| *** stendulker has quit IRC | 00:10 | |
| *** Marga_ has joined #openstack-ironic | 00:14 | |
| devananda | random thought for third-party CI | 00:15 |
|---|---|---|
| *** achanda has joined #openstack-ironic | 00:15 | |
| devananda | what if the vendors used something more like bifrost, rather than a full devstack + tempest run | 00:15 |
| devananda | since the point is to test the interaction between ironic and some special flavor of hardware that can't be virtualized... why require all the layers of openstack above ironic in that test environment? | 00:16 |
| Shrews | devananda: do 3rd parties use tempest? if so, that might prevent that | 00:18 |
| jroll | devananda: that seems fine to me, I'm kind of hacking on something similar | 00:19 |
| jroll | Shrews: what 3rd parties | 00:19 |
| Shrews | the vendor CI | 00:19 |
| jroll | that was mostly a joke | 00:19 |
| *** achanda has quit IRC | 00:20 | |
| jroll | Shrews: is your question "are 3rd parties required to use tempest?' | 00:20 |
| jroll | or "does current ironic 3rd party CI use tempest?" | 00:21 |
| Shrews | jroll: the latter | 00:21 |
| jroll | (the latter of which I'm not terribly concerned about | 00:21 |
| jroll | we only have the xcat thing, if they want to continue using nova/tempest that's fine | 00:21 |
| jroll | but using something bifrosty would make it easier to set up | 00:21 |
| *** meghal has quit IRC | 00:27 | |
| devananda | the ilo team has been trying to get something going | 00:29 |
| devananda | i suspect this would make it much easier for them | 00:29 |
| *** ijw has quit IRC | 01:02 | |
| *** krtaylor has quit IRC | 01:10 | |
| openstackgerrit | Tan Lin proposed openstack/ironic: Improve AMT driver doc https://review.openstack.org/169606 | 01:11 |
| *** lazy_prince has quit IRC | 01:11 | |
| *** stendulker has joined #openstack-ironic | 01:15 | |
| *** chenglch has joined #openstack-ironic | 01:18 | |
| *** spandhe has quit IRC | 01:23 | |
| *** mtanino has quit IRC | 01:26 | |
| *** igordcard_ has quit IRC | 01:31 | |
| *** killer_prince has joined #openstack-ironic | 01:34 | |
| *** killer_prince is now known as lazy_prince | 01:34 | |
| *** krtaylor has joined #openstack-ironic | 01:46 | |
| *** stendulker has quit IRC | 01:53 | |
| *** chenglch has quit IRC | 01:57 | |
| *** chenglch has joined #openstack-ironic | 01:59 | |
| *** stendulker has joined #openstack-ironic | 01:59 | |
| *** Marga_ has quit IRC | 02:11 | |
| *** lazy_prince has quit IRC | 02:18 | |
| *** killer_prince has joined #openstack-ironic | 02:20 | |
| *** killer_prince is now known as lazy_prince | 02:20 | |
| *** lazy_prince has quit IRC | 02:27 | |
| *** killer_prince has joined #openstack-ironic | 02:28 | |
| *** killer_prince is now known as lazy_prince | 02:29 | |
| *** achanda has joined #openstack-ironic | 02:37 | |
| *** chlong has quit IRC | 02:42 | |
| *** ramineni has joined #openstack-ironic | 02:44 | |
| *** lazy_prince has quit IRC | 02:44 | |
| *** chlong has joined #openstack-ironic | 02:44 | |
| *** stendulker has quit IRC | 02:44 | |
| *** killer_prince has joined #openstack-ironic | 02:46 | |
| *** killer_prince is now known as lazy_prince | 02:47 | |
| *** achanda has quit IRC | 02:47 | |
| *** achanda has joined #openstack-ironic | 02:48 | |
| *** meghal has joined #openstack-ironic | 02:54 | |
| *** meghal1 has joined #openstack-ironic | 03:01 | |
| *** meghal has quit IRC | 03:03 | |
| *** Haomeng has joined #openstack-ironic | 03:29 | |
| *** coolsvap|afk is now known as coolsvap | 03:45 | |
| *** jamielennox is now known as jamielennox|away | 03:47 | |
| *** Marga_ has joined #openstack-ironic | 03:54 | |
| *** Marga_ has quit IRC | 03:54 | |
| *** Marga_ has joined #openstack-ironic | 03:55 | |
| *** harlowja is now known as harlowja_away | 03:55 | |
| *** sandhya has joined #openstack-ironic | 03:56 | |
| *** naohirot has joined #openstack-ironic | 04:01 | |
| *** achanda has quit IRC | 04:13 | |
| *** Nisha has joined #openstack-ironic | 04:18 | |
| *** achanda has joined #openstack-ironic | 04:20 | |
| *** coolsvap is now known as coolsvap|afk | 04:41 | |
| *** persia_ has quit IRC | 04:44 | |
| *** persia_ has joined #openstack-ironic | 04:45 | |
| *** coolsvap|afk is now known as coolsvap | 04:48 | |
| *** lazy_prince has quit IRC | 04:48 | |
| *** pradipta has joined #openstack-ironic | 04:58 | |
| *** jcoufal has joined #openstack-ironic | 05:07 | |
| *** ukalifon1 has joined #openstack-ironic | 05:11 | |
| *** jcoufal has quit IRC | 05:13 | |
| *** jcoufal has joined #openstack-ironic | 05:16 | |
| *** kalpase has joined #openstack-ironic | 05:17 | |
| *** kalpase has quit IRC | 05:26 | |
| *** kalpase has joined #openstack-ironic | 05:34 | |
| *** saripurigopi has joined #openstack-ironic | 05:52 | |
| *** killer_prince has joined #openstack-ironic | 06:09 | |
| *** killer_prince is now known as lazy_prince | 06:10 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Imported Translations from Transifex https://review.openstack.org/169184 | 06:12 |
| *** Marga_ has quit IRC | 06:29 | |
| *** spandhe has joined #openstack-ironic | 06:32 | |
| *** pas-ha has joined #openstack-ironic | 06:40 | |
| *** jistr has joined #openstack-ironic | 06:47 | |
| *** achanda has quit IRC | 07:04 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: boot_mode is overwritten in node properties https://review.openstack.org/167952 | 07:04 |
| *** dtantsur|afk is now known as dtantsur | 07:04 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: localboot deploy fails for iscsi_ilo driver https://review.openstack.org/168666 | 07:05 |
| *** spandhe has quit IRC | 07:07 | |
| *** achanda has joined #openstack-ironic | 07:08 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Update ilo drivers documentation for inspection https://review.openstack.org/170065 | 07:13 |
| *** athomas has joined #openstack-ironic | 07:13 | |
| *** meghal has joined #openstack-ironic | 07:19 | |
| *** meghal1 has quit IRC | 07:19 | |
| dtantsur | Morning Ironic! | 07:21 |
| mrda | hi dtantsur | 07:21 |
| dtantsur | o/ | 07:22 |
| *** achanda has quit IRC | 07:23 | |
| *** achanda has joined #openstack-ironic | 07:26 | |
| *** spandhe has joined #openstack-ironic | 07:27 | |
| *** rameshg87 has joined #openstack-ironic | 07:34 | |
| rameshg87 | good afternoon ironic | 07:34 |
| *** andreykurilin_ has joined #openstack-ironic | 07:34 | |
| *** chlong has quit IRC | 07:34 | |
| *** achanda has quit IRC | 07:37 | |
| *** yog__ has joined #openstack-ironic | 07:37 | |
| *** takadayuiko has joined #openstack-ironic | 07:38 | |
| *** subscope has joined #openstack-ironic | 07:41 | |
| dtantsur | rameshg87, g'afternoon | 07:45 |
| rameshg87 | dtantsur: morning .. | 07:46 |
| *** jistr has quit IRC | 07:50 | |
| openstackgerrit | Merged stackforge/ironic-discoverd: Implement basic checks in functest/devstack-test.sh https://review.openstack.org/169776 | 07:52 |
| openstackgerrit | Merged stackforge/ironic-discoverd: Document ipmi_address in HTTP API https://review.openstack.org/170520 | 07:56 |
| *** tiagogomes has joined #openstack-ironic | 08:02 | |
| *** andreykurilin_ has quit IRC | 08:03 | |
| dtantsur | hmm, looks like devstack is broken, e.g. https://review.openstack.org/#/c/170439/ | 08:06 |
| *** lucasagomes has joined #openstack-ironic | 08:08 | |
| *** jistr has joined #openstack-ironic | 08:09 | |
| *** lazy_prince has quit IRC | 08:11 | |
| *** degorenko has joined #openstack-ironic | 08:13 | |
| *** meghal has quit IRC | 08:14 | |
| *** romcheg has joined #openstack-ironic | 08:15 | |
| *** mgoddard has joined #openstack-ironic | 08:16 | |
| *** spandhe has quit IRC | 08:18 | |
| pas-ha | yep, https://bugs.launchpad.net/tempest/+bug/1440984 | 08:27 |
| openstack | Launchpad bug 1440984 in tempest "AttributeError: 'InstallRequirement' when running update.py" [Undecided,New] | 08:27 |
| *** mrda is now known as mrda-away | 08:28 | |
| *** ndipanov has joined #openstack-ironic | 08:30 | |
| *** yuanying has quit IRC | 08:40 | |
| dtantsur | ouch | 08:40 |
| *** killer_prince has joined #openstack-ironic | 08:44 | |
| *** killer_prince is now known as lazy_prince | 08:44 | |
| *** lxsli has left #openstack-ironic | 08:49 | |
| *** killer_prince has joined #openstack-ironic | 08:49 | |
| *** lazy_prince has quit IRC | 08:51 | |
| *** killer_prince is now known as lazy_prince | 08:51 | |
| openstackgerrit | sandhya proposed openstack/ironic: Implement Chassis Level Discovery Tool https://review.openstack.org/158577 | 08:55 |
| *** lazy_prince has quit IRC | 08:55 | |
| *** killer_prince has joined #openstack-ironic | 08:57 | |
| *** killer_prince is now known as lazy_prince | 08:57 | |
| *** edcragg has joined #openstack-ironic | 08:58 | |
| ramineni | pas-ha: yes, was facing this issue on devstack , about to google it :) thanks for the link :) | 08:58 |
| *** killer_prince has joined #openstack-ironic | 09:01 | |
| lucasagomes | gate is broken? :-( | 09:01 |
| *** lazy_prince has quit IRC | 09:02 | |
| *** killer_prince is now known as lazy_prince | 09:02 | |
| dtantsur | yep | 09:03 |
| *** dtantsur is now known as dtantsur|brb | 09:03 | |
| *** pradipta has quit IRC | 09:11 | |
| *** persia_ is now known as persia | 09:14 | |
| lucasagomes | rameshg87, hi, for 170117 yuriyz gave the idea of only doing a reboot if the pxe_ssh driver is used | 09:15 |
| *** meghal has joined #openstack-ironic | 09:15 | |
| lucasagomes | I like it because it's a more self-contained change, so I will update the patch to check for the driver and do that | 09:16 |
| lucasagomes | you can see any problem with it? | 09:16 |
| rameshg87 | lucasagomes: hi | 09:17 |
| rameshg87 | lucasagomes: fine with me, but checking and deciding action for a driver in iscsi_deploy, is it fine ? | 09:17 |
| yuriyz | morning lucasagomes rameshg87 I think also that Ironic should have "total control" over nodes power in the Liberty | 09:19 |
| lucasagomes | rameshg87, yeah that's the part that looks a bit odd | 09:19 |
| lucasagomes | yuriyz, +1 | 09:19 |
| lucasagomes | rameshg87, I was about to add a NOTE explaning why we do that | 09:19 |
| yuriyz | you can add comment w todo inline | 09:19 |
| lucasagomes | like if 'ssh' in task.node.driver: | 09:20 |
| lucasagomes | yuriyz, fair enough | 09:20 |
| lucasagomes | rameshg87, yuriyz or perhaps check if the power interface is ssh interface | 09:20 |
| *** meghal has quit IRC | 09:20 | |
| rameshg87 | yeah, fine with me | 09:21 |
| yuriyz | +1 | 09:21 |
| rameshg87 | anyway it's a todo :) | 09:21 |
| rameshg87 | for pxe_ssh, virsh stores the settings and applies it after reboot. but for virtual box cannot do that :( | 09:22 |
| lucasagomes | :-( | 09:23 |
| lucasagomes | rameshg87, so it doesn't gets applied on vbox? | 09:23 |
| lucasagomes | rameshg87, we will end up having to set the boot device in the power_on() method | 09:23 |
| lucasagomes | not only for vbox, but for all hardwares that doesn't save the boot device option | 09:24 |
| *** pelix has joined #openstack-ironic | 09:24 | |
| *** spandhe has joined #openstack-ironic | 09:25 | |
| rameshg87 | lucasagomes: yeah that's much better :) | 09:28 |
| rameshg87 | lucasagomes: vbox needs it. | 09:28 |
| lucasagomes | yeaj | 09:28 |
| lucasagomes | yeah* | 09:28 |
| *** rameshg871 has joined #openstack-ironic | 09:38 | |
| *** rameshg87 has quit IRC | 09:40 | |
| *** athomas has quit IRC | 09:43 | |
| *** rameshg871 has quit IRC | 09:43 | |
| *** rameshg87 has joined #openstack-ironic | 09:50 | |
| *** stendulker has joined #openstack-ironic | 09:52 | |
| *** firl has quit IRC | 09:52 | |
| *** foexle has joined #openstack-ironic | 09:52 | |
| *** firl has joined #openstack-ironic | 09:52 | |
| *** rameshg871 has joined #openstack-ironic | 09:52 | |
| *** athomas has joined #openstack-ironic | 09:53 | |
