheduled a weekly meeting on Tuesdays to do a bug scrub. | 14:21 |
| jlvillal | s/mrda/mrda and I/ | 14:21 |
| NobodyCam | jlvillal: its got great info and think will be quite helpful | 14:22 |
| NobodyCam | :) | 14:22 |
| jlvillal | rloo, I will send out an email when I get to work in a bit. | 14:22 |
| rloo | jlvillal: part of that wiki (the first part I think) is useful in a wiki. the stuff that doesn't change much :-) For the stuff that could change (status of bugs etc), maybe mention when it was last updated? | 14:22 |
| rloo | jlvillal: oh yeah, I thought it was a bit early for you :) | 14:22 |
| rameshg87 | lucasagomes: okay, getting something now. let me read through. | 14:22 |
| jlvillal | rloo, Okay. I did try to add a date next to each entry. But I will also add a page-updated date. | 14:22 |
| lucasagomes | rameshg87, http://etherboot.org/wiki/pxechaining#breaking_the_infinite_loop | 14:22 |
| lucasagomes | rameshg87, that's a good guide | 14:23 |
| lucasagomes | that's for gPXE but works the same for iPXE | 14:23 |
| jlvillal | rloo, NobodyCam: Thanks for the feedback! | 14:23 |
| * rameshg87 goes through it | 14:23 | |
| NobodyCam | jlvillal: thank you !!! | 14:23 |
| * jlvillal gets ready to go to work. | 14:24 | |
| *** chlong has quit IRC | 14:30 | |
| *** stendulker has joined #openstack-ironic | 14:32 | |
| stendulker | rloo: Hi | 14:32 |
| rloo | hi stendulker | 14:32 |
| stendulker | rloo : Regarding the ilo documentation patch review https://review.openstack.org/#/c/178582/ | 14:33 |
| rloo | stendulker: ? | 14:33 |
| stendulker | rloo: I did not get your comment completely. Are you saying driver documentation should move out of source tree | 14:33 |
| rloo | stendulker: any documentation that isn't related to developing, should move out of ironic source tree and into openstack manuals. | 14:34 |
| *** verdurin has joined #openstack-ironic | 14:34 | |
| rloo | stendulker: somewhere accessible from http://docs.openstack.org/ | 14:35 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 14:35 |
| stendulker | rloo: Ok. So do you suggest to make it a wiki link in the documetation or can we do that when all other non-developer documentation moves into openstack manuals | 14:35 |
| rloo | stendulker: i am just wondering right now, why you want to try to maintain two versions/copies of the same documentation. | 14:35 |
| rloo | stendulker: and also for selfish reasons, if you keep your doc in that wiki, it means fewer patches that I will review. bonus for me! :) | 14:36 |
| stendulker | rloo: I know its painful. But we do follow it so that, user do not have to refer to two places... | 14:36 |
| rloo | stendulker: but they do anyway. I mean, they have to click on the ilo link. does it matter where they land? | 14:36 |
| stendulker | rloo: But that is big disadvantage for us. We will miss close scrutiny of what we write... :( | 14:37 |
| rloo | stendulker: anyway, I didn't -1 on your patch. it was a suggestion. Cuz it seems inefficient. | 14:37 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Add iRMC Virtual Media Deploy module for iRMC Driver https://review.openstack.org/151958 | 14:37 |
| stendulker | rloo: Your coments are very helpful and it makes things better :) | 14:38 |
| rloo | stendulker: do what you want. it was a suggestion. | 14:38 |
| * rloo is going to stop reviewing documentation. Only look at code!!! :) | 14:38 | |
| stendulker | rloo: I think may be we can keep it as it is for now. I will check with the other iLO developers and we can take that call separately. | 14:39 |
| stendulker | rloo: Just wanted to understand your comment and possible options. | 14:39 |
| rloo | stendulker: ok | 14:39 |
| naohirot | lucasagomes: hello | 14:39 |
| stendulker | rloo: Thank you :) | 14:40 |
| stendulker | rloo: Can you please have a look at this spec when you get time. UEFI secure boot support for pxe_ilo driver https://review.openstack.org/#/c/174295/ | 14:41 |
| rloo | stendulker: there are a lot of specs and I am purposely not focusing/reviewing those. I don't have the bandwidth to review the specs and the non-spec patches. | 14:42 |
| lucasagomes | naohirot, hi there | 14:42 |
| naohirot | lucasagomes: hi | 14:42 |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Do not preserve partitions after ephemeral https://review.openstack.org/182688 | 14:42 |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 14:42 |
| naohirot | lucasagomes: I'd like to get your reply regarding https://review.openstack.org/#/c/177632/3/specs/kilo-archive/ironic-generic-raid-interface.rst | 14:43 |
| stendulker | rloo: On a side note, I have also raised the code review for that spec. https://review.openstack.org/#/c/166192/ :) | 14:43 |
| rloo | stendulker: oh. did you want me to -2 it until the spec is approved? | 14:44 |
| stendulker | rloo: Oh no... | 14:44 |
| rloo | stendulker: guess I shouldn't have asked you. that's standard practise. Sorry. | 14:45 |
| lucasagomes | naohirot, will take a look :D | 14:45 |
| naohirot | lucasagomes: okay, thanks | 14:45 |
| naohirot | rameshg87: hello | 14:45 |
| rameshg87 | naohirot: hey | 14:46 |
| naohirot | rameshg87: I've just submitted the update of instance_info, https://review.openstack.org/#/c/151958/17..18/ironic/drivers/modules/irmc/deploy.py | 14:46 |
| rameshg87 | naohirot: oh okay. sure, will have a look. | 14:47 |
| naohirot | rameshg87: you will see why driver_info[irmc_boot_iso] was necessary | 14:47 |
| lucasagomes | naohirot, oh the rst probably was moved to kilo-archieve but the html still there in the specs.openstack.org | 14:48 |
| naohirot | rameshg87: because driver_internal_info[irmc_boot_iso] is necessary instead. | 14:48 |
| lucasagomes | naohirot, so yeah we shouldn't use that... So currently zapping is not merged yet | 14:48 |
| *** uggla has quit IRC | 14:49 | |
| lucasagomes | naohirot, both links seems wrong then, becasue it was merged in Kilo but wasn't implemented so now it's moved to Liberty | 14:49 |
| lucasagomes | but still not merged | 14:49 |
| rameshg87 | naohirot: okay. makes sense if you don't want to use instance_info for both the types of boot isos | 14:49 |
| rameshg87 | naohirot: will have a look | 14:50 |
| BadCub | okay... SO everything I am able to reserve at this point is at minimum 20 minutes from the convention centre... Thoughts???? | 14:50 |
| rameshg87 | BadCub: 20 mins by ..... ? | 14:50 |
| lucasagomes | naohirot, a-ha! well that spec for ironic-generic-raid-interface is also part of the kilo-archive | 14:50 |
| BadCub | by car | 14:50 |
| lucasagomes | perhaps we should just leave it as it? cause it's going to be deleted no? | 14:50 |
| *** e0ne is now known as e0ne_ | 14:50 | |
| rameshg87 | BadCub: oh and a car is being arranged too ? ;-) | 14:51 |
| rameshg87 | BadCub: this is for monday night ironic dinner, right ? | 14:51 |
| BadCub | Yes, for the MOnday dinner | 14:51 |
| *** uggla has joined #openstack-ironic | 14:51 | |
| pshige | Oh by car! | 14:51 |
| BadCub | or taxi | 14:52 |
| *** e0ne_ is now known as e0ne | 14:52 | |
| naohirot | lucasagomes: so would you give me the concrete URL to avoid misunderstanding? | 14:52 |
| rameshg87 | heh | 14:52 |
| lucasagomes | naohirot, I'm unsure... I mean since that spec is also under kilo-archieve why just not use the kilo-archive URL as before? | 14:52 |
| *** hemna_ is now known as hemna | 14:53 | |
| rameshg87 | lucasagomes: back to bug you again. probably a last question if i get it right now | 14:53 |
| rameshg87 | lucasagomes: went through etherboot doc | 14:53 |
| rameshg87 | lucasagomes: i understood chainloading and tagging to avoid that | 14:53 |
| lucasagomes | rameshg87, hey there | 14:53 |
| rameshg87 | lucasagomes: https://review.openstack.org/#/c/182387/2/ironic/common/pxe_utils.py | 14:53 |
| lucasagomes | rameshg87, right | 14:54 |
| BadCub | is everyone okay with a short taxi ride? or should I just kill the idea at this point and see if we can wander about and find something impromtu | 14:54 |
| rameshg87 | lucasagomes: why dhcp_opts.append({'opt_name': 'tag:ipxe,bootfile-name', 'opt_value': ipxe_script_url}) WORKS and dhcp_opts.append({'opt_name': 'bootfile-name', 'opt_value': ipxe_script_url}) DOESN'T WORK | 14:54 |
| rameshg87 | lucasagomes: anything which is !ipxe, is "ipxe", right ? | 14:54 |
| rameshg87 | why do we have to mention it explicitly | 14:55 |
| rameshg87 | or rather how does it work when we mention it explicitly | 14:55 |
| lucasagomes | rameshg87, anything that which is !ipxe can be standard PXE or gPXE | 14:55 |
| lucasagomes | or something else idk | 14:55 |
| *** uggla_ has joined #openstack-ironic | 14:55 | |