| *** rameshg87 has quit IRC | 09:55 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: FIX Vm stuck when deploying with pxe_ssh + local boot https://review.openstack.org/170117 | 09:56 |
| *** naohirot has quit IRC | 09:57 | |
| *** chlong has joined #openstack-ironic | 10:07 | |
| *** dtantsur|brb is now known as dtantsur | 10:07 | |
| *** meghal has joined #openstack-ironic | 10:16 | |
| *** takadayuiko has quit IRC | 10:20 | |
| *** meghal has quit IRC | 10:20 | |
| *** Nisha has quit IRC | 10:25 | |
| *** viktors is now known as viktors|lunch | 10:27 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add append_processing_hooks options https://review.openstack.org/170535 | 10:27 |
| *** jistr has quit IRC | 10:30 | |
| *** romcheg has quit IRC | 10:37 | |
| openstackgerrit | Ramakrishnan G proposed openstack/ironic: Check status of bootloader installation for DIB ramdisk https://review.openstack.org/169419 | 10:39 |
| *** sandhya has quit IRC | 10:42 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add append_processing_hooks options https://review.openstack.org/170535 | 10:46 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add an option to always store ramdisk logs https://review.openstack.org/171122 | 10:46 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add an option to always store ramdisk logs https://review.openstack.org/171122 | 10:49 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add append_processing_hooks options https://review.openstack.org/170535 | 10:49 |
| openstackgerrit | Anusha Ramineni proposed openstack/ironic: iLO driver documentation for node cleaning https://review.openstack.org/169162 | 10:49 |
| *** jistr has joined #openstack-ironic | 10:51 | |
| *** jistr is now known as jistr|biab | 10:52 | |
| dtantsur | lucasagomes, hi! may I ask you review a couple more discoverd patches as Imre seems still on PTO? https://review.openstack.org/#/c/171122/ and https://review.openstack.org/#/c/170535/ | 10:57 |
| *** meghal has joined #openstack-ironic | 10:59 | |
| lucasagomes | dtantsur, sure will do in a sec | 11:01 |
| *** meghal has quit IRC | 11:04 | |
| *** romcheg has joined #openstack-ironic | 11:04 | |
| *** ramineni has quit IRC | 11:04 | |
| lucasagomes | dtantsur, what's the motivation on append_processing_hooks ? | 11:09 |
| lucasagomes | I mean, if someone downstream have a custom processing_hooks I think they want to not change it when discoverd changes | 11:09 |
| lucasagomes | that prevents discoverd downstream from running something they don't want to | 11:09 |
| lucasagomes | or... lemme rephrase, not that they don't want to | 11:10 |
| dtantsur | lucasagomes, not necessary, e.g. we only need to insert a couple of plugins | 11:10 |
| dtantsur | actually the default set of plugins is something you very rarely want to change, if ever | 11:10 |
| lucasagomes | but they don't want to run more or less hooks when they update discoverd if they have a custom processing_hooks, they might just want to run exactly the same | 11:10 |
| lucasagomes | right | 11:10 |
| dtantsur | lucasagomes, they can still use just 'processing_hooks' and override everything like before. But I'd like us (RH) to work with default discoverd + plugins | 11:11 |
| lucasagomes | dtantsur, I find it complicated to have 2 lists of hooks that will run | 11:11 |
| dtantsur | another option would be to have 'processing_hooks' that acts as append, and also some knd of blacklist | 11:12 |
| lucasagomes | right, I don't know. I kinda like the whitelist more | 11:13 |
| lucasagomes | as is right now | 11:13 |
| lucasagomes | it feels safer if you want to have control over ur enviroment, so that new things won't run unless you explicity tell it to run | 11:13 |
| lucasagomes | kinda liek enabled_drivers for ironic | 11:13 |
| *** athomas has quit IRC | 11:13 | |
| lucasagomes | and if one wants to do the default + plugins... perhaps they should script it? | 11:14 |
| TheJulia | good morning | 11:14 |
| dtantsur | lucasagomes, enabled_drivers in Ironic is a completely different beast, e.g. they all are independent | 11:14 |
| * lucasagomes thinks | 11:14 | |
| dtantsur | TheJulia, morning | 11:14 |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Add retry logic to _exec_ipmitool https://review.openstack.org/168120 | 11:14 |
| lucasagomes | TheJulia, morning | 11:15 |
| lucasagomes | dtantsur, right, anyway it's a white list | 11:16 |
| dtantsur | lucasagomes, e.g. if I ever split essential plugin into 2, people will get broken. that can't happen with Ironic. but well, I'm starting to agree that I don't have compelling enough reason for this option... | 11:17 |
| * TheJulia caffinates | 11:17 | |
| lucasagomes | dtantsur, the good thing about this option is that you still can have the old behavior with processing_hooks only | 11:17 |
| lucasagomes | if u just leave the append_processing_hooks empty | 11:18 |
| dtantsur | yep, it's pure opt-in | 11:18 |
| * devananda does email, goes back to bed | 11:18 | |
| lucasagomes | so yeah, I don't have a strong opnion about it either. It just looks a bit odd to have 2 options to enable hooks | 11:18 |
| lucasagomes | dtantsur, plus the other that the hooks are specified matters right? | 11:18 |
| dtantsur | lucasagomes, yep | 11:18 |
| *** athomas has joined #openstack-ironic | 11:19 | |
| lucasagomes | so sometimes one wnat to keep the default hooks, but he has a custom hook that should run before everything else | 11:19 |
| lucasagomes | and that option won't actually help him | 11:19 |
| lucasagomes | because it just appends to the end of the list | 11:19 |
| lucasagomes | devananda, g'night | 11:19 |
| dtantsur | lucasagomes, actually I was thinking about prepend_processing_hooks too :D | 11:22 |
| lucasagomes | heh | 11:22 |
| dtantsur | but yeah, maybe I'm overengineering things... | 11:22 |
| lucasagomes | dtantsur, yeah maybe we should brainstorm see what would be a good way to do it | 11:22 |
| lucasagomes | I think that, if one wants to change the hooks etc he could script it | 11:23 |
| lucasagomes | to apply the configuration with the hooks and order he wants | 11:23 |
| stendulker | lucasgomes: Hi | 11:25 |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: Add an option to always store ramdisk logs https://review.openstack.org/171122 | 11:26 |
| stendulker | lucasgomes: Just found one thing about https://review.openstack.org/#/c/168666 | 11:26 |
| dtantsur | lucasagomes, updated ^^^ | 11:26 |
| dtantsur | lucasagomes, and abandoned append_* for now | 11:26 |
| stendulker | If youc can workflow-1 this review so that I could experimet that option and see if the localboot + iscsi_ilo works | 11:26 |
| stendulker | lucagomes: to block this submittal from going through | 11:27 |
| stendulker | lucasgomes: I will get back to you with my findings shortly. | 11:27 |
| lucasagomes | stendulker, hi yes sure! | 11:28 |
| stendulker | lucasgomes: thank you :) | 11:28 |
| lucasagomes | stendulker, done | 11:29 |
| * lucasagomes will go get some lunch | 11:30 | |
| *** lucasagomes is now known as lucas-hungry | 11:30 | |
| *** viktors|lunch is now known as viktors | 11:33 | |
| *** enikanorov has quit IRC | 11:42 | |
| *** enikanorov has joined #openstack-ironic | 11:43 | |
| tiagogomes | hi, I am having a connectivity problem. The iscsiadm process started by Ironic doesn't seem to run in the namespace where Neutron provided an IP address to the baremetal machine, so Ironic can't communicate with the baremetal machine | 11:46 |
| tiagogomes | I am using a dedicated network for the baremetal provisioning | 11:47 |
| tiagogomes | Do I need to set anything extra? | 11:47 |
| *** chlong has quit IRC | 11:49 | |
| *** jistr|biab is now known as jistr | 11:49 | |
| *** chlong has joined #openstack-ironic | 11:51 | |
| *** trown|outttypeww is now known as trown | 11:53 | |
| *** stendulker has quit IRC | 11:54 | |
| *** lazy_prince has quit IRC | 11:58 | |
| *** coolsvap is now known as coolsvap|afk | 11:59 | |
| *** meghal has joined #openstack-ironic | 11:59 | |
| *** chenglch|2 has joined #openstack-ironic | 12:00 | |
| *** kalpase has left #openstack-ironic | 12:00 | |
| openstackgerrit | Merged stackforge/ironic-discoverd: Add an option to always store ramdisk logs https://review.openstack.org/171122 | 12:03 |
| *** chenglch has quit IRC | 12:03 | |
| *** meghal has quit IRC | 12:04 | |
| * rameshg871 goes home | 12:12 | |
| *** rameshg871 has left #openstack-ironic | 12:12 | |
| *** dprince has joined #openstack-ironic | 12:20 | |
| *** dtantsur_ has joined #openstack-ironic | 12:21 | |
| *** dtantsur has quit IRC | 12:21 | |
| *** dtantsur_ is now known as dtantsur | 12:21 | |
| *** alexpilotti has joined #openstack-ironic | 12:22 | |
| *** pcrews has quit IRC | 12:24 | |
| *** killer_prince has joined #openstack-ironic | 12:25 | |
| *** killer_prince is now known as lazy_prince | 12:25 | |
| *** lucas-hungry is now known as lucasagomes | 12:26 | |
| *** killer_prince has joined #openstack-ironic | 12:28 | |
| *** lazy_prince has quit IRC | 12:30 | |
| *** keekz has quit IRC | 12:30 | |
| *** keekz has joined #openstack-ironic | 12:32 | |
| *** killer_prince has quit IRC | 12:33 | |
| *** Haomeng|2 has joined #openstack-ironic | 12:34 | |
| *** killer_prince has joined #openstack-ironic | 12:36 | |
| *** Haomeng has quit IRC | 12:36 | |
| *** killer_prince is now known as lazy_prince | 12:36 | |
| *** saripurigopi has quit IRC | 12:38 | |
| *** killer_prince has joined #openstack-ironic | 12:48 | |
| *** lazy_prince has quit IRC | 12:49 | |
| *** killer_prince is now known as lazy_prince | 12:49 | |
| *** kkoski has joined #openstack-ironic | 12:54 | |
| *** killer_prince has joined #openstack-ironic | 12:56 | |
| *** lazy_prince has quit IRC | 12:58 | |
| *** killer_prince is now known as lazy_prince | 12:58 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: FIX Vm stuck when deploying with pxe_ssh + local boot https://review.openstack.org/170117 | 12:59 |
| *** meghal has joined #openstack-ironic | 13:00 | |
| *** meghal has quit IRC | 13:05 | |
| *** openstackgerrit has quit IRC | 13:07 | |
| *** openstackgerrit has joined #openstack-ironic | 13:07 | |
| jroll | morning all :) | 13:23 |
| lucasagomes | jroll, morning | 13:23 |
| jroll | \o | 13:24 |
| *** Marga_ has joined #openstack-ironic | 13:24 | |
| jroll | welcome back :P | 13:24 |
| *** Marga_ has quit IRC | 13:24 | |
| *** Marga_ has joined #openstack-ironic | 13:25 | |
| dtantsur | jroll, morning! | 13:25 |
| jroll | hiya dtantsur | 13:25 |
| lucasagomes | thanks :D | 13:27 |
| lucasagomes | yeah bank holidays r good ;-) | 13:27 |
| jroll | heh | 13:29 |
| *** lazy_prince has quit IRC | 13:29 | |
| *** chenglch|2 has quit IRC | 13:30 | |
| *** ndipanov has quit IRC | 13:31 | |
| *** killer_prince has joined #openstack-ironic | 13:35 | |
| *** killer_prince is now known as lazy_prince | 13:35 | |
| *** jgrimm is now known as zz_jgrimm | 13:37 | |
| dtantsur | folks in #tripleo are getting "Not Acceptable" in response to node-list: http://paste.openstack.org/show/199450/. How could that happen? Oo | 13:41 |
| jroll | wat | 13:42 |
| *** lazy_prince has quit IRC | 13:46 | |
| lucasagomes | dtantsur, odd, it's something related to microversioning apparently | 13:47 |
| dtantsur | yep... | 13:47 |
| NobodyCam | Good mornign Ironicers | 13:48 |
| lucasagomes | NobodyCam, morning | 13:48 |
| jroll | I can't see in the code where this could possibly happen | 13:49 |
| NobodyCam | hey hey lucasagomes dtantsur and jroll | 13:49 |
| jroll | morning NobodyCam :) | 13:49 |
| NobodyCam | :) | 13:49 |
| dtantsur | NobodyCam, morning | 13:49 |
| *** chlong has quit IRC | 13:51 | |
| openstackgerrit | Merged openstack/ironic: Add missing mock for test_create_cleaning_ports_fail https://review.openstack.org/170999 | 13:57 |
| *** mtanino has joined #openstack-ironic | 14:00 | |
| jroll | turns out the tripleo thing was ironic at max 1.4, new client (which defaults to 1.6) | 14:01 |
| jroll | I'm really sad the client defaults to a specific version | 14:01 |
| *** meghal has joined #openstack-ironic | 14:01 | |
| lucasagomes | +1 | 14:01 |
| dtantsur | jroll, yeah, it does not sound right | 14:01 |
| *** ndipanov has joined #openstack-ironic | 14:02 | |
| dtantsur | until we have version negotiation, that is | 14:02 |
| *** rloo has joined #openstack-ironic | 14:03 | |
| jroll | we do have version negotiation, it's called 'latest' | 14:05 |
| jroll | :P | 14:05 |
| *** meghal has quit IRC | 14:05 | |
| openstackgerrit | Dmitry Tantsur proposed stackforge/ironic-discoverd: README: move out release notes and start troubleshooting guide https://review.openstack.org/171211 | 14:06 |
| NobodyCam | brb | 14:07 |
| *** Marga_ has quit IRC | 14:07 | |
| *** ndipanov has quit IRC | 14:08 | |
| *** ndipanov has joined #openstack-ironic | 14:12 | |
| *** killer_prince has joined #openstack-ironic | 14:15 | |
| *** killer_prince is now known as lazy_prince | 14:15 | |
| *** zz_jgrimm is now known as jgrimm | 14:23 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic-specs: Updated Ironic Microversions Spec to fix minor errors. https://review.openstack.org/171217 | 14:24 |
| *** Sukhdev has joined #openstack-ironic | 14:25 | |
| *** Nisha has joined #openstack-ironic | 14:26 | |
| jroll | it just occurred to me that MANAGEABLE -> AVAILABLE invokes a clean cycle | 14:35 |
| jroll | which kinda sucks, I wish that was optional | 14:36 |
| jroll | so you could do things like caching in manageable | 14:36 |
| jroll | also, couldn't a clean cycle annihilate things that a zap cycle might do? | 14:36 |
| lucasagomes | jroll, it should be optional no? | 14:37 |
| lucasagomes | I mean from manageable you can invoke clean directly from the api | 14:37 |
| jroll | lucasagomes: between manageable and available specifically? | 14:37 |
| lucasagomes | but just moving states should be optional | 14:37 |
| lucasagomes | yeah | 14:37 |
| jroll | hm, need to look | 14:37 |
| lucasagomes | IIRC that's what the machine state spec says | 14:37 |
| * lucasagomes looks | 14:37 | |
| jroll | nope | 14:38 |
| lucasagomes | state machine* | 14:38 |
| lucasagomes | http://specs.openstack.org/openstack/ironic-specs/specs/kilo/new-ironic-state-machine.html | 14:38 |
| jroll | https://github.com/openstack/ironic/blob/master/ironic/common/states.py#L272-273 | 14:38 |
| *** devlaps has quit IRC | 14:38 | |
| lucasagomes | oh yeah it's trhought clenaing | 14:38 |
| jroll | code says no, spec says "AVAILABLE (through CLEANING) via the provide API call." | 14:38 |
| jroll | right | 14:38 |
| lucasagomes | yeah | 14:39 |
| jroll | :/ | 14:39 |
| rloo | morning ironic'ers | 14:39 |
| NobodyCam | good morning rloo :) | 14:39 |
| jroll | heya rloo :) | 14:39 |
| jroll | lucasagomes: this makes me sad :/ | 14:39 |
| *** Nisha has quit IRC | 14:39 | |
| lucasagomes | jroll, why is that? | 14:39 |
| rloo | jroll: yeah, I think it is coded to manageable -> clean -> available | 14:39 |
| jroll | lucasagomes: well, I just had a thought about caching images in manage and/or zap | 14:40 |
| jroll | but cleaning will blow that away | 14:40 |
| jroll | I'm sure there are other cases where that's also true | 14:40 |
| rloo | jroll: i tried to ping JayF and JoshNang about that last Thursday. The doc doesn't mention that. and i think there was something else but i don't recall now. | 14:40 |
| lucasagomes | jroll, right yeah | 14:40 |
| lucasagomes | rloo, morning | 14:41 |
| dtantsur | morning, rloo | 14:41 |
| rloo | jroll: oh yeah. the doc mentions looking at clean-step or something like that, to see what step is being done, but that isn't (wasn't last week) being exposed in the api | 14:41 |
| jroll | rloo: what doc doesn't mention what? | 14:41 |
| rloo | jroll: JayF's cleaning documentation | 14:42 |
| BadCub | morning folks | 14:42 |
| NobodyCam | morning BadCub | 14:42 |
| jroll | rloo: doesn't mention the manageable -> available stuff? | 14:42 |
| rloo | jroll: sorry, just starting to catch up on stuff since last Thursday. | 14:42 |
| rloo | jroll: yeah, the cleaning doc doesn't mention that. | 14:42 |
| pshige | morning | 14:42 |
| rloo | jroll: I'll try to review the cleaning doc today. | 14:42 |
| NobodyCam | morning pshige | 14:42 |
| jlvillal | dtantsur: My wife just got her tourist visa for Canada. It took about 22 days. | 14:43 |
| NobodyCam | morning jlvillal | 14:43 |
| jroll | rloo: ah, right, ok | 14:43 |
| dtantsur | jlvillal, congrats! I wonder how long it's gonna take to get my business visa... well, 22 days should be fine :) | 14:43 |
| jlvillal | Good morning: NobodyCam, jroll, rloo, dtantsur, lucasagomes, pshige, and everybody else :) | 14:43 |
| dtantsur | and morning :) | 14:43 |
| jroll | heya jlvillal :) | 14:44 |
| *** mgoddard1 has joined #openstack-ironic | 14:44 | |
| lucasagomes | pshige, jlvillal good morning | 14:44 |
| jlvillal | dtantsur: Yeah the visa center told her 50/50 chance. Luckily it worked out :) | 14:44 |
| pshige | 22 days! | 14:44 |
| jlvillal | dtantsur: I would assume business visa will be more likely to be approved. | 14:44 |
| jroll | lucasagomes: at any rate, I'm debating trying to fix this asap or next cycle, regardless I'll probably end up hacking it downstream | 14:44 |
| *** mgoddard has quit IRC | 14:44 | |
| lucasagomes | jroll, asap == before summit? | 14:45 |
| dtantsur | jlvillal, oh yeah... | 14:45 |
| jroll | lucasagomes: asap == before rc1 | 14:45 |
| lucasagomes | perhaps it's something we could discuss in the summit? | 14:45 |
| lucasagomes | jroll, oh | 14:45 |
| jroll | but like, idk if it's a problem for others | 14:45 |
| dtantsur | I guess our scripts assume that "provide" works quickly | 14:46 |
| lucasagomes | you have a valid use case for that, so it worth fixing | 14:46 |
| lucasagomes | the problem before rc1 is that, it looks more like a feature | 14:46 |
| lucasagomes | not a bug] | 14:47 |
| pshige | NobodyCam, jlvillal, lucasgomes good morning | 14:47 |
| NobodyCam | :) | 14:47 |
| jroll | lucasagomes: yeah, indeed | 14:47 |
| jroll | dtantsur: when cleaning can fall back to shred :P | 14:47 |
| dtantsur | oh | 14:48 |
| devananda | morning, all | 14:48 |
| jroll | heya devananda | 14:49 |
| jroll | devananda: we had a user this morning upgrade their client, and everything broke because their ironic only supported up to 1.4 | 14:49 |
| jroll | so that's fun. | 14:49 |
| devananda | erm | 14:50 |
| devananda | "everything" ? | 14:50 |
| jroll | yes | 14:50 |
| jroll | like node-list. | 14:50 |
| NobodyCam | morning devananda | 14:51 |
| jroll | because the client passes 1.6 by default | 14:51 |
| TheJulia | good morning | 14:51 |
| lucasagomes | yeah :/ | 14:51 |
| dtantsur | devananda, morning | 14:51 |
| jroll | hey TheJulia :) | 14:51 |
| jlvillal | morning devananda TheJulia | 14:51 |
| pshige | morning devananda TheJulia | 14:52 |
| lucasagomes | the bright side is that the Ironic behavior was correct, user was asking for 1.6 and Ironic failed with 406 which seems the right thing to do if you can't fulfill the request | 14:52 |
| devananda | oh. right. because we didn't actually implement auto-negotiate-down | 14:53 |
| devananda | arrrrgh | 14:53 |
| *** mgoddard1 has quit IRC | 14:53 | |
| * devananda quits | 14:54 | |
| lucasagomes | perhaps we should document the upgrade process | 14:54 |
| * devananda quits typing to go make coffee | 14:54 | |
| lucasagomes | first update Ironic, then client etc... | 14:54 |
| *** mgoddard has joined #openstack-ironic | 14:54 | |
| jroll | lucasagomes: I tend to think the latest client should work with most ironic releases | 14:54 |
| lucasagomes | jroll, right, that was the case when we used the minimum version as the default | 14:55 |
| lucasagomes | because that's backward supported | 14:55 |
| lucasagomes | but having a fixed version is hard to have it | 14:55 |
| lucasagomes | perhaps we should rethink about the "latest" | 14:55 |
| jroll | lucasagomes: or have the client discover the best version | 14:56 |
| jroll | which I guess is "pass latest" | 14:57 |
| lucasagomes | heh yeah | 14:57 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add maintenance check before call do_node_deploy https://review.openstack.org/170149 | 14:57 |
| lucasagomes | the latest seems to be the way to do that | 14:57 |
| *** rloo_ has joined #openstack-ironic | 14:58 | |
| *** rloo has quit IRC | 14:58 | |
| devananda | very simply, the latest client should always Just Work (tm) | 14:59 |
| devananda | a client relese should never ever break anyone | 14:59 |
| jroll | right | 15:00 |
| jroll | but that just happened :/ | 15:00 |
| devananda | anyone fixing it yet? | 15:00 |
| lucasagomes | devananda, right, we are trying to figure what would be the right fix for it | 15:01 |
| lucasagomes | devananda, can you remember why pointing the version to "latest" wasn't used in the first place? | 15:01 |
| rloo_ | didn't we already go through that exercise? (behavior of client + server)? there's a spec... | 15:01 |
| *** meghal has joined #openstack-ironic | 15:02 | |
| rloo_ | lucasagomes: http://specs.openstack.org/openstack/ironic-specs/specs/kilo/api-microversions.html | 15:02 |
| * lucasagomes looks | 15:02 | |
| devananda | yah. there's a spec. we just didn't actually implement it all yet | 15:02 |
| devananda | the right thing to do is for hte client to receive the 406, lookat the header, then re-initiate the request for version 1.4 | 15:03 |
| *** Sukhdev has quit IRC | 15:04 | |
| devananda | and cache 'server {address} supports up to {version} so dont re-negotiate again for another {period}' | 15:04 |
| rloo_ | devananda: hmm. the spec doesn't mention that http://specs.openstack.org/openstack/ironic-specs/specs/kilo/api-microversions.html#use-case-6-new-client-new-ironic-unsupported-client-version | 15:04 |
| devananda | rloo_: one further -- http://specs.openstack.org/openstack/ironic-specs/specs/kilo/api-microversions.html#use-case-7a-new-client-new-ironic-negotiated-version-not-user-specified | 15:05 |
| BadCub | heya NobodyCam devananda jroll JayF dtantsur lucasagomes, you guys up for a Spec review meeting early next week to start approving specs for L? | 15:05 |
| devananda | Ironic responds with a 406 Not Acceptable, along with the -Min- and -Max- headers that it can support (in this case 1.1 and 1.10) | 15:05 |
| devananda | The client should transparently proceed, having negotiated that both client and server will use v1.10. The client should also cache this microversion, so that subsequent attempts do not need to renegotiate microversions. | 15:05 |
| dtantsur | BadCub, o/ | 15:05 |
| jroll | BadCub: sure, let me just submit some specs ;D | 15:06 |
| devananda | ^^ that ^^ is what we didn't implement. well one of the things. | 15:06 |
| lucasagomes | BadCub, +1 | 15:06 |
| *** meghal has quit IRC | 15:06 | |
| devananda | but that's what we NEED right now | 15:06 |
| rloo_ | devananda: ah yes. hmm. we should probably implement it (and the other stuff in that spec). soon... | 15:06 |
| devananda | rloo_: indeed | 15:06 |
| devananda | what do we do in the meantime? pin to v1.1 again? | 15:06 |
| BadCub | awesome guys :-) I'll throw out a couple of days and see what looks best for everyone :) | 15:07 |
| lucasagomes | devananda, it sounds fine, the only bit I worry is about caching the version | 15:07 |
| lucasagomes | like how the client would actually cache it | 15:08 |
| rloo_ | I can't keep track of what breaks if we do whatever. dtantsur said something broke discoverd w/o the latest client that was released, right? | 15:08 |
| jlvillal | lucasagomes: ~/.ironicclient.cfg ? | 15:08 |
| lucasagomes | hmm | 15:09 |
| * lucasagomes has hard times with cache invalidation and stuff | 15:09 | |
| dtantsur | rloo_, well, with new client our scripts broke | 15:09 |
| rloo_ | dtantsur: oh :-( | 15:09 |