| lucasagomes | rameshg87, with tag:ipxe,bootfile-name... I make sure the request coming is iPXE | 14:55 |
| lucasagomes | "!ipxe" just means it's not ipxe | 14:56 |
| naohirot | lucasagomes: As I commented, Dmitry prefers to change it to https://review.openstack.org/#/c/140826/, is this problem ? | 14:56 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: When boot option is not persisted, set boot on next power on https://review.openstack.org/177642 | 14:57 |
| *** killer_prince has joined #openstack-ironic | 14:57 | |
| lucasagomes | naohirot, yeah I saw his comment... I mean, all those specs under kilo-achieve will be deleted after they are merged in L no? | 14:57 |
| lucasagomes | naohirot, I'm a bit confused to be honest... | 14:57 |
| lucasagomes | IMO I would just live the specs under kilo-archive as is... I wouldn't update them | 14:58 |
| lucasagomes | because they are temporary stuff | 14:58 |
| lucasagomes | AFAIUI | 14:58 |
| lucasagomes | rameshg87, got it? | 14:59 |
| lucasagomes | rameshg87, like being "!ipxe" doesn't means it's ipxe | 14:59 |
| rameshg87 | oh :) | 14:59 |
| naohirot | lucasagomes: I prefer to move foreword by taking Dmitry's idea, and then later we can fix it if it settled in the final place. | 14:59 |
| rameshg87 | if that's the case got it | 14:59 |
| rloo | BadCub: I doubt that wandering around will work with a big crowd/group. | 14:59 |
| lucasagomes | rameshg87, cause !ipxe can also be gPXE | 14:59 |
| *** Marga_ has joined #openstack-ironic | 14:59 | |
| lucasagomes | or starndard PXE | 15:00 |
| *** uggla has quit IRC | 15:00 | |
| rloo | BadCub: so I take it all those 5 choices you mentioned are out. | 15:00 |
| rloo | BadCub: any way to cater/bring food in? | 15:00 |
| lucasagomes | naohirot, right... ok no problem with that | 15:00 |
| lucasagomes | naohirot, I will change my vote | 15:00 |
| naohirot | lucasagomes: great thanks :) | 15:00 |
| *** lazy_prince has quit IRC | 15:00 | |
| rloo | BadCub: but a taxi ride shouldn't be too bad. is 20 min accounting for the traffic there? | 15:01 |
| *** kkoski has joined #openstack-ironic | 15:02 | |
| rameshg87 | lucasagomes: in case of !ipxe, we always handover undionly.kpxe, right ? | 15:02 |
| BadCub | rloo: some of them are no longer available. I am having to redo the search now | 15:02 |
| rameshg87 | i mean even for gPXE | 15:02 |
| *** killer_prince has quit IRC | 15:02 | |
| rloo | BadCub: if you find something, just grab it! | 15:02 |
| BadCub | if everyone is good with 20 minute travel, I will book something that works | 15:02 |
| *** Sukhdev has joined #openstack-ironic | 15:02 | |
| *** e0ne is now known as e0ne_ | 15:02 | |
| lucasagomes | rameshg87, yup, because then we hand over one iPXE iamge | 15:03 |
| *** Marga_ has quit IRC | 15:03 | |
| lucasagomes | rameshg87, and the second request will come from iPXE | 15:03 |
| NobodyCam | 20 minutes is not that bad. I'm in! | 15:03 |
| rameshg87 | yeah i got it | 15:03 |
| lucasagomes | and then we will match the tag ipxe, and return the boot script | 15:03 |
| lucasagomes | rameshg87, I know it's a bit confusing to get the head around it | 15:03 |
| lucasagomes | :-( | 15:03 |
| *** Marga_ has joined #openstack-ironic | 15:03 | |
| lucasagomes | but nothing we can do really, it's the way it works | 15:03 |
| rameshg87 | lucasagomes: i will definitely try ipxe once and learn it | 15:03 |
| rloo | BadCub: I think walking is preferable to taxi, but do what you can. Oh, if this is for Monday, traffic won't be so bad because it is a holiday. | 15:04 |
| lucasagomes | rameshg87, cool :-) | 15:04 |
| rameshg87 | lucasagomes: thanks for taking pains for explanation :) | 15:04 |
| *** kkoski has quit IRC | 15:04 | |
| lucasagomes | rameshg87, any time! no problem | 15:04 |
| BadCub | rloo: indeed. I will go ahead and book us in [somewhere] :) | 15:04 |
| naohirot | rameshg87: The point of my change is that I switched driver_info['irmc_boot_iso'] to instance_info['irmc_boot_iso'], however I needed to introduce driver_internal_info['irmc_boot_iso']. | 15:05 |
| rameshg87 | naohirot: yeah but i think use-case was same. we could have lived with instance_info only | 15:05 |
| rameshg87 | naohirot: in the end there is only one irmc_boot_iso for the instance, right ? | 15:06 |
| naohirot | rameshg87: But I don't know how. | 15:06 |
| rameshg87 | naohirot: need to go now. if you aren't around when i come back, i will surely go through and comment. | 15:06 |
| *** rameshg87 is now known as rameshg87-away | 15:06 | |
| naohirot | rameshg87-away: I have to go to bed too, please make comment on the gerrit, I'll check it tomorrow. | 15:07 |
| naohirot | good night all | 15:07 |
| *** naohirot has quit IRC | 15:08 | |
| pshige | naohirot: night! | 15:09 |
| *** kan has joined #openstack-ironic | 15:14 | |
| *** jistr is now known as jistr|mtg | 15:19 | |
| nagyz | rameshg87-away, you're not the only one not understanding ipxe... | 15:22 |
| nagyz | ;-) | 15:22 |
| *** Nisha has joined #openstack-ironic | 15:23 | |
| *** stickpin has left #openstack-ironic | 15:23 | |
| *** kkoski has joined #openstack-ironic | 15:24 | |
| *** kkoski has quit IRC | 15:26 | |
| *** jistr|mtg is now known as jistr | 15:29 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic-specs: exposing driver capabilities https://review.openstack.org/182601 | 15:30 |
| jlvillal | BadCub: 20 minutes is fine for me. | 15:30 |
| BadCub | I'm workin on it. *groans* | 15:32 |
| nagyz | lucasagomes, once a node is stuck in waiting for callback and I issue a nova delete that also is stuck forever... | 15:33 |
| lucasagomes | nagyz, hmm try to ironic node-set-provision-state <node uuid> deleted | 15:34 |
| lucasagomes | nagyz, ironic shoudl support stopping an deployment in WAIT call back | 15:35 |
| lucasagomes | sounds like nova is not sending that command to ironic? | 15:35 |
| * lucasagomes tries that out | 15:35 | |
| *** rameshg87-away is now known as rameshg87 | 15:35 | |
| rameshg87 | nagyz: :) | 15:36 |
| lucasagomes | rameshg87, you can test ipxe in devstack | 15:36 |
| NobodyCam | lol: should we land https://review.openstack.org/#/c/173218 or wait until summit? there is a item on the Friday meetup to go over specs, I added raid to it. but it has four plus two's | 15:37 |
| lucasagomes | rameshg87, apply this to devstack https://review.openstack.org/#/c/182595/ and that patch in ironic under review to ironic | 15:37 |
| lucasagomes | rameshg87, then just enable add IRONIC_IPXE_ENABLED=True to the local.conf | 15:37 |
| lucasagomes | and devstack will cofigure an ipxe eviroment for u | 15:37 |
| lucasagomes | so you can poke with it | 15:37 |
| rloo | NobodyCam: what's there to discuss. Land it! | 15:38 |
| NobodyCam | rloo: there are two Raid specs up | 15:38 |
| rameshg87 | lucasagomes: sure. iwill try it out soon | 15:38 |
| rloo | NobodyCam: but I don't think any work can be done on that until another spec is approved. | 15:38 |
| rameshg87 | NobodyCam: hold on please | 15:38 |
| rameshg87 | NobodyCam: there is parent spec which as got to merge | 15:38 |
| rameshg87 | NobodyCam: https://review.openstack.org/#/c/173214/ | 15:38 |
| nagyz | lucasagomes, right, deleted works | 15:38 |
| lucasagomes | rameshg87, cool... or follow the install guide if u want to configure for ur current enviroment | 15:39 |
| lucasagomes | rameshg87, and test on baremetla | 15:39 |
| lucasagomes | nagyz, ack, I will take a look in nova | 15:39 |
| rameshg87 | lucasagomes: yeah will test on bare metal | 15:39 |
| rloo | rameshg87, NobodyCam: this spec can still be landed? Or should it have a depends on thing, on the other spec? | 15:39 |
| lucasagomes | see why it's not sending the delete to ironic | 15:39 |
| lucasagomes | rameshg87, ack :-) | 15:39 |
| nagyz | lucasagomes, appreciated | 15:39 |
| rameshg87 | rloo: yeah it has dependency | 15:39 |
| NobodyCam | rameshg87: ya thats why I added it to the Friday meetup agenda | 15:40 |
| rloo | rameshg87: I don't see the dependency in the commit message :-( | 15:40 |
| rameshg87 | i think when i raised review to move it from kilo-archive to liberty | 15:40 |
| rameshg87 | i forgot to put dependency | 15:40 |
| rameshg87 | may be i will put gerrit dependency | 15:40 |
| *** ukalifon has quit IRC | 15:40 | |
| rloo | rameshg87: maybe not bother doing that, cuz if you update the commit msg, it'll wipe out all the votes. | 15:40 |
| rameshg87 | rloo: :( | 15:41 |
| * rameshg87 has never got 4 x +2s in my life | 15:41 | |
| NobodyCam | rameshg87: maybe just set WIP with a comment as to why | 15:41 |
| rloo | rameshg87: we don't usually wait for 4 +2s before approving :-) | 15:41 |
| rloo | NobodyCam: i say just approve this. rameshg87 can update it if the parent one changes. | 15:42 |
| rameshg87 | rloo: hopefully i doesn't change | 15:42 |
| rameshg87 | :) | 15:42 |
| rameshg87 | NobodyCam: so W-1 or should i wait for +A ? | 15:42 |
| rameshg87 | NobodyCam: rloo: i think better to wait | 15:42 |
| rameshg87 | NobodyCam: rloo: i am not going to gain anything in landing that spec without it's parent | 15:43 |
| *** e0ne_ is now known as e0ne | 15:43 | |
| NobodyCam | rameshg87: I'm good with waiting... | 15:43 |
| rameshg87 | NobodyCam: rloo: i think i will wait | 15:43 |
| rameshg87 | i will do W-1 | 15:43 |
| NobodyCam | :) that way you wont lose the votes you have | 15:43 |
| rloo | rameshg87, NobodyCam: I personally don't care, whatever works for you both. I will try to remember not to read it so I don't hold it back ;) | 15:43 |
| * rameshg87 takes screenshot of review with 4 x +2s | 15:44 | |
| NobodyCam | lol :) ++++ | 15:44 |
| * rloo thinks rameshg87 will have a tshirt made ... | 15:44 | |
| rameshg87 | hehe | 15:45 |
| *** Nisha has quit IRC | 15:47 | |
| nagyz | lucasagomes, sometimes I see that I wanted to get an instance (from nova) it boots one ironic node and then later on after a while it switches to an other node. is that the built-in fail mechanism? | 15:48 |
| lucasagomes | nagyz, nova has a retry filter | 15:48 |
| lucasagomes | which is enabled by default, so i think it will retry 3 times | 15:48 |
| lucasagomes | and then it fails complete to boot the instance | 15:49 |
| lucasagomes | nagyz, look at the scheduler filters you can tweak it or disable it | 15:49 |
| *** shakamunyi has quit IRC | 15:51 | |
| *** barra204 has quit IRC | 15:51 | |
| NobodyCam | open question: looking at the Stable Branches section of https://etherpad.openstack.org/p/liberty-ironic-design-summit-ideas these look like they should go in / on the https://etherpad.openstack.org/p/liberty-ironic-functional-testing pad.. there is a item on fridays meetup agenda stable branches too: https://etherpad.openstack.org/p/liberty-ironic-contributor-meetup | 15:52 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Remove unused CONF variable in test_ipminative.py https://review.openstack.org/182735 | 15:52 |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Update iLO driver documentation related to Swift https://review.openstack.org/182736 | 15:53 |
| nagyz | so generally speaking what if I have 2x10gbit in the servers? can I add both nics to ironic? I guess ironic is really only used for the provisioning state; if I want LACP I need to have the image with the correct network options set. | 15:55 |
| rameshg87 | good night ironic | 15:57 |
| *** rameshg87 has quit IRC | 15:57 | |
| *** dprince has quit IRC | 15:58 | |
| JayF | nagyz: We actually run ironic like that | 16:03 |
| BadCub | Okay folks... Venue for dinner reserved... It is 10 minutes from convention centre: eddyspinkpeppercorn.com | 16:03 |
| BadCub | 1485 Kingsway | 16:03 |
| BadCub | Vancouver, BC, V5N 2R6 | 16:03 |
| rloo | NobodyCam: yes, I think the stable branches section fits with the testing. The question of stopping support for juno is separate I think. | 16:04 |
| JayF | nagyz: but none of the network switching pieces needed to have provisioning on a separate network than tenants is upstream/generic right now | 16:04 |
| JayF | nagyz: there's a spec you'd be interested in, and I have a patch to make cloud-init understand how to make LACP configs you might be interested in too :) | 16:04 |
| NobodyCam | rloo: ya I'm putting it on the testing pad | 16:04 |
| NobodyCam | morning to JayF too | 16:05 |
| rloo | NobodyCam: thanks for doing all this! | 16:05 |
| NobodyCam | :-p | 16:05 |
| JayF | good mornign | 16:05 |
| JayF | see you all in a few days :P | 16:05 |
| NobodyCam | oh ya | 16:06 |
| jlvillal | BadCub: Thanks for organizing the dinner! :) | 16:06 |
| BadCub | I have updated the pad with address, date, time. We have space for 20 ppl. :) | 16:06 |
| BadCub | jlvillal: yw :) | 16:06 |
| rloo | thx BadCub | 16:07 |
| BadCub | rloo: yw :) | 16:07 |
| BadCub | They have a pretty good menu too! Should be something for everyone :) | 16:07 |
| nagyz | JayF, sorry I'm not sure I fully understand the first sentence - there is no physical switch setup, right? in kilo | 16:08 |
| rloo | BadCub: you deserve a seat next to deva this time :-) | 16:08 |
| *** jcoufal has quit IRC | 16:08 | |
| *** kan has quit IRC | 16:08 | |
| BadCub | rloo: lol | 16:09 |
| rloo | BadCub: although I have the best conversations with you! | 16:10 |
| JayF | nagyz: Ironic doesn't support using one network for provisioning, and another network for tenants | 16:10 |
| BadCub | rloo: well thank ya kindly ma'am. May I reply in kind :) | 16:10 |
| JayF | nagyz: and given it's (afaik) impossible to PXE off LACP, that's a requirement unless you're running one of the virtual media drivers | 16:10 |
| nagyz | JayF right. you can PXE boot off LACP if you don't have active just passive LACP | 16:11 |
| rloo | BadCub: ;) | 16:11 |
| NobodyCam | we don't have a official slot for: Should we ditch pecan/WSME? should that be a friday meetup thing or do we thing a larger audience would be better to get a broader view? | 16:11 |
| devananda | JayF: that is something I'd like to change (separate networks). | 16:11 |
| nagyz | JayF, one option would be that for provisioning one network is provisioned for the node and then once that's successful the physical switch is reconfigured so the node is on a new network | 16:11 |
| nagyz | shouldn't be that hard if we leverage neutron | 16:11 |
| JayF | nagyz: that's what we do downstream in Rackspace OnMetal today; but that code isn't generic or upstream yet | 16:11 |
| devananda | nagyz: there's already proposals for that up :) | 16:11 |
| JayF | nagyz: I was looking for a spec I thought jroll was working on to add support for that; but I couldn't find it | 16:12 |
| NobodyCam | morning devananda :) | 16:12 |
| *** rwsu has joined #openstack-ironic | 16:13 | |
| rloo | JayF, nagyz: this spec: http://specs.openstack.org/openstack/ironic-specs/specs/kilo-archive/network-provider.html | 16:13 |
| nagyz | JayF, are you guys using neutron's ml2 driver to actually change the switch config and just changing the network via neutron calls for the port? | 16:13 |
| JayF | nagyz: I don't understand all the technical details of that bit of our implementation | 16:13 |
| devananda | nagyz, JayF: see https://etherpad.openstack.org/p/liberty-ironic-network-isolation | 16:14 |
| devananda | nagyz: in short, that is exactly what's proposed. | 16:14 |
| JayF | That's pretty awesome | 16:14 |
| devananda | but the details of cross-service interaction need to be worked out | 16:14 |
| devananda | does nova tell neutron? or does ironic tell neutron? | 16:14 |
| NobodyCam | just fyi: I'm still doing cleanup and what nots on the summit pads | 16:14 |
| nagyz | too bad I won't be able to go to the current summit | 16:15 |
| devananda | i think there's a limit to how many etherpads Firefox can have open at once ... and I think I crossed it a while ago, lol | 16:15 |
| BadCub | lol | 16:15 |
| NobodyCam | nagyz: if you have thoughts or concerns please add then to the 'pad so we know about them | 16:16 |
| BadCub | I found the limit on how many opentable and yelp windows one can have open this morning! | 16:16 |
| NobodyCam | devananda: safari is like 5 | 16:16 |
| NobodyCam | I now limit my self to 3 open at a time | 16:16 |
| devananda | i think i have 20 open ... | 16:16 |
| NobodyCam | ieek | 16:16 |
| * NobodyCam thinks his Mac book air would explode | 16:17 | |
| nagyz | thanks guys, have a great evening - bbl. | 16:18 |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Execute "parted" from root in list_partitions() https://review.openstack.org/182741 | 16:18 |
| NobodyCam | night nagyz | 16:18 |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Do not preserve partitions after ephemeral https://review.openstack.org/182688 | 16:21 |
| *** rloo has quit IRC | 16:22 | |
| *** jistr has quit IRC | 16:25 | |
| *** dprince has joined #openstack-ironic | 16:29 | |
| openstackgerrit | John L. Villalovos proposed openstack/python-ironicclient: Refactor resource_fields.py https://review.openstack.org/181985 | 16:29 |
| *** derekh_ has quit IRC | 16:37 | |
| *** kkoski has joined #openstack-ironic | 16:40 | |
| *** praneshp has joined #openstack-ironic | 16:41 | |
| *** ppal has quit IRC | 16:50 | |
| praneshp | Hello Ironic. Is there a feature that is similar to host aggregates in nova? Ideally, i should be able to group ironic nodes into groups, and based on some input while doing nova boot it should pick up a node from the right pool | 16:54 |