| jlvillal | lucasagomes: Though I would think occasionally the client should re-check in case Ironic gets upgraded. | 15:09 |
| NobodyCam | BadCub: sounds good here | 15:09 |
| lucasagomes | I wonder if a warning message or something indicating that the default version is higher than the server isn't a better way to signlize the problem | 15:10 |
| devananda | dtantsur: broke how? | 15:10 |
| devananda | lucasagomes: version negotiation should be transparent to the end user | 15:10 |
| devananda | like they should not get a warning or log message or -anything- | 15:10 |
| devananda | if both client and server support X.YY, then they should just use it | 15:10 |
| dtantsur | devananda, I've sent that to the ML before weekend, you even answered :) nodes started appearing in AVAILABLE state and discoverd refused to start inspection on them | 15:11 |
| lucasagomes | devananda, right, I agree. But it seems that for version negociation we should use the "latest" key as the version | 15:11 |
| lucasagomes | so it will figure what is the latest supported version by the server | 15:11 |
| devananda | dtantsur: ah, right. just checking if this was something else. also - not done with my first cup of coffee yet :) | 15:12 |
| dtantsur | :) | 15:12 |
| lucasagomes | having a fixed version in one of the ends makes it a bit complicated | 15:12 |
| *** pcrews has joined #openstack-ironic | 15:14 | |
| *** stendulker has joined #openstack-ironic | 15:17 | |
| openstackgerrit | Shivanand Tendulker proposed openstack/ironic: boot_mode is overwritten in node properties https://review.openstack.org/167952 | 15:17 |
| *** ijw has joined #openstack-ironic | 15:19 | |
| devananda | did we do anything in the client (other than this) that would prevent it from working with older versions of the server right now? | 15:19 |
| jroll | not that I'm aware of | 15:20 |
| NobodyCam | romcheg: Happy Birthday! | 15:21 |
| romcheg | NobodyCam: Thank you! | 15:21 |
| NobodyCam | :) | 15:21 |
| BadCub | Happy B-Day romcheg :) | 15:22 |
| * jroll passes romcheg a beer and says happy birthday | 15:24 | |
| romcheg | Thank you all guys! | 15:24 |
| devananda | oh right, take a look at 162834 | 15:24 |
| devananda | we added new columns to the CLI output that will be empty if talking to old servers | 15:25 |
| devananda | which isn't the same as broken, but we could hide them ... | 15:25 |
| romcheg | Unfortunately I don't know where are your B-days but if you tell me that, I'll try to send you a postcard :) | 15:25 |
| devananda | romcheg: yea! happy bday! | 15:25 |
| romcheg | Btw, good evening/morning/whatever everyone :) | 15:25 |
| romcheg | devananda: thans! | 15:25 |
| romcheg | *thanks even :) | 15:25 |
| NobodyCam | :) | 15:26 |
| *** romcheg has left #openstack-ironic | 15:27 | |
| NobodyCam | hey hey romcheg before you head off to drink, (i always do on my birthday so I assume others do too) I was looking at https://bugs.launchpad.net/ironic/+bug/1327260 | 15:27 |
| openstack | Launchpad bug 1327260 in Ironic "Disk layout is not being verified on re-provisioning" [Medium,Triaged] | 15:27 |
| NobodyCam | \what driver is that with ? | 15:28 |
| *** krtaylor has quit IRC | 15:28 | |
| *** mtanino has quit IRC | 15:29 | |
| lucasagomes | NobodyCam, I believe the pxe_ipmitool one... That's about the rebuild | 15:32 |
| lucasagomes | apparently (I asked it a couple of days before here in the #) | 15:32 |
| NobodyCam | ahh TY lucasagomes | 15:33 |
| NobodyCam | :) | 15:33 |
| *** foexle has quit IRC | 15:33 | |
| rloo_ | devananda: wrt 162834, we should hide the new(er) fields | 15:34 |
| rloo_ | devananda: but it isn't urgent | 15:35 |
| NobodyCam | brb | 15:35 |
| jroll | lucasagomes: did you say our gate was broken this morning or am I making stuff up? | 15:36 |
| lucasagomes | jroll, it apparently was some devstack problem which is now fixed | 15:37 |
| jroll | ah, ok | 15:37 |
| jroll | lucasagomes: is that what hit 169419? | 15:37 |
| jroll | (and should we recheck) | 15:37 |
| * lucasagomes looks | 15:37 | |
| *** firl has quit IRC | 15:37 | |
| lucasagomes | jroll, yeah apparently on patch-set 13 | 15:38 |
| lucasagomes | before the recheck | 15:38 |
| jroll | ah ok | 15:38 |
| jroll | thanks :) | 15:39 |
| lucasagomes | np | 15:39 |
| *** devlaps has joined #openstack-ironic | 15:40 | |
| *** ijw_ has joined #openstack-ironic | 15:42 | |
| jroll | anybody else want eyes on https://review.openstack.org/#/c/169419/ before I +A? | 15:42 |
| jroll | devananda: ^ | 15:42 |
| * devananda clicks | 15:42 | |
| devananda | jroll: double check me on something | 15:43 |
| devananda | jroll: does'nt htis break if used with an old dib ramdisk? | 15:43 |
| jroll | devananda: when using localboot, yes | 15:44 |
| devananda | i get that local boot may not have been supported in juno, but | 15:44 |
| lucasagomes | devananda, only for localboot | 15:44 |
| jroll | so the ramdisk would have to be just new enough | 15:44 |
| devananda | what if i upgrade ironic and nova and start trying to use localboot | 15:44 |
| jroll | devananda: but today the current DIB ramdisk is broken, so | 15:44 |
| devananda | but i dont change the ramdisk -- how does the failure behave? | 15:44 |
| devananda | right. so its broken either way | 15:44 |
| devananda | gah | 15:44 |
| *** ijw has quit IRC | 15:44 | |
| lucasagomes | I think it will get sutck on the DEPLOYWAIT (cause it goes back to that state when it's installing the bootloader with this patch) | 15:45 |
| lucasagomes | and it will timeout eventually | 15:45 |
| lucasagomes | :/ | 15:45 |
| devananda | right | 15:45 |
| devananda | that's what i thought | 15:45 |
| *** ijw_ has quit IRC | 15:45 | |
| jroll | so feb 6 - apr 2 ramdisks work without this, and break with this. apr 2 - today ramdisks break without this, work with this. | 15:45 |
| devananda | can this patch check in some way that the ramdisk supports this option, and at least error helpfully if not? | 15:45 |
| *** ijw has joined #openstack-ironic | 15:46 | |
| jroll | devananda: I wonder if port 10000 isn't listening, if we could key on that. though that could also indicate some failure. | 15:46 |
| devananda | upgrades that break users like this are the worst upgrades | 15:46 |
| *** Marga_ has joined #openstack-ironic | 15:47 | |
| dtantsur | g'night folks | 15:48 |
| *** dtantsur is now known as dtantsur|afk | 15:48 | |
| NobodyCam | night dtantsur|afk | 15:48 |
| *** yog__ has quit IRC | 15:48 | |
| * lucasagomes thinks about the breakage | 15:48 | |
| *** romcheg has joined #openstack-ironic | 15:49 | |
| lucasagomes | I thought it was okish to land it since localboot was introduced in this realease | 15:49 |
| lucasagomes | and we didn't cut it yet | 15:49 |
| lucasagomes | but yeah would be nice to be backward compat even within the release | 15:49 |
| * lucasagomes thinks see if there's a way | 15:49 | |
| lucasagomes | problem is that with dib Ironic knows nothing about it when it boots :/ | 15:50 |
| lucasagomes | ipa at least does a lookup and heartbeat | 15:50 |
| lucasagomes | maybe we could pass someting in the pass_deploy_info | 15:50 |
| lucasagomes | if that specific key is passed we can assume it's a new ramdisk, and wait it to call the pass_bootloader_*... if not assume it's an old ramdisk | 15:51 |
| lucasagomes | (not sure how overcomplex that can get) | 15:51 |
| romcheg | devananda: Do you know whether there is a way to test whether all release jobs will work fine w/o actually make a release? | 15:55 |
| *** stendulker_ has joined #openstack-ironic | 15:58 | |
| *** stendulker has quit IRC | 15:58 | |
| *** alexpilotti has quit IRC | 15:59 | |
| devananda | jroll: I hae commented on that review. do what you think is right | 16:00 |
| *** spandhe has quit IRC | 16:01 | |
| *** meghal has joined #openstack-ironic | 16:02 | |
| jroll | devananda: hrm. | 16:03 |
| devananda | oh gawd, roofers are here today | 16:03 |
| * devananda starts getting ready to go into the office where it will be quieter | 16:03 | |
| jroll | devananda: I'm going to be pragmatic and approve | 16:03 |
| devananda | romcheg: nope | 16:03 |
| romcheg | devananda: thanks | 16:03 |
| *** meghal has quit IRC | 16:05 | |
| *** lazy_prince has quit IRC | 16:05 | |
| jroll | devananda: any plans to update the spec template for liberty? | 16:05 |
| jroll | or can I start writing things? | 16:06 |
| devananda | jroll: nothing comes to mind right now | 16:06 |
| jroll | cool, thanks | 16:06 |
| *** jcoufal has quit IRC | 16:07 | |
| *** killer_prince has joined #openstack-ironic | 16:07 | |
| *** killer_prince is now known as lazy_prince | 16:08 | |
| *** alexpilotti has joined #openstack-ironic | 16:11 | |
| *** Nisha has joined #openstack-ironic | 16:12 | |
| * lucasagomes spent at least one hour trying to figure out why his apache site wasn't working... tl;dr selinux | 16:13 | |
| jroll | man | 16:14 |
| jroll | selinux screwed me really hard setting up mpd the other day | 16:14 |
| NobodyCam | lucasagomes: ugggh :-p | 16:14 |
| lucasagomes | jroll, :( | 16:15 |
| JayF | rloo_: What exactly is wrong with the docs? | 16:15 |
| lucasagomes | yeah I created a rule 'sudo semanage fcontext -a -t httpd_sys_content_t "/httpboot(/.*)?"' | 16:15 |
| JoshNang | JayF: i think it was that we don't have anything about cleaning happening from managed->available | 16:15 |
| lucasagomes | but I had to apply it using restorecon | 16:15 |
| JoshNang | (which i agree, making optional is reasonable) | 16:15 |
| rloo_ | JayF: I should review them, but last week I noticed that it doesn't mention that cleaning is done from manageable -> available | 16:15 |
| jroll | lucasagomes: that's all jibberish to me :D | 16:15 |
| *** meghal has joined #openstack-ironic | 16:15 | |
| JayF | rloo_: for "when does cleaning happen" I mostly wanted to reference the state machine doc | 16:16 |
| lucasagomes | jroll, yeah, I'm not familiar with selinux at all | 16:16 |
| rloo_ | JayF: and there was something about getting the clean step but we don't yet expose the node's cleaning_step(or whatever it is called) | 16:16 |
| JayF | rloo_: basically the doc is "X happens during cleaning state" | 16:16 |
| lucasagomes | one of the first things I do after installing a OS is to disable it | 16:16 |
| JayF | rloo_: JoshNang: I think we expose /current/ clean_step, just not listing all the steps beforehand | 16:16 |
| lucasagomes | (personal use ofc) | 16:16 |
| lucasagomes | for rh I gotta deal with selinux | 16:16 |
| * lucasagomes weep | 16:16 | |
| jroll | lucasagomes: lol, +1 | 16:16 |
| *** meghal1 has joined #openstack-ironic | 16:17 | |
| rloo_ | JayF: looking at http://docs.openstack.org/developer/ironic/deploy/cleaning.html | 16:18 |
| devananda | BadCub: there are 7 nottin-progress bugs on the RC list | 16:18 |
| devananda | BadCub: I believe it is time to bump all 7 of them | 16:18 |
| rloo_ | JayF: there's nothing there about the state machine doc? | 16:18 |
| JayF | rloo_: did I not reference it? gah | 16:18 |
| openstackgerrit | Thierry Carrez proposed openstack/ironic: Open Liberty development https://review.openstack.org/171274 | 16:19 |
| BadCub | devananda: I was thinkng he same thing considering we plan on cutting RC1 in a day or two | 16:19 |
| rloo_ | JayF: and I don't think this is true: http://docs.openstack.org/developer/ironic/deploy/cleaning.html#what-clean-step-is-running | 16:19 |
| JayF | JoshNang: ^ | 16:19 |
| *** meghal has quit IRC | 16:20 | |
| JoshNang | hmm did that get dropped? | 16:21 |
| rloo_ | JayF, JoshNang: josh has a patch that was going to expose clean_step, ah yes, it was part of https://review.openstack.org/#/c/159322/ | 16:21 |
| JayF | JoshNang: yeah idk, that was a section you wrote :P | 16:21 |
| JayF | JoshNang: rloo_: That's for exposing a list of clean steps | 16:23 |
| BadCub | devananda: all have been bumped | 16:23 |
| JayF | JoshNang: rloo_: I thought getting the current clean step was already in there though | 16:23 |