| praneshp | I understand that I can add arbitrary key value pairs to ironic nodes when creating them, and write a scheduler filter to read those. Was just wondering if there was an alternative that was similar to aggregates so I can avoid patching and stay close to community code | 16:55 |
| praneshp | jroll: devananda ^^ | 16:55 |
| JayF | Chassis? | 16:57 |
| *** Marga_ has quit IRC | 16:57 | |
| jlvillal | NobodyCam, BadCub: Maybe we can add to some agenda item a discussion about httpretty. Infra people don't seem to like that we keep black-listing every new release that comes out. Not sure what kind of effort it would take to stop using it in python-ironicclient. | 17:00 |
| praneshp | JayF: ok let me look into chassis. thanks for the tip | 17:01 |
| *** jcoufal has joined #openstack-ironic | 17:01 | |
| BadCub | jlvillal: That is probably a wise idea. devananda NobodyCam ^^ thoughts? | 17:01 |
| *** Sukhdev has quit IRC | 17:02 | |
| NobodyCam | humm | 17:02 |
| *** jcoufal has quit IRC | 17:02 | |
| NobodyCam | sounds like a friday thing! | 17:03 |
| * NobodyCam adds | 17:03 | |
| *** Sukhdev has joined #openstack-ironic | 17:03 | |
| BadCub | NobodyCam: thank you :) | 17:03 |
| *** ijw has joined #openstack-ironic | 17:04 | |
| praneshp | JayF: looks like there is a chassis level discovery tool in progress: https://review.openstack.org/#/c/158577/ | 17:04 |
| praneshp | thanks for the tip though JayF | 17:04 |
| *** davideag_ has joined #openstack-ironic | 17:04 | |
| jlvillal | BadCub: NobodyCam: clarkb suggests we just lock down the version to the working version. And when httpretty finally fixes the bug we allow new versions. Which makes sense to me. | 17:04 |
| jlvillal | But maybe we also consider alternative libraries? | 17:05 |
| NobodyCam | added.. Thank you jlvillal | 17:05 |
| jlvillal | NobodyCam: Thanks | 17:05 |
| jlvillal | NobodyCam: FYI: https://review.openstack.org/#/c/182340/ is the patch | 17:05 |
| *** gridinv_ has joined #openstack-ironic | 17:06 | |
| *** davideagnello has quit IRC | 17:07 | |
| *** gridinv_ has quit IRC | 17:11 | |
| *** davideagnello has joined #openstack-ironic | 17:12 | |
| *** achanda has joined #openstack-ironic | 17:14 | |
| *** davideag_ has quit IRC | 17:15 | |
| NobodyCam | jlvillal: I tagged you as person who knows about this topic. :) | 17:16 |
| *** pelix has quit IRC | 17:16 | |
| *** pas-ha has quit IRC | 17:17 | |
| NobodyCam | I see cinder has a session about integration, any reason for us to talk after there session on wed? | 17:18 |
| *** harlowja has quit IRC | 17:20 | |
| *** harlowja has joined #openstack-ironic | 17:20 | |
| *** rloo has joined #openstack-ironic | 17:23 | |
| openstackgerrit | Merged stackforge/pyghmi: Flesh out and rework some sensor descriptions https://review.openstack.org/182074 | 17:24 |
| lucasagomes | folks I will call it a day | 17:25 |
| lucasagomes | have a good night | 17:25 |
| *** lucasagomes is now known as lucas-dinner | 17:25 | |
| NobodyCam | night lucas-dinner | 17:26 |
| BadCub | g'night lucas-dinner | 17:27 |
| rloo | night lucas-dinner | 17:28 |
| NobodyCam | I'm going to add a "other project session impact discussion" item to Fridays meetup .. just in case! | 17:28 |
| rloo | NobodyCam: not sure I understand, what is that 'other project session impact discussion' meant for? | 17:30 |
| rloo | jlvillal: wrt the httpretty thing. I don't know that it is worth discussing at the summit. unless there are some proposals and we are discussing which to pick. | 17:34 |
| *** 17SACKCIT has joined #openstack-ironic | 17:34 | |
| *** 7YUAAALBR has joined #openstack-ironic | 17:34 | |
| *** 7YUAAALBR has quit IRC | 17:34 | |
| *** 17SACKCIT has quit IRC | 17:34 | |
| jlvillal | rloo: Okay. I don't think it has to be a summit thing. It could be a weekly meeting agenda item. | 17:34 |
| rloo | jlvillal: if the only reason is cuz of this particular bug, i don' tknow that it is worth discussing? | 17:35 |
| jlvillal | rloo: To me what is worth discussing is do we want to keep using httpretty or replace it? | 17:35 |
| jlvillal | rloo: Based on the issues that have come up. | 17:35 |
| rloo | jlvillal: that could be a discussion we have with every package we are using. why httppretty? there is only this one issue or are there more? | 17:36 |
| jlvillal | rloo: It has happened three releases in a row from httpretty where it doesn't work. And then breaking the gate. | 17:36 |
| *** uggla_ has quit IRC | 17:36 | |
| rloo | jlvillal: the same bug in httpretty hasn't been fixed. we could have blocked it the first time. or we could have patched/fixed our client the first time. | 17:37 |
| rloo | jlvillal: i am not convinced that a bug in py3.4.2 that causes this issue with httpretty is a reason for switching packages. | 17:37 |
| jlvillal | rloo: So that is the reason I thought it could be discussed. At the moment httpretty is only used in test_shell.py. | 17:38 |
| jlvillal | rloo: I am fine if you feel we should just keep it as is. But I think we need to lock it down to < 0.8.7 so we don't have the issue occur again. | 17:39 |
| rloo | jlvillal: maybe just send out email. cuz i'm not really sure it is worth discussing but maybe i'm missing something. | 17:39 |
| rloo | jlvillal: it seems like we could 1. < 0.8.7; 2. push for blocking 0.8.9; 3. patch our client | 17:39 |
| rloo | jlvillal: i'm fine doing any of the 3, but i'm not going to spend time doing all three to see which one people like. | 17:40 |
| rloo | jlvillal: my only question is whether anyone cares to get this fixed before the summit starts. | 17:40 |
| jlvillal | rloo: Fair enough. 1) gets my vote. 3 looks interesting but I don't know how much work is involved. | 17:40 |
| jlvillal | rloo: I would hope we could do 1 or 2 this week. To unblock the gate. | 17:41 |
| rloo | jlvillal: 3 is somewhat simple. i'll change the patch to do 1 and see what infra sez. | 17:41 |
| *** gridinv_ has joined #openstack-ironic | 17:41 | |
| jlvillal | rloo: thanks | 17:41 |
| *** alexpilotti has quit IRC | 17:41 | |
| NobodyCam | rloo: I intended it for stuff like what comes out of the cinder session on integration | 17:42 |
| NobodyCam | more of a sharing type thing.. | 17:43 |
| *** dguerri is now known as _dguerri | 17:46 | |
| *** cdearborn has quit IRC | 17:49 | |
| rloo | NobodyCam: hmm, maybe we need a recap/priorities thingy in Fri. | 17:54 |
| *** kkoski has quit IRC | 17:55 | |
| *** Nisha has joined #openstack-ironic | 17:56 | |
| rloo | NobodyCam: I just looked at the Fri stuff. In the last summit on the Fri, there was a bit of time (30 min or less) where we discussed/recap'd misc/priorities/heard from our fearless leader. | 17:59 |
| NobodyCam | rloo: :) I added a "other project session impact wrap-up discussion" | 18:00 |
| NobodyCam | think that covers it or? | 18:00 |
| *** e0ne is now known as e0ne_ | 18:00 | |
| rloo | NobodyCam: well, I wasn't thinking specifically about cross-project issues. | 18:00 |
| rloo | NobodyCam: oh, i didn't realize there was a separate etherpad for Fri. Am looking at the wrong thing then? | 18:01 |
| NobodyCam | https://etherpad.openstack.org/p/liberty-ironic-contributor-meetup | 18:01 |
| NobodyCam | oh | 18:01 |
| *** kkoski has joined #openstack-ironic | 18:02 | |
| NobodyCam | I should have links on the idea's pad... trying to keep things up to date as I shuffel them around | 18:02 |
| rloo | NobodyCam: does that etherpad cover everything from https://etherpad.openstack.org/p/liberty-ironic-design-summit-ideas? If so, maybe we can delete the 3.1-3.11 stuff | 18:02 |
| NobodyCam | rloo: yes it covers it all. | 18:02 |
| *** Sukhdev_ has joined #openstack-ironic | 18:03 | |
| rloo | NobodyCam: there's too much stuff; I can't grok it all! We have the top part of the main ether pad with stuff, then below that there are notes, and then there are other etherpads. which is the source of truth. | 18:03 |
| NobodyCam | hehhe: anything I have moved I cut the orginal notes and (should have) put a link to the pad where I moved it to. | 18:04 |
| rloo | NobodyCam: wrt documentation on Fri. I can't remember if I put that there or not, but from my point of view, if we have a liason with the docs folks (and we do, pshige), unless pshige wants to discuss something at the summit, I don't think we need that. | 18:04 |
| *** e0ne_ is now known as e0ne | 18:04 | |
| *** kkoski has quit IRC | 18:05 | |
| rloo | NobodyCam: and I just discussed with jlvillal about httpretty, we don't need to discuss that on Fri so I will delete it | 18:05 |
| *** Sukhdev has quit IRC | 18:05 | |