| JayF | JoshNang: rloo_: Note the difference between doing a node-show and a call to cleaning/steps | 16:23 |
| JoshNang | JayF: i thought so too, as its part of the node object. | 16:23 |
| rloo_ | JayF, JoshNang: you have to add the clean_step thing to the node, and it isn't there. | 16:23 |
| *** alexpilotti_ has joined #openstack-ironic | 16:24 | |
| JoshNang | hrm | 16:24 |
| *** alexpilotti has quit IRC | 16:24 | |
| *** alexpilotti_ is now known as alexpilotti | 16:24 | |
| rloo_ | JoshNang: oh, see, I added a comment about it in your patch. around line 566: https://review.openstack.org/#/c/159322/10/ironic/api/controllers/v1/node.py | 16:26 |
| devananda | BadCub: ty | 16:26 |
| devananda | adam_g: https://bugs.launchpad.net/bugs/1433727 is assigned t oyou but does not show any change in progress to address it | 16:26 |
| openstack | Launchpad bug 1433727 in Ironic "partial upgrade not possible, 'reason': u'Unknown argument: "configdrive" (HTTP 400)'" [High,Confirmed] - Assigned to Adam Gandelman (gandelman-a) | 16:26 |
| BadCub | devananda: yw. Pad has been updated as well | 16:26 |
| *** jistr has quit IRC | 16:27 | |
| JayF | rloo_: we're both trying to wrangle devstack to see the actual behavior | 16:27 |
| JayF | rloo_: because we both thought it was already there and if not that's bad enough we should try to get a fix into L | 16:28 |
| jroll | JayF: these are all the fields the api can return https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L466-544 | 16:29 |
| devananda | jroll, JayF, dtantsur|afk, lucasagomes -- regarding releases of ipa and discoverd, ttx has just pointed out to me that there is now a guide for how we do stable releases of libraries and related projects -- http://specs.openstack.org/openstack/openstack-specs/specs/library-stable-branches.html | 16:29 |
| JayF | Ugh | 16:29 |
| JoshNang | i know i wrote the code to add the clean_step field, but i can't find it | 16:30 |
| jroll | devananda: cool | 16:30 |
| devananda | i'm not entirely sure that these qualify as libraries, though, so i'm not sure how much of that applies -- but it would be good for ya'll to be aware of | 16:30 |
| lucasagomes | devananda, oh cool, I will take a look | 16:30 |
| JayF | devananda: I'm honestly not a huge fan of having a stable branch for IPA :-/ | 16:30 |
| lucasagomes | devananda, yeah definetly | 16:30 |
| devananda | particularly since we will need a stable branch of both services that corresponds to stable/kilo of Iroinc | 16:30 |
| devananda | JayF: if IPA does backwards compat right, then yea, that's fine | 16:31 |
| jroll | JoshNang: https://review.openstack.org/#/c/155561/ | 16:31 |
| *** meghal1 has quit IRC | 16:32 | |
| rloo_ | jroll, JoshNang: that 155561 only adds clean_step to the sample, not to the node | 16:33 |
| jroll | rloo_: right | 16:33 |
| JoshNang | hrm. i mean i can write the patch to add it pretty quick. not sure if we want to bump the api version a couple days before rc | 16:33 |
| jlvillal | Anyone else see this occasional test error? http://paste.openstack.org/show/199616/ in File "ironic/tests/drivers/ilo/test_deploy.py", line 689, in test_tear_down_cleaning | 16:34 |
| JoshNang | i may have just had it local and not pushed it up after get_clean_steps got bumped | 16:34 |
| jlvillal | Second time I run the test it goes away. But I see it every day or so. | 16:34 |
| jroll | JoshNang: you'll need to write the patch very soon anyway :P | 16:34 |
| lucasagomes | devananda, yeah and we have the ironic-libs now too | 16:34 |
| lucasagomes | we need to start populating it | 16:34 |
| lucasagomes | JayF, btw re ^, on L, let's start working on the disk partitioning bits there? | 16:34 |
| JayF | JoshNang: I think we should file a bug about it and try to get it into K, but devananda is here and we can ask his opinion :) | 16:34 |
| devananda | ooh yes | 16:34 |
| rloo_ | JoshNang: so I think the cleaning step info is avail in driver_internal_info but maybe not obvious cuz it has all the clean steps or some of them or something like that. | 16:34 |
| *** kkoski has quit IRC | 16:35 | |
| JayF | devananda: exposing clean_step on the node object was missed during implementation of cleaning. Would you consider that a big enough bug to get in before the RC? From an operators' perspective; I say very yes :/ | 16:35 |
| JoshNang | rloo_: it has the remaining clean steps, so the first one is what's being executed | 16:35 |
| jroll | *current* clean step, to be clear | 16:35 |
| rloo_ | JoshNang: is that "good enough" for now? (although adding a microversion etc would be nice) | 16:36 |
| JayF | JoshNang: rloo_: driver_internal_info, is it returned by the API today? | 16:36 |
| jroll | JayF: yes | 16:36 |
| jroll | (given a particular version iirc) | 16:36 |
| JayF | then that's probably OK as long as we document it properly and add clean_step later? | 16:37 |
| rloo_ | jroll, JayF, JoshNang: driver_internal_info avail in 1.3+ | 16:37 |
| JayF | JoshNang: ^ you onboard with that? update docs to tell them to look at driver_internal_info? Plus I'll add the ref to the state machine diagram that apparently is missing | 16:37 |
| JoshNang | lemme verify quick | 16:37 |
| rloo_ | what bothers me is that even with our microversioning, you get cleaning w/o having to specify any microversion. | 16:37 |
| *** Marga_ has quit IRC | 16:38 | |
| jroll | cleaning isn't just an api :/ | 16:38 |
| *** Marga_ has joined #openstack-ironic | 16:38 | |
| rloo_ | jroll: I know. i just don't know what exactly is covered/should be micro-version controlled. | 16:38 |
| jroll | rloo_: the API semantics | 16:39 |
| *** devlaps has quit IRC | 16:39 | |
| rloo_ | jroll: so *only* API semantics | 16:39 |
| jroll | as far as I'm concerned | 16:39 |
| jroll | the way we talk to the API, the way it resopnds | 16:39 |
| JoshNang | yeah, driver_internal_info should work. | 16:40 |
| * jroll finds breakfast | 16:40 | |
| JoshNang | either way, i'll write the patch soon | 16:40 |
| JayF | Once my stack finishes stacking; I'll confirm behavior and update the docs | 16:40 |
| adam_g | devananda, re https://bugs.launchpad.net/bugs/1433727 i was going to push a doc update today | 16:41 |
| openstack | Launchpad bug 1433727 in Ironic "partial upgrade not possible, 'reason': u'Unknown argument: "configdrive" (HTTP 400)'" [High,Confirmed] - Assigned to Adam Gandelman (gandelman-a) | 16:41 |
| JoshNang | JayF: thank you! | 16:41 |
| JayF | JoshNang: ^ make sure when you expose clean_step the patch includes doc fixes, since I'm going to land a real-fix today | 16:41 |
| JoshNang | JayF: ++ | 16:41 |
| devananda | adam_g: BadCub: if it's just a doc change, let's untarget it from RC1, but do it soon anyway | 16:42 |
| devananda | doc updates go live on the site when ever they land, and dont strictly need to be in the stable/kilo branch | 16:42 |
| *** spandhe has joined #openstack-ironic | 16:42 | |
| BadCub | devananda: will do now | 16:42 |
| lucasagomes | jroll, devananda since the dependent patch was approved, mind taking a look at https://review.openstack.org/#/c/170117/ ? | 16:43 |
| *** Marga_ has quit IRC | 16:43 | |
| lucasagomes | it's quite small | 16:43 |
| tiagogomes | Hi, I am stuck as I can't make the ironic-conductor communicate with the baremetal machine to do the iscsi stuff. Is there any instructions on how to setup network properly? | 16:43 |
| *** Marga_ has joined #openstack-ironic | 16:49 | |
| NobodyCam | brb | 16:51 |
| *** lazy_prince has quit IRC | 16:51 | |
| *** kkoski has joined #openstack-ironic | 16:51 | |
| BadCub | brb | 16:52 |
| *** krtaylor has joined #openstack-ironic | 16:54 | |
| rloo_ | hi, this patch seems to have gotten a lot of +2s (except from me): https://review.openstack.org/#/c/165788/ | 16:56 |
| rloo_ | the bug associated with it isn't targetted for rc1: https://bugs.launchpad.net/ironic/+bug/1407888 | 16:56 |
| openstack | Launchpad bug 1407888 in Ironic "Implement get_boot_device for seamicro driver" [Medium,In progress] - Assigned to Harshada Mangesh Kakad (harshada-kakad) | 16:56 |
| rloo_ | so should I remove my -1? (I don't think I'm holding anything/anyone up, but thought I'd check) | 16:57 |
| NobodyCam | tiagogomes: have you seen http://docs.openstack.org/developer/ironic/deploy/install-guide.html#configure-neutron-to-communicate-with-the-bare-metal-server | 17:04 |
| *** edcragg has quit IRC | 17:05 | |
| *** kkoski has quit IRC | 17:07 | |
| *** romcheg has quit IRC | 17:11 | |
| JayF | 2015-04-07 17:04:46.137 | [ERROR] /opt/stack/devstack/lib/ironic:565 Timed out waiting for Nova hypervisor-stats count >= 3 <- I just got this trying to stack locally | 17:12 |
| JayF | I know this is basically what we've been seeing in the gate | 17:12 |
| *** meghal has joined #openstack-ironic | 17:12 | |
| JayF | curious if there's anything I should do to dig in since I have it locally? | 17:12 |
| JayF | adam_g: ^ thoughts? | 17:12 |
| adam_g | JayF, the n-cpu should be polling ironic for resources and reporting them in the logs | 17:13 |
| adam_g | you should see some in nova hypervisor-stats | 17:13 |
| adam_g | ill stack locally and see if its affecting me | 17:13 |
| JayF | adam_g: I'm asking more from a "diagnose this so we can fix the gate" perspective | 17:13 |
| adam_g | it looked like, in the gate, it was due to slow slaves | 17:13 |
| JayF | this is a blazing fast overpowered cloud server | 17:13 |
| JayF | like a Rackspace perf 8GB node | 17:14 |
| JayF | it's not slow :/ | 17:14 |
| adam_g | JayF, did the resources eventually get picked up? or did n-cpu reach its 60 attempts and bail out? | 17:14 |
| rloo_ | JayF: there's a bug open for that. https://bugs.launchpad.net/ironic/+bug/1441007 | 17:14 |
| openstack | Launchpad bug 1441007 in Ironic "Gate failure: Timed out waiting for Nova hypervisor-stats count >= 1" [Medium,Confirmed] | 17:14 |
| rloo_ | JayF: see the comment from Vladyslav | 17:14 |
| rloo_ | JayF: that has been happening since last week I think | 17:15 |
| jroll | 17:14:03 JayF | like a Rackspace perf 8GB node <- that's what the gate uses | 17:16 |
| *** kkoski has joined #openstack-ironic | 17:16 | |
| JayF | jroll: :( | 17:16 |
| JayF | adam_g: I can't run nova hypervisor-stats after sourcing ~/devstack/openrc for credentials, getting 403s | 17:16 |
| adam_g | JayF, source openrc admin admin | 17:17 |
| jroll | source openrc admin admin | 17:17 |
| jroll | JayF: slow is relative, I think better words are "slower than devstack expects" | 17:17 |
| JayF | NovaException: Error contacting Ironic server for 'node.list'. Attempt 60 of 60 | 17:17 |
| JayF | and n-cpu died as a result | 17:17 |
| jroll | right, is ir-api even up? | 17:17 |
| JayF | I gotta remember how to use screen :/ | 17:18 |
| NobodyCam | JayF: screen -dr | 17:19 |
| NobodyCam | lol | 17:19 |
| NobodyCam | *ducks* | 17:19 |
| JayF | more for scrolling, but I'm asking google | 17:19 |
| jroll | ctrl-a [ | 17:19 |
| NobodyCam | ^^ +++ | 17:19 |
| JayF | ironic-api started at 17:00 | 17:19 |
| JayF | Nova timed out at 16:52 | 17:19 |
| NobodyCam | wow | 17:20 |
| JayF | I wonder if JoshNang's addition to get the cleaning network is causing slowdowns | 17:20 |
| jroll | sho nuff, same problem we've been seeing | 17:20 |
| jroll | JayF: how dare you imply neutron is slow | 17:20 |
| JoshNang | heh | 17:20 |
| devananda | huh? | 17:20 |
| JoshNang | it's possible though. | 17:20 |
| JayF | like it didn't even *try* to start until 17:00:39, 8 minutes after n-cpu crashed from being unable to talk to it | 17:20 |
| NobodyCam | TheJulia: I am assuming bifrost will not create virsh nodes it self | 17:21 |
| JayF | devananda: https://bugs.launchpad.net/ironic/+bug/1441007 is what we're talking about | 17:21 |
| openstack | Launchpad bug 1441007 in Ironic "Gate failure: Timed out waiting for Nova hypervisor-stats count >= 1" [Medium,Confirmed] | 17:21 |