| NobodyCam | :) +++ | 18:05 |
| *** e0ne has quit IRC | 18:06 | |
| rloo | NobodyCam: wrt Fri -- are these ideas, and/or what is the order in which we do them? | 18:06 |
| rloo | NobodyCam: and/or are there going to be parallel things happening on Fri? | 18:06 |
| *** ijw has quit IRC | 18:07 | |
| *** vdrok_ has joined #openstack-ironic | 18:07 | |
| NobodyCam | they are Idea's an I have not thought about order, execpt to put the "show and tell" items on the top, but thats it (so far) | 18:07 |
| *** ijw has joined #openstack-ironic | 18:07 | |
| rloo | NobodyCam: ok. I suspect things may change between now and then. I'm already exhausted looking at the list of stuff :-) | 18:09 |
| *** stendulker has quit IRC | 18:09 | |
| NobodyCam | rloo: +++, just trying to make sure we're ready to roll at the summit. please feel free to edit anything you see wrong :) | 18:10 |
| *** ijw_ has joined #openstack-ironic | 18:10 | |
| rloo | NobodyCam: nothing wrong I don't think. Just trying to think of what is the best use of all our time face-to-face blah blah ;) | 18:10 |
| NobodyCam | rloo +++ :) | 18:11 |
| NobodyCam | wrong was wrong word | 18:11 |
| rloo | NobodyCam: :) No worries. | 18:11 |
| NobodyCam | wow wounders what "W"'s we can wield | 18:12 |
| *** ijw has quit IRC | 18:12 | |
| rloo | NobodyCam: what? :) | 18:12 |
| NobodyCam | lol | 18:12 |
| BadCub | NobodyCam: "W"'s? I don;t think Starwood would approve! | 18:14 |
| BadCub | lol | 18:14 |
| NobodyCam | lol | 18:14 |
| * BadCub still feels guilt for leaving Starwood for Hilton :( | 18:14 | |
| *** kkoski has joined #openstack-ironic | 18:14 | |
| NobodyCam | but we're staying at the 4 seasons this summit | 18:15 |
| BadCub | I know... It's Not a St. Regis. But we can slum it for a week! | 18:15 |
| NobodyCam | hehheehe | 18:15 |
| * NobodyCam wounders where mordred is staying for the summit? | 18:16 | |
| mordred | one of the westins | 18:16 |
| * BadCub is willing to bet it is a Starwood brand | 18:16 | |
| BadCub | Yep. I was right! | 18:16 |
| BadCub | lol | 18:16 |
| mordred | :) | 18:16 |
| NobodyCam | :) ahh morning mordred | 18:16 |
| mordred | morning NobodyCam ! | 18:16 |
| NobodyCam | hehehe :) +++ | 18:16 |
| BadCub | mornin mordred :) | 18:16 |
| mordred | morning BadCub ... how's things in the desert? | 18:17 |
| * NobodyCam notes its always morning in ironic-landa | 18:17 | |
| BadCub | mordred: they are well! We are getting new walls built this week! Hows things [wherever] you are today? | 18:17 |
| mordred | BadCub: they're lovely! if only I knew where I was | 18:18 |
| BadCub | mordred: lol~ | 18:18 |
| NobodyCam | lol | 18:18 |
| BadCub | We are getting ready for our summer migration back to washington next month | 18:18 |
| *** praneshp has quit IRC | 18:19 | |
| * BadCub runs out for smoke and to review contractor progress | 18:20 | |
| *** praneshp has joined #openstack-ironic | 18:24 | |
| Nisha | NobodyCam, rloo , devananda i raised https://review.openstack.org/182572, https://review.openstack.org/182581 and https://review.openstack.org/182601 for handling cpabilities across nova and ironic. This is just an approach i thought of about capabilities. along with these three it requires fixing https://bugs.launchpad.net/ironic/+bug/1421261 also. Just a FYI to give view on overall approach | 18:25 |
| openstack | Launchpad bug 1421261 in Ironic " node's properties['capabilities'] value should be a dictionary" [Wishlist,Triaged] - Assigned to Michael Turek (mjturek) | 18:25 |
| *** Marga_ has joined #openstack-ironic | 18:28 | |
| NobodyCam | hi / morning Nisha | 18:30 |
| Nisha | Good Morning NobodyCam | 18:30 |
| BadCub | hi Nisha :) | 18:31 |
| NobodyCam | Nisha: that is down on, one of the summit pads already, so I expect we'll go over it there.. | 18:31 |
| Nisha | hi BadCub , ^^^^ | 18:31 |
| Nisha | NobodyCam, yes i expect that | 18:31 |
| *** ukalifon has joined #openstack-ironic | 18:31 | |
| NobodyCam | will you be there too? | 18:32 |
| Nisha | i just wanted to bring in the approach i just thought of to solve the issue cleanly | 18:32 |
| Nisha | No NobodyCam :( | 18:32 |
| NobodyCam | :( | 18:32 |
| Nisha | rameshg87, wanyen and stendulker will be there | 18:32 |
| NobodyCam | ahh okay ;) | 18:33 |
| BadCub | bummer Nisha Would be really cool to meet yo in person :) | 18:33 |
| Nisha | :) | 18:33 |
| Nisha | BadCub, next time :) | 18:33 |
| BadCub | Nisha: indeed!!!!! :) | 18:34 |
| NobodyCam | :) +++ | 18:34 |
| *** ijw_ has quit IRC | 18:34 | |
| *** Marga_ has quit IRC | 18:35 | |
| *** ijw has joined #openstack-ironic | 18:35 | |
| Nisha | BadCub, NobodyCam just a request to go through the approach in above reviews, from above reviews two are in nova (nova-ironic-virt driver, nova scheduler and it touches the nova database too) | 18:35 |
| *** Marga_ has joined #openstack-ironic | 18:35 | |
| *** ijw has quit IRC | 18:35 | |
| Nisha | and one spec is in ironic which is just the one i picked up from backlog | 18:35 |
| *** ijw has joined #openstack-ironic | 18:36 | |
| BadCub | oh cool. Are you re-proposing it for L? | 18:36 |
| NobodyCam | ack !! | 18:36 |
| Nisha | nova approach is necessary to be thought upon as that drives the ironic also | 18:36 |
| Nisha | yes | 18:36 |
| Nisha | BadCub, yes | 18:37 |
| BadCub | Nisha: coolness | 18:37 |
| BadCub | :) | 18:37 |
| Nisha | NobodyCam, BadCub the bug 1421261 is equally required for this overall approach | 18:38 |
| openstack | bug 1421261 in Ironic " node's properties['capabilities'] value should be a dictionary" [Wishlist,Triaged] https://launchpad.net/bugs/1421261 - Assigned to Michael Turek (mjturek) | 18:38 |
| Nisha | mjturek1, are u working on above bug? | 18:39 |
| * BadCub likes seeing bugs get fixed :) | 18:39 | |
| *** ifarkas has quit IRC | 18:40 | |
| Nisha | BadCub, me too | 18:40 |
| Nisha | :) | 18:40 |
| BadCub | ++ :) | 18:40 |
| *** e0ne has joined #openstack-ironic | 18:45 | |
| rloo | hi NobodyCam, would you mind approving this? (Or should I update it. Not sure.) https://review.openstack.org/#/c/180662/ | 18:45 |
| mjturek1 | Nisha haven't spent much time on it would you like me to give it up? | 18:47 |
| * NobodyCam looks | 18:47 | |
| mjturek1 | Nisha, but planning to look into it | 18:47 |
| Nisha | mjturek1, if you are planning then fine else i can take it up | 18:48 |
| mjturek1 | Nisha, yeah I'd like to still take a go at it :) | 18:48 |
| mjturek1 | I'll focus on it a bit more | 18:48 |
| Nisha | mjturek1, thanks but that requires Nova also to change else nova boot will start failing for capabilities | 18:49 |
| Nisha | i have already proposed the spec for it... | 18:49 |
| Nisha | mjturek1, https://review.openstack.org/#/c/182572 | 18:50 |
| Nisha | changing capabilities to accept values as dictionary alone in ironic is not sufficient | 18:51 |
| *** Sukhdev_ has quit IRC | 18:51 | |
| mjturek1 | Nisha ahhhh I see | 18:51 |
| Nisha | mjturek1, will wait for the update.... | 18:51 |
| NobodyCam | rloo: lgtm, did you want to fix the spelling nit in the commit message or ? | 18:51 |
| rloo | NobodyCam: not particularly. You're welcome to do it and then +A. | 18:52 |
| mjturek1 | Nisha, just to confirm, before this bug can be fix, need to have this blueprint implemented in nova? | 18:52 |
| mjturek1 | making sure I'm understanding | 18:52 |
| Nisha | hmm not exactly as we will have to maintain compatibility... so the current format will still be supported for backward compatibility in ironic | 18:53 |
| openstackgerrit | Chris Krelle proposed openstack/ironic-specs: Add a 'Client (CLI) impact' section to the spec https://review.openstack.org/180662 | 18:53 |
| mjturek1 | Nisha, right, understood | 18:54 |
| NobodyCam | rloo: done | 18:55 |
| Nisha | hence it will work. but it breaks the functionality as nova scheduler cannot select ironic node with capabilities defined accepting values as dictionary even if flavor has been created for a key in the capabilities | 18:55 |
| NobodyCam | :) thank you :) | 18:55 |
| rloo | thx NobodyCam! I'm not core reviewer there so if I updated, I'd have lost all the votes and I'd have to go begging, which I didn't want to do :) | 18:55 |
| NobodyCam | hehehe :) | 18:56 |
| mjturek1 | Nisha ahhhh I see | 18:56 |
| NobodyCam | thats a easy one. I commented what I did so others could follow | 18:56 |
| devananda | victor_lowther: ping | 18:56 |
| Nisha | mjturek1, so IMO both of them shud be implemented together but we need to see how things should behave in such situations...might be we have to convert the dictionary to string in ironic virt driver | 18:57 |
| Nisha | for compatibility but thats again a change in nova | 18:57 |
| Nisha | i dont think ironic is intelligent to know if nova understand string form of capabilities or dictionary format | 18:58 |
| *** kkoski has quit IRC | 18:58 | |
| *** kkoski has joined #openstack-ironic | 18:59 | |
| Nisha | so thats a challenge how do we handle such situations.... | 18:59 |
| Nisha | the dictionary format is required to handle multiple values for a key | 18:59 |
| *** _dguerri is now known as dguerri | 18:59 | |
| mjturek1 | hmmm, so the goal wouldn't be storing the capabilities as a dict on the ironic side and then sending them as a string when requested by nova? | 19:01 |
| mjturek1 | Nisha ^ | 19:01 |
| Nisha | No | 19:01 |
| openstackgerrit | Merged openstack/ironic-specs: Add a 'Client (CLI) impact' section to the spec https://review.openstack.org/180662 | 19:01 |
| Nisha | it shud be dictionary | 19:01 |
| mjturek1 | I see | 19:02 |
| Nisha | we need to actually discuss this point ...thats why i was thinking if the bug should be really a bug or a spec | 19:02 |
| Nisha | it will touch almost complete ironic | 19:02 |
| Nisha | the bug needs to cover the impact at each place | 19:03 |
| mjturek1 | right, might be worth a spec as it's a little more involved than the original bug intended | 19:03 |
| Nisha | :( | 19:03 |
| Nisha | NobodyCam, rloo ^^^ what do u say | 19:03 |
| Nisha | mjturek1, i think nova spec and code should get merged first for it. and then ironic | 19:04 |
| mjturek1 | Nisha agreed | 19:05 |
| Nisha | so we dont have compatibility issues | 19:05 |
| Nisha | in that way it might be simpler and straight fwd | 19:05 |
| mjturek1 | yeah totally, otherwise values will need to be different types on each end (if I'm understanding properly) | 19:06 |
| Nisha | yes correct | 19:07 |
| NobodyCam | Nisha: I have looked at the reviews you've linked yet. but if I'm following the chat. I would say yes it is going to require a spec | 19:07 |
| Nisha | NobodyCam, i was thinking so, should i throw up a spec for it even though it is filed as bug? | 19:08 |
| NobodyCam | Nisha: in this case I would say yes. | 19:08 |
| NobodyCam | please also but the bug in the refference section | 19:09 |
| Nisha | Ok. I would add up a spec for it asap before summit | 19:09 |
| Nisha | yes | 19:09 |
| NobodyCam | :) TY Nisha :) | 19:09 |
| vdrok_ | hi everyone | 19:10 |
| mjturek1 | Nisha, I'll unassign myself from the bug for now. I'd love to help down the road though! | 19:10 |
| vdrok_ | could anyone take a look at this tempest change - https://review.openstack.org/#/c/178607/ | 19:11 |
| Nisha | mjturek1, :) i will add you as coworker for the spec | 19:11 |
| Nisha | :) | 19:11 |
| mjturek1 | Nisha that'd be wonderful! :) | 19:11 |
| Nisha | :) | 19:11 |
| vdrok_ | there hasn't been much reviews about it, i'll start pinging qa guys tomorrow | 19:11 |
| *** Sukhdev has joined #openstack-ironic | 19:16 | |
| NobodyCam | devananda: have you looked at the work on the summit pads... | 19:17 |
| devananda | NobodyCam: so one of the areas where I feel we could do better at the summit is having some sort of goal for each of the sessions | 19:17 |
| devananda | what do we want to get out of each one? | 19:17 |
| BadCub | fluffy things? | 19:18 |
| devananda | (and actually state that in the 'pad) | 19:18 |
| NobodyCam | ahh yes I have not done anything along that line | 19:18 |
| devananda | BadCub: for this trip, I'm leaving my fluffy things at home. they'll come out again in august ... :) | 19:19 |
| NobodyCam | ++ lol | 19:19 |
| BadCub | LOL | 19:19 |
| BadCub | I always bring fluffy things with me | 19:19 |
| NobodyCam | devananda: something along the lines of what I just added to https://etherpad.openstack.org/p/liberty-ironic-cycle-goals Line #45 | 19:26 |
| NobodyCam | ahh | 19:29 |
| *** pcrews has quit IRC | 19:29 | |
| * krotscheck has finally managed to get the ironic webclient to multi-cloud. | 19:30 | |
| rloo | NobodyCam: "spear reviews"? do you mean "steer reviews"? (I don't know if it is you or not) | 19:30 |
| devananda | krotscheck: woot! | 19:32 |
| *** e0ne has quit IRC | 19:32 | |
| krotscheck | I don't suppose someone here can send me a dump of an ironic database so I can start working with data? | 19:33 |
| *** e0ne has joined #openstack-ironic | 19:33 | |
| BadCub | we are reviewing cows? Huh? | 19:35 |
| * BadCub needs more coffee.... Possibly lots more coffee..... | 19:36 | |
| rloo | BadCub: ha, that's good, I didn't think of that! | 19:36 |
| BadCub | rloo: lol | 19:36 |
| devananda | krotscheck: fake data or real data? | 19:36 |
| krotscheck | devananda: fake'll be fine | 19:37 |
| BadCub | glad I grew up on a ranch... I can review cows! | 19:37 |
| krotscheck | I need to make things pretty | 19:37 |
| krotscheck | Not accurate | 19:37 |
| krotscheck | (Story of my life, really....) | 19:37 |
| BadCub | brb | 19:37 |
| NobodyCam | rloo: doh | 19:37 |
| devananda | krotscheck: TheJulia? ^^ | 19:38 |
| TheJulia | krotscheck: where would you like to receive a database dump? | 19:39 |
| krotscheck | TheJulia: DCC? Email (michael.krotscheck@hp.com)? Carrier Pidgeon? | 19:40 |
| TheJulia | DCC over IP over Carrier Pidgeon sounds a little too painful | 19:41 |
| krotscheck | TheJulia: So.... I shouldn't start up the rotisserie then? | 19:42 |
| *** praneshp has quit IRC | 19:42 | |
| * krotscheck was looking forward to having pidgeon for lunch. | 19:42 | |
| *** praneshp has joined #openstack-ironic | 19:43 | |
| devananda | krotscheck: if TheJulia sent a carrier pidgeon now, somehow I doubt you'd get it in time for lunch ... | 19:43 |
| TheJulia | krotscheck: no... creating it now | 19:43 |
| krotscheck | TheJulia: Cool, thanks | 19:43 |
| TheJulia | yeah, north carolina is a decent distance from... ??? | 19:43 |
| *** e0ne has quit IRC | 19:43 | |
| krotscheck | TheJulia: Seattle | 19:43 |
| devananda | krotscheck: you dont happen to be downtown today, by chance? | 19:43 |
| krotscheck | devananda: Nope. | 19:44 |
| krotscheck | devananda: Will likely be on friday though. | 19:44 |
| praneshp | hey ironic. I’m trying ot understand the availability zone concept in Ironic. I see that AvailabilityZoneFilter is included in the ironic_host_manager (https://github.com/openstack/nova/blob/master/nova/scheduler/ironic_host_manager.py#L36) | 19:44 |
| krotscheck | The argument in my head goes something like: "That would require getting out of my bathrobe...." | 19:44 |
| praneshp | but looks like ironic nodes are not associated to any partucular nova-compute | 19:45 |
| devananda | krotscheck: let me guess - the bathrobe will win? | 19:45 |
| devananda | praneshp: correct. there is no correlation between nova-compute and ironic node. In fact, running more than one nova-compute process with ironic is not supported | 19:45 |
| krotscheck | devananda: Well, it's really more that by the time I've decided to get dressed, I'm already three conversations into a technical discussion and can't exactly stop to take a shower. | 19:46 |
| praneshp | and you can only put nova-compute hosts in availabilty zones | 19:46 |
| devananda | praneshp: doing so is possible, but leads but some race conditions | 19:46 |
| krotscheck | Or commute, for that matte.r | 19:46 |
| praneshp | devananda: then why the AvailabilityZoneFilter? | 19:46 |
| devananda | praneshp: you could run some KVM and some Ironic nova-computes | 19:47 |
| devananda | praneshp: or perhaps you could associate different nova-computes with different ironic clusters -- I haven't tested this, but it's an interesting possibility | 19:47 |
| *** Sukhdev_ has joined #openstack-ironic | 19:47 | |
| praneshp | devananda: also a related question. WHat’s the best way to do avail-zone like things in ironic? A simple way I can think of is put it in Capabilities and add a scheduler filter to parse out in nova-scheduler side | 19:48 |
| praneshp | devananda: I see oyur point about kvm/ironice, etc | 19:48 |
| *** ijw_ has joined #openstack-ironic | 19:49 | |
| devananda | praneshp: depends on what "az-like-things" you want to do | 19:49 |
| devananda | but probably yes, capabilities and a filter is a reasonable way to go. you can also stash information in ironic's node.properties and then let nova filter on that (custom filter, probably) | 19:49 |
| praneshp | devananda: I want to group ironic hosts into pools based on hardware spec, etc. (similar to what an aggregate would be in nova). Then based on wht the user passes, pick the right hypervisor. | 19:50 |
| *** Sukhdev has quit IRC | 19:50 | |
| praneshp | devananda: yeah, thanks. That makes sense. Pretty straightforward I guess. | 19:50 |