| jroll | JayF: I wonder if some things got re-arranged in that nova starts much earlier compared to ironic than it used to | 17:21 |
| jroll | not sure if that sentence parses, but I tried | 17:21 |
| JayF | idk, vdrok appears to have diagnosed a thing in the bug | 17:21 |
| JayF | " | 17:21 |
| JayF | There are some nova x509 requests which fail, and it takes a minute to fail each request. That seems to cause this late start of ironic. | 17:21 |
| jroll | oh, interesting | 17:22 |
| NobodyCam | oh looks like there is a create_vm_nodes script in tools | 17:22 |
| rloo_ | JayF: looks like this is a dup bug? https://bugs.launchpad.net/ironic/+bug/1440188 | 17:22 |
| openstack | Launchpad bug 1440188 in Ironic "Error contacting Ironic server for 'node.list' during check-tempest-dsvm-ironic-pxe_ssh" [Medium,Confirmed] | 17:22 |
| JayF | rloo_: yep | 17:23 |
| adam_g | JayF, which x509 cert requests are blocking? | 17:23 |
| TheJulia | NobodyCam: there is a playbook that will do it and a helper script in contrib | 17:23 |
| TheJulia | NobodyCam: err, tools | 17:24 |
| JayF | adam_g: idk, I was quoting a comment in bug 1441007 | 17:24 |
| openstack | bug 1441007 in Ironic "Gate failure: Timed out waiting for Nova hypervisor-stats count >= 1" [Medium,Confirmed] https://launchpad.net/bugs/1441007 | 17:24 |
| adam_g | oh | 17:24 |
| NobodyCam | yep found it :) | 17:24 |
| *** Sukhdev has joined #openstack-ironic | 17:26 | |
| *** harlowja_away is now known as harlowja | 17:31 | |
| *** athomas has quit IRC | 17:34 | |
| lucasagomes | folks I will call it a day | 17:36 |
| lucasagomes | have a good night everybody | 17:36 |
| NobodyCam | have a good night lucasagomes | 17:36 |
| adam_g | JayF, do you have your local devstack.log handy? | 17:36 |
| jroll | night lucas :) | 17:36 |
| *** achanda has joined #openstack-ironic | 17:36 | |
| *** lucasagomes is now known as lucas-dinner | 17:37 | |
| adam_g | im wondering if these openstack credentials commands taking 3 seconds each is whats eating up our timeout http://logs.openstack.org/39/170439/2/gate/gate-tempest-dsvm-ironic-agent_ssh/66849d2/logs/devstacklog.txt.gz#_2015-04-06_16_47_28_093 | 17:37 |
| *** rloo_ has quit IRC | 17:39 | |
| *** kkoski has quit IRC | 17:39 | |
| JayF | adam_g: https://gist.githubusercontent.com/jayofdoom/03f12d0720f1a6ee0df6/raw/gistfile1.txt | 17:40 |
| *** kkoski has joined #openstack-ironic | 17:40 | |
| adam_g | 2015-04-07 16:50:37.189 | + nova x509-get-root-cert /opt/stack/devstack/accrc/cacert.pem | 17:41 |
| adam_g | 2015-04-07 16:51:38.270 | ERROR (ClientException): The server has either erred or is incapable of performing the requested operation. (HTTP 500) (Request-ID: req-4ebd2970-7062-4ff2-a7b1-f8b2ad1e0381) | 17:41 |
| adam_g | tho i dont see the same in the gate logs im looking at | 17:41 |
| BadCub | heya folks... is anyone available to review: https://review.openstack.org/#/c/168120/ ? | 17:42 |
| *** trown is now known as trown|lunch | 17:43 | |
| devananda | wtf? https://bugs.launchpad.net/ironic/+bug/1439663 | 17:44 |
| openstack | Launchpad bug 1439663 in Ironic "localboot with whole disk image using dib ramdisk tries to install bootloader" [Undecided,New] - Assigned to Ramakrishnan G (rameshg87) | 17:44 |
| *** athomas has joined #openstack-ironic | 17:44 | |
| devananda | so that feature has yet another omg-im-broken bug | 17:44 |
| *** romcheg has joined #openstack-ironic | 17:44 | |
| jroll | devananda: all of these are related to DIB ramdisk, let's kill it with fire | 17:45 |
| devananda | yah | 17:45 |
| devananda | given that dib-ramdisk + local boot was a) broken by a change in dib that landed after feature freeze b) fixed with a rush patch that is still in the gate c) still broken if used with a whole disk image ... i'm quite frustrated with this feature | 17:50 |
| *** dprince has quit IRC | 17:51 | |
| *** meghal has quit IRC | 17:52 | |
| greghaynes | ^ is that related to a dib change? | 17:52 |
| greghaynes | We did touch some bootloadery bits kind of recently | 17:52 |
| jroll | devananda: indeed | 17:52 |
| jroll | greghaynes: no, it was all us | 17:52 |
| greghaynes | ah, ok | 17:52 |
| stendulker_ | Hi | 17:53 |
| stendulker_ | Upon rebasing the review https://review.openstack.org/#/c/167952/ ; py27 tests are failing. But same are passing in my dev environment. | 17:54 |
| jroll | jlvillal: ^ isn't that the test you mentioned earlier? http://logs.openstack.org/52/167952/15/check/gate-ironic-python27/725083b/console.html | 17:55 |
| jlvillal | jroll: Looking | 17:56 |
| stendulker_ | jroll: yes | 17:56 |
| stendulker_ | jroll, jvillal : they look to be related to last 2 submittals related to testcases in prepare_cleaning | 17:57 |
| jroll | seems like maybe a leaky test somewhere | 17:57 |
| jroll | changing dhcp configs or something maybe? | 17:57 |
| jlvillal | jroll: I think different from what I saw: http://paste.openstack.org/show/199616/ | 17:58 |
| jroll | dhcp provider | 17:58 |
| jlvillal | jroll: Oh no | 17:58 |
| jroll | jlvillal: yeah, though similar test | 17:58 |
| jlvillal | jroll: I see it has one of the failures | 17:58 |
| jlvillal | jroll: The link you posted has two failures. One of which matches the one I have seen on occasion | 17:58 |
| jroll | right | 17:58 |
| jroll | and that uses dhcp_factory.DHCPFactory().provider, maybe related | 17:59 |
| jroll | could be using the none provider | 17:59 |
| BadCub | heya folks. devananda just added a section to the Review pad called BUGS THAT LOOK CONCERNING. He would like folks to take a look and see if we need any of these in RC1 | 17:59 |
| *** ramineni has joined #openstack-ironic | 18:00 | |
| stendulker_ | jroll: yes its related to dhcp_factory.DHCPFactory().provider where it tries to get *_cleaning_ports | 18:01 |
| jroll | stendulker_: perhaps in setUp() set the dhcp provider configs? probably not the best way to solve this, but I would give it a try | 18:01 |
| *** jamielennox|away is now known as jamielennox | 18:01 | |
| ramineni | JayF: hi | 18:02 |
| stendulker_ | jroll: other one in test_agent.py is also same case for delete_cleaning_ports | 18:02 |
| NobodyCam | BadCub: https://bugs.launchpad.net/ironic/+bug/1439663 might be one | 18:02 |
| openstack | Launchpad bug 1439663 in Ironic "localboot with whole disk image using dib ramdisk tries to install bootloader" [Undecided,New] - Assigned to Ramakrishnan G (rameshg87) | 18:02 |
| *** absubram has joined #openstack-ironic | 18:04 | |
| BadCub | NobodyCam: I have marked that as HIGH and tagged it for RC1 | 18:05 |
| ramineni | JayF: Regarding your comment on https://review.openstack.org/#/c/169162/6/doc/source/drivers/ilo.rst | 18:05 |
| JayF | ramineni: I'm not sure how or if it's possible at all that iLo has a separate config option for erase_devices | 18:06 |
| JayF | ramineni: and if it's true, I wonder how that interacts with the agent-side priority change config option JoshNang added | 18:06 |
| *** meghal has joined #openstack-ironic | 18:06 | |
| ramineni | JayF: We had discussion on that at the time of review, having seperate options for iLo would be better , rather than using agent option in iLo drivers | 18:07 |
| JayF | that's the crux of my comment, that I'm 90% sure that the document isn't reflecting reality, and if it is, I think reality is broken | 18:07 |
| *** kejiahu has quit IRC | 18:07 | |
| jroll | JayF: ilo has it's own agent/secure erase implementation iirc | 18:07 |
| jroll | or something | 18:07 |
| JoshNang | i think the big part is ilo and stock agent share a lot of code, but not all of it | 18:07 |
| ramineni | jroll: ^^^ , remember? | 18:07 |
| jroll | maybe get_clean_steps implementation is different | 18:07 |
| JayF | jroll: the way JoshNang implemented the priority change I suspect only one of those options will actually work if both are set | 18:07 |
| JayF | jroll: which goes back to my comment about how reality is broken if this is true | 18:07 |
| jroll | JayF: I don't believe so | 18:08 |
| jroll | but I'm happy to look | 18:08 |
| JayF | join me :P I'm looking too | 18:08 |
| JoshNang | https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/ilo/deploy.py#L567 | 18:08 |
| jroll | JayF: they override get_clean_steps https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/ilo/deploy.py#L567 | 18:08 |
| jroll | heh, nice link | 18:09 |
| JoshNang | which is almost exactly https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/agent.py#L355 | 18:09 |
| JoshNang | heh | 18:09 |
| JayF | That makes me sad | 18:09 |
| jroll | we explicitly decided this | 18:09 |
| JayF | I feel bad for the operators who have to sort that our | 18:09 |
| JayF | Well I didn't explicitly decide it, and I disagree, but apparently I'm way too late | 18:09 |
| *** dprince has joined #openstack-ironic | 18:09 | |
| JayF | but that's going to be confusing as hell for operators | 18:09 |
| jroll | ilo and non-ilo servers may need different priorities | 18:09 |
| JayF | agent_* drivers behave one way, agent_ilo drivers behave another | 18:09 |
| JayF | I'm not saying there's not a use case; I'm saying it's confusing as hell | 18:10 |
| jroll | are we sure this isn't resource tracker lag? https://bugs.launchpad.net/ironic/+bug/1435153 | 18:11 |
| openstack | Launchpad bug 1435153 in Ironic "ironic hypervisor is still available when the node is in maintenance status." [Undecided,New] | 18:11 |
| jroll | devananda: Haomeng|2 ^ | 18:11 |
| JayF | ramineni: changed my vote on that patchset, although it looks like you have to convince jenkins too | 18:11 |
| JayF | JoshNang: I /strongly/ suggest we try to deprecate the iLo stuff being different when we implement a more generic agent priority change config option | 18:12 |
| JayF | JoshNang: maybe we can design it so it matches the use case, but is less confusing :/ | 18:12 |
| JoshNang | JayF: i'm hoping with boot interface, we can refactor a bunch of this | 18:13 |
| JayF | I don't know how boot interface helps in this case | 18:13 |
| JayF | I just don't like that which OOB driver you choose impacts options for the IB driver | 18:13 |
| JoshNang | i'd imagine boot interface will handle most of the agent interaction | 18:13 |
| JayF | maybe I have improper assumptions in my head around that, but they're assumptions we encourage with inband_outofband style driver names | 18:13 |
| JoshNang | at least, that's what i got from the spec ramesh put up | 18:14 |
| *** subscope_ has joined #openstack-ironic | 18:16 | |
| ramineni | JayF: ok, thanks | 18:18 |
| *** ramineni1 has joined #openstack-ironic | 18:24 | |
| *** mtanino has joined #openstack-ironic | 18:25 | |
| *** ramineni has quit IRC | 18:27 | |
| *** ramineni2 has joined #openstack-ironic | 18:30 | |
| *** pas-ha has quit IRC | 18:31 | |
| openstackgerrit | Merged openstack/ironic: Fix help string for glance auth_strategy option https://review.openstack.org/170439 | 18:32 |
| *** athomas has quit IRC | 18:32 | |
| openstackgerrit | Merged openstack/ironic: Document ports creating configuration for in-band inspection https://review.openstack.org/169253 | 18:32 |
| *** ramineni1 has quit IRC | 18:33 | |
| *** NobodyCam has quit IRC | 18:37 | |
| *** meghal has quit IRC | 18:38 | |
| *** ramineni has joined #openstack-ironic | 18:39 | |
| *** rloo has joined #openstack-ironic | 18:39 | |
| *** NobodyCam has joined #openstack-ironic | 18:39 | |
| *** trown|lunch is now known as trown | 18:40 | |
| *** ramineni2 has quit IRC | 18:40 | |
| *** ijw has quit IRC | 18:53 | |
| *** enikanorov has quit IRC | 19:01 | |
| *** Marga_ has quit IRC | 19:04 | |
| *** kbs1 has joined #openstack-ironic | 19:05 | |
| *** kbs has quit IRC | 19:06 | |
| *** jistr has joined #openstack-ironic | 19:07 | |
| openstackgerrit | Devananda van der Veen proposed openstack/python-ironicclient: Client should fall back to the lower versions if necessary https://review.openstack.org/171334 | 19:08 |
| devananda | that unbreaks the client | 19:08 |
| devananda | I need to run -- anyone want to pick ^ up and finish it? | 19:08 |
| * BadCub adds to pad | 19:09 | |
| devananda | I've tested with juno, early kilo, kilo with v1.4, and current master -- all work | 19:09 |
| BadCub | awesomeness | 19:10 |
| devananda | needs unit tests, and the code isn't al lthat clean. it probably fails pep | 19:10 |
| devananda | but it does the negotiation | 19:10 |
| * devananda runs afk for a couple hours | 19:10 | |
| adam_g | devananda, does that negate the need for https://review.openstack.org/#/c/167038/ | 19:10 |
| adam_g | er, cya | 19:10 |
| devananda | adam_g: yes | 19:10 |
| *** ramineni1 has joined #openstack-ironic | 19:10 | |
| adam_g | devananda, nice | 19:11 |
| devananda | wait, maybe | 19:11 |
| devananda | adam_g: I think it does. becayuse of htis | 19:12 |
| devananda | if not resp.getheader('X-OpenStack-Ironic-API-Version'): | 19:12 |
| adam_g | ya | 19:12 |
| devananda | self.os_ironic_api_version = 1.1 | 19:12 |
| NobodyCam | brb | 19:12 |
| devananda | k, now i'm really afk | 19:12 |
| *** ramineni has quit IRC | 19:13 | |
| JayF | adam_g: 2015-04-07 19:17:16.988 | + nova x509-get-root-cert /opt/stack/devstack/accrc/cacert.pem -- that's where it appears to be hanging for a full minute | 19:18 |
| *** igordcard_ has joined #openstack-ironic | 19:19 | |
| JayF | adam_g: and all this appears to be for the ec2/eucalyptus support | 19:19 |
| adam_g | JayF, ya, i see that in your local gist but haven't found an occurrence upstrea | 19:19 |
| JayF | similarly there's another x509-create-cert on the nova call that's slow as well | 19:19 |
| adam_g | JayF, yea, i wonder if out-of-tree ec2-api migratino has happened | 19:19 |
| JayF | and they appear to be failing at that | 19:19 |
| JayF | 2015-04-07 19:18:26.140 | + nova --os-password password --os-username demo --os-tenant-name invisible_to_admin x509-create-cert /opt/stack/devstack/accrc/invisible_to_admin/demo-pk.pem /opt/stack/devstack/accrc/invisible_to_admin/demo-cert.pem | 19:19 |
| JayF | 2015-04-07 19:19:27.242 | ERROR (ClientException): The server has either erred or is incapable of performing the requested operation. (HTTP 500) (Request-ID: req-d81fb513-cf17-42e1-8566-13f25fa75a76) | 19:19 |
| JayF | that's a full minute to do *nothing* | 19:19 |
| adam_g | the openstackclient ec2 commands take a good long time in the failed pxe job i was looking at. 3 seconds each, across a total of 24 calls | 19:20 |
| JayF | and it happens multiple times | 19:20 |
| adam_g | JayF, id be interested to see the n-api logs there, or w/e is erroring | 19:20 |
| JayF | adam_g: I'll gladly drop your ssh key on this machine if you want | 19:20 |
| adam_g | JayF, sure | 19:20 |
| adam_g | JayF, ssh-import-id gandelman-a | 19:20 |
| *** ramineni1 has quit IRC | 19:21 | |
| adam_g | also, i put up https://review.openstack.org/#/c/171313/ to bump the API retrires. realistically we can't count on devstack not adding a bunch of stuff todo between n-cpu and ir-api come up | 19:21 |
| JayF | adam_g: root at 104.130.169.38 or 2001:4802:7803:104:be76:4eff:fe20:376a | 19:21 |
| JayF | adam_g: stack is in /opt/stack where it should be, and after this stack.sh finishes I won't run anything else outside of a tmux session | 19:21 |
| *** subscope_ has quit IRC | 19:22 | |
| *** openstackgerrit has quit IRC | 19:22 | |
| JayF | adam_g: hmm. I may have local breakage too somehow | 19:23 |
| adam_g | JayF, is there a tmux/screen going now? | 19:23 |
| *** openstackgerrit has joined #openstack-ironic | 19:23 | |
| JayF | adam_g: as stack user, the one devstack creates | 19:23 |
| JayF | adam_g: none otherwise | 19:23 |
| JayF | adam_g: This stack.sh is still going, and n-cpu has already died | 19:23 |
| JayF | adam_g: and this is still in the middle of setting up a bucketload of EC2 crap | 19:23 |
| adam_g | JayF, n-api.log has some rpc timeouts | 19:24 |
| adam_g | File "/opt/stack/nova/nova/cert/rpcapi.py", line 89, in generate_x509_cert | 19:24 |
| JayF | adam_g: probably oto the n-cpu that crashed partway through, perhaps? | 19:24 |
| * NobodyCam is back | 19:24 | |
| JayF | adam_g: /me checks for RPC timeouts in gate | 19:25 |
| *** achanda has quit IRC | 19:25 | |
| JayF | adam_g: have any good failures for me to look at, perhaps? | 19:26 |
| adam_g | JayF, i haven't seen the x509 things happen in upstreams gate | 19:26 |
| *** Sukhdev has quit IRC | 19:26 | |
| adam_g | JayF, oh, it looks like you're not running the n-cert process? | 19:26 |
| JayF | adam_g: like they don't happen at all? | 19:26 |
| adam_g | i haven't found them, no | 19:27 |
| JayF | adam_g: what is n-cert? | 19:27 |
| JayF | adam_g: This is 100% shiny fresh devstack clone, with configs identical to those prescribed here: http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html | 19:27 |
| adam_g | JayF, the nova certificate service that listens on the rpc bus and generates certificates :) | 19:27 |
| adam_g | i dont remember *why* it exists | 19:28 |
| JayF | adam_g: well shouldn't devstack configure it by default if it's required? | 19:28 |
| adam_g | JayF, yeah? maybe it was removed from the default services. let me look | 19:28 |
| JayF | adam_g: I'm generally saying I don't think *I* did something wrong in this case; code is broken or docs are broken because I did a reasonable job of following the docs :P | 19:28 |
| adam_g | JayF, https://review.openstack.org/#/c/169709/ | 19:31 |
| *** meghal has joined #openstack-ironic | 19:31 | |
| adam_g | this smells like a devstack bug... it shouldn't be calling that stuff if the service isnt active | 19:31 |
| JayF | adam_g: or alternatively; should ec2 be disabled by default as well? | 19:31 |
| *** harlowja is now known as harlowja_away | 19:31 | |
| adam_g | JayF, AFAIK it is disabled by default now that it lives outside of nova and in its own devstack plugin | 19:32 |
| *** stendulker_ has quit IRC | 19:32 | |
| adam_g | JayF, none of this should affect our ironic deployment | 19:32 |
| JayF | adam_g: this is a fresh clone of devstack from this morning, with localrc exactly as documented | 19:32 |
| adam_g | right | 19:32 |
| *** kbs has joined #openstack-ironic | 19:32 | |
| JayF | so that's not a valid assumption, right? | 19:32 |
| adam_g | they removed n-cert from the default services that gets installed, but there's still code running by default that requires n-cert | 19:33 |
| JayF | And because that code fails, it's preventing Ironic from starting up fast enough | 19:33 |
| JayF | at least that's the local cause for that right now, correct? | 19:33 |
| *** meghal has quit IRC | 19:35 | |
| JayF | oh man this rabbithole is deep :x | 19:35 |
| *** kbs1 has quit IRC | 19:35 | |
| adam_g | JayF, yeah, in your case. i haven't seen it happen upstream tho | 19:36 |
| *** subscope_ has joined #openstack-ironic | 19:37 | |
| JayF | adam_g: So do you want me to get a fresh VM, go through our steps one by one, and prove out it's happening in this case? | 19:37 |
| *** achanda has joined #openstack-ironic | 19:37 | |
| *** harlowja_away is now known as harlowja | 19:37 | |
| JayF | adam_g: because I have nfi what would be different about these servers vs anyone else following Ironic docs on how to devstack | 19:37 |
| JayF | 2015-04-07 19:37:55.341 | /opt/stack/devstack/lib/ironic: line 560: [: -ge: unary operator expected | 19:38 |
| JayF | that is interesting, and bugged | 19:38 |
| * JayF looks | 19:38 | |
| adam_g | JayF, nothing is different. it looks like that devstack commit broke devstack when running with an unspecified ENABLED_SERVICES. im happy to start looking at fixing it there | 19:39 |
| JayF | adam_g: I think that's the case with me, at least directly | 19:39 |
| adam_g | yes, and likely anyone else following the docs | 19:39 |
| JayF | I set enable_service and disable_service | 19:39 |
| JayF | but never ENABLED_SERVICES="blah" | 19:39 |
| adam_g | including vlad @ https://bugs.launchpad.net/ironic/+bug/1441007 | 19:39 |
| openstack | Launchpad bug 1441007 in devstack "Gate failure: Timed out waiting for Nova hypervisor-stats count >= 1" [Undecided,In progress] | 19:39 |
| JayF | adam_g: so you're saying, set ENABLED_SERVICES="" in the top of my local.conf, and the stack works? | 19:40 |
| adam_g | JayF, as a workaround for now, ya. | 19:41 |
| adam_g | http://logs.openstack.org/38/167038/2/check/check-tempest-dsvm-ironic-pxe_ssh/c268f7e/logs/localrc.txt.gz | 19:41 |
| JayF | I'm trying my way before copying out that whole ENABLED_SERVICES var from your link | 19:41 |
| JayF | adam_g: Yeah; that doesn't work either. Looks like maybe we assume ENABLED_SERVICES= is sane and has default services started, so we don't dupe them in our docs | 19:43 |
| JayF | ENABLED_SERVICES=key,n-api,n-cpu,n-cond,n-sch,g-api,g-reg,rabbit,tempest,mysql,dstat,ironic,ir-api,ir-cond,q-svc,q-agt,q-dhcp,q-l3,q-meta,s-proxy,s-object,s-container,s-account # what my failing runs were running with | 19:44 |
| JayF | and gate runs with n-crt and n-obj | 19:46 |
| * JayF adds those to his services | 19:46 | |
| *** achanda has quit IRC | 19:47 | |
| adam_g | JayF, something tells me this isn't limited to ironic users. /me tries with no localrc | 19:47 |
| JayF | adam_g: I think because of the timing issue, we expose it in a way that can't be ignored | 19:47 |
| JayF | adam_g: whereas other folks would only notice ec2 stuff failed if testing it | 19:48 |
| *** ukalifon1 has quit IRC | 19:48 | |
| adam_g | JayF, create_userrc.sh assumes ec2 api and n-cert | 19:48 |
| JayF | adam_g: I don't see ec2 API in my enabled services that was failing | 19:48 |
| JayF | adam_g: so you're saying EC2 things are happening even though they're 100% unneeded? | 19:49 |
| adam_g | JayF, i dont know. gimme a few to dig through this and ill get back to you | 19:50 |
| JayF | okay cool, trying to help troubleshoot some but lots of demands on my time today (meeting in 10m followed by another meeting, hooray) | 19:50 |
| JayF | adam_g: fwiw; with that fixed ir-api came up with literally only 48 seconds to spare before n-cpu timed out | 19:54 |
| JayF | adam_g: so it's a /very/ tight fit timewise already | 19:54 |
| *** jistr has quit IRC | 19:55 | |
| JayF | adam_g: so if you have a patch that raises the number of attempts or makes the retries have a longer delay; I'd +1 it as a temp measure | 19:55 |
| adam_g | JayF, https://review.openstack.org/#/c/171313/ yeah, that should be increased regardless. | 19:56 |
| * devananda misses the bus, decides to stay home another hour and attend TC meeting from here | 19:56 | |
| *** Nisha has quit IRC | 19:56 | |
| devananda | adam_g: just wondering, why dont we start ir-api before n-cpu ? | 19:58 |
| *** dprince has quit IRC | 19:59 | |
| adam_g | devananda, just the way devstack is implemented. the core services startup and then it processes others via extras.d | 19:59 |
| devananda | adam_g: mmm. core services. | 20:00 |
| *** ijw has joined #openstack-ironic | 20:01 | |
| *** meghal has joined #openstack-ironic | 20:02 | |
| *** Shrews has quit IRC | 20:02 | |
| devananda | adam_g: ceilometer and heat both have hooks in stack.sh, within the service startup for keystone | 20:03 |
| devananda | adam_g: i dont see why ironic couldn't | 20:03 |
| adam_g | devananda, yeah, no idea why those are special | 20:03 |
| *** Shrews has joined #openstack-ironic | 20:03 | |
| devananda | adam_g: this is probably legacy structure from pre-integration | 20:04 |
| devananda | adam_g: *pre-ironic-integration. I'd suggest just fixing that | 20:04 |
| devananda | bumping the timeout is a reasonable stop-gap, but not a fix | 20:04 |
| adam_g | devananda, with the effort to move everything out of devstack and into project's plugins, i dont think that'd be a welcomed change | 20:05 |
| devananda | adam_g: won't we still need a way to express dependencies? | 20:05 |