| devananda | praneshp: node.properties already covers cpu/ram/disk/architecture | 19:50 |
| devananda | and yea, just extend that as needed, like, for GPU type, etc | 19:50 |
| praneshp | devananda: thanks for the quick replies! | 19:50 |
| devananda | you may also want to take a look at the work Nisha has proposed for more advanced capabilities filtering in Nova | 19:51 |
| *** ijw has quit IRC | 19:51 | |
| praneshp | devananda:I’ll take a look. Is there a filter that does not depend on getting the entire ironic node list? | 19:51 |
| praneshp | I understand that it is the way nova does it, but I’m seeing that it could be a bottleneck | 19:52 |
| praneshp | with large-ish deployments | 19:52 |
| devananda | nova should already have cached the node.properties as capabilities | 19:52 |
| devananda | nova scheduler will filter that like it filters with any other hypervisor | 19:52 |
| devananda | which, yes, is inefficient | 19:52 |
| NobodyCam | brb | 19:52 |
| praneshp | Hmmm. thanks devananda! | 19:53 |
| devananda | there's certainly room for improvement in the scheduler itself, but that's outside the scope of Ironic ... there are a couple sessions planned next week to discuss scheduler changes, fwiw | 19:53 |
| praneshp | devananda: correct, I agree that’s probably something in nova rather than ironic. I was just picking your brain. | 19:53 |
| devananda | :) | 19:54 |
| devananda | NobodyCam: to your question, yes, sort of | 19:54 |
| *** kkoski has quit IRC | 19:54 | |
| devananda | NobodyCam: I think the driver session has some good questions and goals, too | 19:54 |
| devananda | NobodyCam: but eg. https://etherpad.openstack.org/p/liberty-ironic-state-machine-next-steps is pretty empty | 19:56 |
| *** kkoski has joined #openstack-ironic | 19:57 | |
| *** dprince has quit IRC | 19:57 | |
| devananda | adam_g: ping re: https://etherpad.openstack.org/p/liberty-ironic-functional-testing -- will you have time to jot anything down before the summit, or shall we just take really good notes after performing a mass vulcan mind meld in the session? :) | 19:57 |
| * devananda steps afk for a minute | 19:58 | |
| *** e0ne has joined #openstack-ironic | 19:59 | |
| *** ijw_ has quit IRC | 19:59 | |
| *** vdrok_ has quit IRC | 20:05 | |
| NobodyCam | :) | 20:07 |
| adam_g | devananda, ill jot some stuff down and happy to be there at the session, too | 20:11 |
| adam_g | is there a time set? | 20:11 |
| devananda | yah | 20:11 |
| NobodyCam | :) | 20:11 |
| devananda | adam_g: http://libertydesignsummit.sched.org/event/0d680d665396cf6bb91f65bc41f5f164 | 20:11 |
| NobodyCam | thank you adam_g | 20:11 |
| devananda | adam_g: thanks much | 20:12 |
| adam_g | np | 20:13 |
| NobodyCam | oh ya state machine is kinda empty looking | 20:14 |
| * NobodyCam dig up random news feed puller to apend to the bottom | 20:14 | |
| NobodyCam | lol j/k | 20:14 |
| *** kkoski has quit IRC | 20:17 | |
| *** kkoski has joined #openstack-ironic | 20:18 | |
| *** kkoski1 has joined #openstack-ironic | 20:20 | |
| *** kkoski has quit IRC | 20:20 | |
| *** gridinv_ has quit IRC | 20:22 | |
| *** pcrews has joined #openstack-ironic | 20:25 | |
| openstackgerrit | Julia Kreger proposed openstack/bifrost: Unmask and prevent failure. https://review.openstack.org/182462 | 20:25 |
| *** ijw has joined #openstack-ironic | 20:36 | |
| *** ijw has quit IRC | 20:37 | |
| *** ijw has joined #openstack-ironic | 20:37 | |
| *** achanda has quit IRC | 20:48 | |
| *** Nisha has quit IRC | 20:49 | |
| openstackgerrit | Jarrod Johnson proposed stackforge/pyghmi: Leave numbers intact https://review.openstack.org/182837 | 20:54 |
| *** kkoski1 has quit IRC | 20:54 | |
| *** Sukhdev has joined #openstack-ironic | 20:59 | |
| *** thrash is now known as thrash|g0ne | 21:00 | |
| *** Sukhdev_ has quit IRC | 21:01 | |
| *** jjohnson2 has quit IRC | 21:04 | |
| *** NobodyCa1 has joined #openstack-ironic | 21:07 | |
| NobodyCa1 | hey hey JoshNang just looking at https://review.openstack.org/#/c/159322 & https://review.openstack.org/#/c/161457 any chance you have time to update them before the summit | 21:07 |
| *** GheRiver1 has joined #openstack-ironic | 21:07 | |
| NobodyCa1 | oh joy | 21:07 |
| *** ekarlso has quit IRC | 21:09 | |
| *** NobodyCam has quit IRC | 21:09 | |
| *** greghaynes has quit IRC | 21:09 | |
| *** GheRivero has quit IRC | 21:09 | |
| *** greghayn1 has joined #openstack-ironic | 21:09 | |
| *** ekarlso has joined #openstack-ironic | 21:10 | |
| *** NobodyCa1 is now known as NobodyCam | 21:11 | |
| *** achanda has joined #openstack-ironic | 21:11 | |
| *** Marga_ has quit IRC | 21:11 | |
| JoshNang | NobodyCam: probably not | 21:11 |
| *** Marga_ has joined #openstack-ironic | 21:12 | |
| NobodyCam | :) NP I just added them as ref to https://etherpad.openstack.org/p/liberty-ironic-state-machine-next-steps | 21:12 |
| NobodyCam | so you may get questions on them | 21:12 |
| NobodyCam | :-p | 21:12 |
| JoshNang | NobodyCam: gotcha. the api one...no idea what to do there. probably has to be an async api with a later return value, but i haven't seen anything like that in our api | 21:13 |
| JoshNang | zapping should be relatively quick, just nee to find time after the summit to do it | 21:14 |
| NobodyCam | :) +++ time whats that | 21:15 |
| JoshNang | heh | 21:15 |
| JoshNang | :D | 21:15 |
| *** praneshp has quit IRC | 21:16 | |
| *** greghayn1 is now known as greghaynes | 21:29 | |
| NobodyCam | devananda: when you back I have a couple of questions on / about https://etherpad.openstack.org/p/liberty-ironic-nova-virt-driver | 21:41 |
| *** Sukhdev_ has joined #openstack-ironic | 21:44 | |
| mrda | Morning | 21:44 |
| NobodyCam | good morning mrda | 21:45 |
| mrda | hey NobodyCam | 21:45 |
| NobodyCam | mrda: jlvillal: great start on the Nova bug stuff :) | 21:45 |
| NobodyCam | TY both | 21:46 |
| mrda | \o/ | 21:46 |
| *** trown is now known as trown|outttypeww | 21:46 | |
| jlvillal | NobodyCam: Thank you. | 21:46 |
| *** Sukhdev has quit IRC | 21:46 | |
| NobodyCam | :) | 21:47 |
| *** lucas-dinner has quit IRC | 22:03 | |
| *** igordcard has joined #openstack-ironic | 22:10 | |
| *** Marga_ has quit IRC | 22:11 | |
| *** Marga_ has joined #openstack-ironic | 22:11 | |
| *** achanda has quit IRC | 22:13 | |
| *** jamielennox|away is now known as jamielennox | 22:19 | |
| *** e0ne has quit IRC | 22:28 | |
| *** Sukhdev has joined #openstack-ironic | 22:33 | |
| NobodyCam | JayF: ahhh you session conflicts with another one I wanted to attend... /me wounder if he can clone himself in time | 22:34 |
| *** ekarlso has quit IRC | 22:34 | |
| *** ekarlso has joined #openstack-ironic | 22:34 | |
| NobodyCam | s/you/toyr/ | 22:34 |
| NobodyCam | gah typoed the correction even | 22:34 |
| *** praneshp has joined #openstack-ironic | 22:35 | |
| *** Sukhdev_ has quit IRC | 22:37 | |
| -openstackstatus- NOTICE: Gerrit and Zuul are going offline for reboots to fix a security vulnerability. | 22:39 | |
| *** ChanServ changes topic to "Gerrit and Zuul are going offline for reboots to fix a security vulnerability." | 22:39 | |
| *** achanda has joined #openstack-ironic | 22:40 | |
| *** Sukhdev has quit IRC | 22:43 | |
| *** openstackgerrit has quit IRC | 22:47 | |
| *** openstackgerrit has joined #openstack-ironic | 22:49 | |
| *** ChanServ changes topic to "Bare Metal Provisioning | Status: http://bit.ly/ironic-whiteboard | Docs: http://docs.openstack.org/developer/ironic/ | Bugs: https://bugs.launchpad.net/ironic" | 22:57 | |
| -openstackstatus- NOTICE: Gerrit and Zuul are back online. | 22:57 | |
| *** yuanying has joined #openstack-ironic | 23:01 | |
| *** achanda has quit IRC | 23:09 | |
| * BadCub thinks it is time to call it a day [or a night] and go do unproductive things | 23:13 | |
| * NobodyCam thinks there are quite a few sessions / talks about storage this summit.. | 23:15 | |
| NobodyCam | and lol https://xkcd.com/908/ | 23:18 |
| *** russell_h has quit IRC | 23:18 | |
| *** russell_h has joined #openstack-ironic | 23:19 | |
| *** russell_h has quit IRC | 23:19 | |
| *** russell_h has joined #openstack-ironic | 23:19 | |
| *** Marga_ has quit IRC | 23:20 | |
| *** Marga_ has joined #openstack-ironic | 23:21 | |
| harlowja | periodic probe for others to look at https://review.openstack.org/#/c/141961/ (which can hopefully get some of the FSM code out of ironic and share it with taskflow + ironic + others...) | 23:23 |
| harlowja | joshbot at your service :-P | 23:23 |
| NobodyCam | lol | 23:24 |
| harlowja | speak now or forever hold your peace | 23:24 |
| NobodyCam | I'm looking | 23:24 |
| harlowja | or whatever that saying is, lol | 23:24 |
| NobodyCam | lol | 23:25 |
| *** russell_h has quit IRC | 23:26 | |
| *** russell_h has joined #openstack-ironic | 23:26 | |
| *** achanda has joined #openstack-ironic | 23:26 | |
| *** russell_h has quit IRC | 23:27 | |
| *** russell_h has joined #openstack-ironic | 23:27 | |
| *** russell_h has quit IRC | 23:28 | |
| *** russell_h has joined #openstack-ironic | 23:28 | |
| *** igordcard has quit IRC | 23:29 | |
| NobodyCam | harlowja: luv the you? comemnts :) wow great effort | 23:29 |
| *** russell_h has quit IRC | 23:30 | |
| *** russell_h has joined #openstack-ironic | 23:30 | |
| harlowja | i only speak in python | 23:30 |
| harlowja | english fail many times for me, lol | 23:30 |
| harlowja | oh | 23:30 |
| harlowja | the "you?" | 23:30 |
| harlowja | i try to catch my reader, with subliminal messages | 23:31 |
| harlowja | not really that subliminal i guess | 23:31 |
| harlowja | hmmm | 23:31 |
| * harlowja needs to make them more subliminal | 23:31 | |
| *** praneshp has quit IRC | 23:31 | |
| harlowja | if i scatter "you want to help" every now and then, wonder if that would do it | 23:31 |
| NobodyCam | no no you did just that | 23:32 |
| NobodyCam | I like that you did it!!! | 23:32 |
| harlowja | ha | 23:32 |
| harlowja | thx | 23:32 |
| NobodyCam | :) humm mdon't see the work item to intergrate that lib with Ironic | 23:33 |
| NobodyCam | hehehehe j/k | 23:33 |
| harlowja | rloo u gonna help right | 23:33 |
| harlowja | righhhht | 23:33 |
| harlowja | ? | 23:33 |
| *** JayF has quit IRC | 23:33 | |
| rloo | harlowja: huh? what are you guys talking about? | 23:35 |
| *** JayF has joined #openstack-ironic | 23:35 | |
| harlowja | rloo fix it all | 23:36 |
| rloo | harlowja: oh. the automaton library. yeah. sure. maybe. will see. | 23:36 |
| harlowja | nuff said | 23:36 |
| harlowja | that doesn't sound very confidence inspiring | 23:36 |
| jlvillal | jroll: In regard to: https://review.openstack.org/#/c/130877/ Would you rather 1) Restore the patch (from Abandoned) or 2) Have me upload a brand new patch? Either works for me. | 23:36 |
| rloo | harlowja: depends on my priorities. go talk to my manager :) | 23:36 |
| harlowja | yea, sure, yeehawww, | 23:36 |
| harlowja | ^ much improved | 23:36 |
| harlowja | or in canadian, yes | 23:36 |
| * harlowja runs away | 23:37 | |
| rloo | jlvillal: not sure if jroll is 'listening'; he said he's away the rest of the week | 23:37 |
| jlvillal | rloo: Ah. thanks. | 23:37 |
| *** BadCub is now known as BadCub-Hiding | 23:38 | |
| rloo | jlvillal: I think it makes sense to use the same patch. it has all the comments etc with it | 23:38 |
| rloo | jlvillal: thx for taking it on :-) | 23:38 |
| jlvillal | rloo: Okay. I lack the power to restore it. But I still need to play around with it locally. So I can wait on that. | 23:38 |
| * rloo wonders if we should abandon patches like "they" do! | 23:38 | |
| mrda | rloo: I think jroll is off doing vacation and not listening at all for the rest of the week | 23:39 |
| jlvillal | rloo: I could see abandoning patches after X weeks of inactivity. | 23:39 |
| rloo | jlvillal: oh, joe mentioned a 'restore' button but I don't see it. think it is only avail for the author? | 23:39 |
| JayF | I can confirm he's not here the rest of the week | 23:39 |
| JayF | but you never know if jroll is listening | 23:39 |
| BadCub-Hiding | abandoning patches is cruel and inhumane. Just think how many patches out there are lonely and in need of love LOL | 23:39 |
| JayF | he is pro at IRC lurking | 23:39 |
| jlvillal | rloo: Yeah, I think so. When I'm ready I can ping Joe. | 23:39 |
| mrda | "Pro lurking" is totally a sport | 23:39 |
| * BadCub-Hiding lurks often | 23:39 | |
| JayF | everyone in channel right now is bad at it | 23:40 |
| rloo | mrda: do you know how to reactivate abandoned patches? is it only cores + author that can do it? | 23:40 |
| mrda | Please hold your call is important to us | 23:40 |
| harlowja | https://review.openstack.org/#/c/179780/ was something that popped up | 23:40 |
| mrda | (I'll go find out) | 23:40 |
| harlowja | 179780 seems releated to what u gus are talking about (abandon things) | 23:40 |
| harlowja | poor patches | 23:40 |
| harlowja | homless | 23:40 |
| rloo | harlowja: I would rather it be manually, although the tool doesn't discriminate or anything I guess. | 23:40 |
| harlowja | rloo well u can combine that with https://github.com/openstack-infra/release-tools/blob/master/release_notes.py#L28 | 23:41 |
| rloo | harlowja: we're trying to adopt an abandoned patch but we don't seem to have powers or we don't know how. | 23:41 |
| harlowja | and make a happy abandon? | 23:41 |
| rloo | harlowja: https://review.openstack.org/#/c/130877/ | 23:41 |
| harlowja | "We are sorry but we are $emotion to abandon your patch, have a nice day" | 23:41 |
| JayF | It's like mad libs | 23:42 |
| rloo | harlowja: yeah, I do like the emotion thingy. | 23:42 |
| harlowja | ya, someone needs to make better sayings :-P | 23:42 |
| harlowja | "We are sorry but you are $crap_emotion, have a nice day" (maybe not this one) | 23:42 |
| mrda | rloo: I think it's only the owner who can un-abandon | 23:43 |
| harlowja | someone with politeness can make all those templates :-P | 23:43 |
| JayF | jlvillal: let me have a sec | 23:43 |
| mrda | rloo: So I would suggest starting a new patch, referencing the original unabandoned one in the commit message so that any previous discussion can be easily found. | 23:44 |
| jlvillal | JayF: Okay. Though confused. | 23:44 |
| JayF | jlvillal: We were G+'ing a sec ago. Was going to see if I could get jroll to unabandon it | 23:45 |
| mrda | s/unabandoned/abandoned/ | 23:45 |
| JayF | but he seems to be offline | 23:45 |
| jlvillal | JayF: Ah thanks! | 23:45 |
| jlvillal | JayF: No hurry. I am still trying to get unittests to pass. | 23:45 |
| rloo | mrda: hmm. that doesn't seem right (that only the owner can unabandon). Oh well. | 23:45 |
| JayF | well no need to thank me, it didn't work | 23:45 |
| rloo | jlvillal: plan B -- see what mrda said. New patch :-) Whatever works for you when you're ready. | 23:46 |
| mrda | right as in legally correct? or right as in morally correct? | 23:46 |
| jlvillal | rloo: Okay. I have the assumption that the person (joe) had the power to abandon it probably also has the power to un-abandon. | 23:46 |
| * mrda likes it that jlvillal is fixing nova bugs | 23:47 | |
| jlvillal | mrda: Uh... I haven't yet got that far! :) | 23:48 |
| jlvillal | mrda: I'm in the attempting to fix stage :P | 23:48 |
| rloo | jlvillal: i suppose we could test out that theory by abandoning someone's patch ;) | 23:48 |
| jlvillal | :) | 23:48 |
| * NobodyCam has one you can use to test ... should you wish | 23:49 | |
| * mrda looks for an abandoned ironic change that a core could try and unabandon | 23:49 | |
| rloo | sure NobodyCam | 23:49 |
| JayF | mrda: lets do some science, 1s | 23:49 |
| mrda | ooooh, science | 23:50 |
| * mrda rubs hands together | 23:50 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: test post please ignore https://review.openstack.org/182893 | 23:50 |
| JayF | ^ is now abandoned | 23:50 |
| mrda | https://review.openstack.org/#/c/111118/ or this one | 23:51 |
| rloo | JayF: so I see a 'Restore Change' button. Anyone else see it? | 23:51 |
| mrda | (sorry JayF, I'm too slow) | 23:51 |
| * BadCub-Hiding holds up protest signs saying give Patches a Chance. Don't abandon the patches you love | 23:51 | |
| mrda | rloo: I can't see the Restore button | 23:51 |
| mrda | rloo: so it appears cores can | 23:52 |
| rloo | JayF: do you see it? | 23:52 |
| JayF | rloo: I do, but I'm the owner | 23:52 |
| rloo | mrda: yeah, that's good to know | 23:52 |
| rloo | ok, so our experiment 'worked'. cores+owner can unabandon. | 23:52 |
| mrda | jlvillal: you might like to see if there's a core in #openstack-nova and ask if they can restore your patch series | 23:52 |
| *** russell_h has quit IRC | 23:52 | |
| *** russell_h has joined #openstack-ironic | 23:53 | |
| NobodyCam | https://www.youtube.com/watch?v=Jm-upHSP9KU | 23:53 |
| jlvillal | mrda: Okay. I will wait until I am ready to have something to submit first. | 23:53 |
| mrda | jlvillal: dansmith is active right now, just sayin' | 23:53 |
| jlvillal | mrda: Ok I'll try :) | 23:53 |
| JayF | NobodyCam: seen the movie :D | 23:54 |
| jlvillal | mrda: Done! :) | 23:55 |
| *** mtanino has quit IRC | 23:57 | |
| NobodyCam | JayF: :) yep, remembers seeing it the theater | 23:59 |
| JayF | I am not 100% sure when it was released | 23:59 |
| JayF | but if my guess is right I wasn't alive | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!