| devananda | i see tons of things like... if is_this_service_enabled: if that-service_is_enabled: if something_else: do special thing; | 20:06 |
| adam_g | devananda, yea, but its bash not puppet/ansible | 20:07 |
| devananda | right | 20:09 |
| devananda | adam_g: it's just odd that ironic has no hooks anywhere in stack.sh. I think it's the only integrated project that doesn't? | 20:12 |
| devananda | adam_g: well, let me ask -- are heat, ceilometer, etc, pulling *out* their hooks from stack.sh? | 20:12 |
| adam_g | devananda, trove, sahara, tempest, are all installed via extras.d | 20:12 |
| adam_g | devananda, i dont know the plan TBH. i imagine that stuff is leagcy from a time before devstack was modularized | 20:13 |
| devananda | adam_g: k k, thanks | 20:13 |
| openstackgerrit | Merged openstack/ironic: Check status of bootloader installation for DIB ramdisk https://review.openstack.org/169419 | 20:13 |
| NobodyCam | brb | 20:15 |
| *** subscop__ has joined #openstack-ironic | 20:20 | |
| *** subscope_ has quit IRC | 20:23 | |
| JayF | JoshNang: I think you were wrong w/r/t driver_internal_info; see: | 20:23 |
| JayF | JoshNang: https://gist.github.com/jayofdoom/bf52c957ea54b5093d8a note no clean_steps | 20:24 |
| JoshNang | JayF: have you run through cleaning? | 20:24 |
| JoshNang | also, i'm pretty sure it deletes them at the end of cleaning. | 20:24 |
| JayF | JoshNang: look at state | 20:24 |
| *** subscop__ has quit IRC | 20:25 | |
| JoshNang | JayF: hmmmm | 20:25 |
| jroll | where does it put steps then? they have to go somewhere | 20:25 |
| JayF | aha | 20:26 |
| JoshNang | yeah, they definitely get stored in there | 20:26 |
| JayF | JoshNang: https://gist.github.com/jayofdoom/bf52c957ea54b5093d8a look at new comment | 20:26 |
| JayF | it was added as soon as the node exited cleaning | 20:27 |
| JayF | which the node would've exited as soon as it got the steps | 20:28 |
| jroll | JayF: is erase_disks disabled because devstack | 20:28 |
| JayF | so I think this works | 20:28 |
| jroll | ? | 20:28 |
| JoshNang | it would be in cleaning state for a couple minutes before it actually gets the clean steps | 20:28 |
| JayF | jroll: yah | 20:28 |
| jroll | yeah, ok | 20:28 |
| JoshNang | ^ that too | 20:28 |
| JayF | aight cool | 20:28 |
| *** achanda has joined #openstack-ironic | 20:29 | |
| *** alex_xu has quit IRC | 20:34 | |
| *** lintan_ has quit IRC | 20:35 | |
| openstackgerrit | Devananda van der Veen proposed openstack/python-ironicclient: Client should fall back to the lower versions if necessary https://review.openstack.org/171334 | 20:41 |
| * devananda fixed error messages | 20:41 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Set hash seed to 0 in gendocs environment https://review.openstack.org/171360 | 20:44 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Fix bad quoting in quickstart guide https://review.openstack.org/171361 | 20:44 |
| JayF | RFR ^ Two really simple changes to docs | 20:44 |
| NobodyCam | w00 h00 my first bifrost deploy: "result": "node activated" | 20:46 |
| TheJulia | NobodyCam: of course, the real question is, can you ping it? :) | 20:47 |
| *** igordcard_ has quit IRC | 20:47 | |
| NobodyCam | TheJulia: not yet | 20:49 |
| NobodyCam | :-p | 20:49 |
| TheJulia | you had a bunch of other things going on, so I would expect you to find issues | 20:49 |
| *** Marga_ has joined #openstack-ironic | 20:50 | |
| NobodyCam | lol so how long does shread take on a 10 g vm lol... doh | 20:58 |
| JayF | NobodyCam: forevah | 20:59 |
| JayF | NobodyCam: seriously, like 1h+ | 20:59 |
| NobodyCam | lol | 20:59 |
| NobodyCam | oh really | 20:59 |
| *** meghal has quit IRC | 20:59 | |
| * NobodyCam cries just a bit | 21:00 | |
| *** pelix has quit IRC | 21:03 | |
| devananda | anyone feel like reviewing an oslo sync? 165940 looks reasonable to me, but could use one or two more eyes | 21:10 |
| devananda | BadCub: ^ | 21:10 |
| * BadCub looks | 21:11 | |
| JayF | devananda: ty for the quick approval on those fixes, trying to get things fixed before I forget rst syntax :P | 21:11 |
| BadCub | devananda: ^ already has +A on it | 21:11 |
| devananda | oh, heh | 21:12 |
| devananda | so it does | 21:12 |
| devananda | rloo: cheers | 21:12 |
| BadCub | hehehe | 21:12 |
| rloo | huh? | 21:12 |
| rloo | oh. :D | 21:12 |
| *** jxiaobin has joined #openstack-ironic | 21:12 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Fixup post-merge comments on cleaning document https://review.openstack.org/171372 | 21:13 |
| JayF | rloo: ^ this patch should fix your concerns | 21:13 |
| rloo | JayF: looking... | 21:13 |
| JayF | please feel free to mark that review with any other issues in that doc you want fixed | 21:13 |
| JayF | since you didn't get a review the first time around. | 21:14 |
| JayF | I'd like these docs to be maximum-awesome | 21:14 |
| rloo | JayF: ha ha. 'maximum-awesome', I dunno ... | 21:15 |
| BadCub | anyone have to time to do some reviews on the last two bug patches we have? https://review.openstack.org/#/c/167952/ & https://review.openstack.org/#/c/168120/ | 21:17 |
| *** trown is now known as trown|outttypeww | 21:17 | |
| devananda | notes from post-TC-meeting discussions -- we are going to have to release and then maintain a stable branch of the client | 21:18 |
| NobodyCam | devananda: for K too? | 21:19 |
| devananda | yes | 21:19 |
| devananda | this release | 21:19 |
| rloo | ouch | 21:20 |
| NobodyCam | JayF: only about 20 minutes | 21:21 |
| *** mrda-away is now known as mrda | 21:21 | |
| mrda | Morning Ironic | 21:22 |
| NobodyCam | morning mrda | 21:22 |
| BadCub | heya mrda | 21:22 |
| devananda | morning, mrda | 21:22 |
| JayF | NobodyCam: Not that I'm saying the disks in my hardware lab are slow ... but yeah :P | 21:23 |
| mrda | o/ | 21:23 |
| NobodyCam | JayF: lol thats from my "taco" laptop... 7200 rpm spinning disk I think | 21:23 |
| *** chlong has joined #openstack-ironic | 21:25 | |
| rloo | JayF: wrt the cleaning doc, I'm in a meeting and need to listen, so only commented on the changes you made. | 21:26 |
| * devananda notices that the small oslo-incubator change we just approved triggers a much larger dependent patch that Shrews approved a few weeks ago | 21:26 | |
| devananda | rloo: ^ | 21:26 |
| devananda | 157602 | 21:26 |
| rloo | devananda: what? sigh. | 21:26 |
| JayF | rloo: I can wait to merge that if you want a closer look | 21:27 |
| rloo | JayF: don't wait for me. well, up to you. at least look at the comments I made. | 21:27 |
| *** pelix has joined #openstack-ironic | 21:28 | |
| rloo | devananda: well, your call if you want to try to get 157602 in. Looks like it needs to be rebased. | 21:29 |
| devananda | yuea, looks like it's failing rebase | 21:29 |
| *** meghal has joined #openstack-ironic | 21:29 | |
| devananda | which is good. i dont htink it should land right now | 21:29 |
| rloo | devananda: oh yeah. that was the touch-almost-all-files change. | 21:30 |
| *** meghal1 has joined #openstack-ironic | 21:31 | |
| *** Marga_ has quit IRC | 21:31 | |
| *** Marga_ has joined #openstack-ironic | 21:31 | |
| *** ijw has quit IRC | 21:33 | |
| *** meghal has quit IRC | 21:34 | |
| *** ijw has joined #openstack-ironic | 21:34 | |
| *** meghal1 has quit IRC | 21:35 | |
| *** openstackgerrit has quit IRC | 21:37 | |
| *** openstackgerrit has joined #openstack-ironic | 21:38 | |
| devananda | rloo, BadCub:165788 is a driver fix for seamicro. I reviewed it several weeks ago, and no one's made noise inthe meetings about it ... it has a bunch of +2's but rloo caught a small issue, as did I just now | 21:39 |
| devananda | would be eeasy for someone to fix those and ten wwe can merge it | 21:39 |
| rloo | devananda: the seamicro folks haven't asked for it to be in rc1, so honestly, i'm fine punting to Liberty. | 21:40 |
| devananda | k k | 21:40 |
| devananda | it's so close, but i'm fine either way. like i said, no one's raise any concerns | 21:40 |
| rloo | devananda: unless we haven't been clear, but i think in the meeting notes, etc, BadCub has mentioned people should bring up bugs for rc1. | 21:40 |
| devananda | (I'm just doing a quick review of old patches with +2's on them) | 21:41 |
| rloo | devananda: yeah, I know what you mean, I looked at that one earlier today ;) | 21:41 |
| devananda | when I unhide reviewed patchsets, I see soooo many old things I've forgotten about... | 21:41 |
| BadCub | I wasn't actually tracking that one for rc1 | 21:42 |
| rloo | devananda: ha ha. yeah. the community has done (and started) a lot of stuff :-) | 21:43 |
| devananda | indeed | 21:43 |
| devananda | BadCub: np. let's track it for early-liberty | 21:44 |
| BadCub | will tag the bug now | 21:44 |
| BadCub | devananda: the other bugs that are of concern listed. Should we go ahead and tag them for L at this point? | 21:46 |
| BadCub | there is 3 of them and comments have been added to most | 21:46 |
| *** pelix has quit IRC | 21:47 | |
| devananda | ++ | 21:47 |
| BadCub | done | 21:48 |
| *** ijw has quit IRC | 21:49 | |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: Fix VM stuck when deploying with pxe_ssh + local boot https://review.openstack.org/170117 | 21:49 |
| *** ijw has joined #openstack-ironic | 21:49 | |
| * devananda goes afk and catches a bus | 21:51 | |
| *** meghal has joined #openstack-ironic | 21:54 | |
| *** ijw_ has joined #openstack-ironic | 21:55 | |
| *** ijw has quit IRC | 21:59 | |
| NobodyCam | TheJulia: upon redeploy the node is now pingable :) | 21:59 |
| *** Marga_ has quit IRC | 22:01 | |
| *** Marga_ has joined #openstack-ironic | 22:02 | |
| TheJulia | Intresting... | 22:03 |
| TheJulia | NobodyCam: can you redeploy and see if its not pingable? | 22:04 |
| NobodyCam | i can try that | 22:04 |
| NobodyCam | running deploy.yaml now | 22:05 |
| TheJulia | thanks | 22:05 |
| TheJulia | persumably you have cleaning disabled by now :) | 22:05 |
| NobodyCam | oh yes :) | 22:05 |
| NobodyCam | heheheh | 22:06 |
| *** absubram has quit IRC | 22:06 | |
| TheJulia | good :) | 22:07 |
| TheJulia | because I'm going to be sadpanda if its not pingable | 22:07 |
| adam_g | rloo, ping | 22:08 |
| NobodyCam | :-p | 22:08 |
| rloo | hi adam_g, what's up? | 22:09 |
| adam_g | rloo, so the new client release seems to be causing problems. http://logs.openstack.org/05/162805/2/check/check-tempest-dsvm-ironic-pxe_ssh/d4f4219/logs/devstacklog.txt.gz#_2015-04-07_12_30_27_903 | 22:09 |
| adam_g | rloo, it looks like https://review.openstack.org/#/c/155031/ hoped to make it backward compat with the old args, right? | 22:10 |
| rloo | adam_g: yes, it was supposed to. I didn't think it modified the port stuff though. | 22:10 |
| rloo | adam_g: hmm, let me think about this... | 22:11 |
| *** alexpilotti has quit IRC | 22:12 | |
| rloo | adam_g: so --node_uuid should still work, just that the message won't show it cuz it is deprecated. | 22:12 |
| *** alexpilotti has joined #openstack-ironic | 22:13 | |
| rloo | adam_g: ahh. it shows --node as required though. | 22:13 |
| *** alexpilotti has quit IRC | 22:14 | |
| adam_g | rloo, bug? | 22:14 |
| adam_g | JayF, FYI https://review.openstack.org/#/c/171406/ | 22:14 |
| rloo | adam_g: yup, bug. | 22:14 |
| adam_g | rloo, ok. let me file one | 22:14 |
| rloo | adam_g: thx. did we just start using the new client? | 22:14 |
| adam_g | rloo, yeah, i think it went out yesterday or today | 22:15 |
| rloo | adam_g: I guess we don't use client from master. sigh. | 22:15 |
| adam_g | rloo, this is showing up on stable/juno | 22:15 |
| adam_g | master devstack has been updated already to use the new args | 22:15 |
| JayF | adam_g: https://review.openstack.org/#/c/171406/1/tools/create_userrc.sh,cm looks like you inadvertantly removed something on line 26, otherwise lgtm | 22:15 |
| jroll | adam_g: what if stable/juno pinned an older agent release | 22:16 |
| adam_g | JayF, oops. thanks | 22:16 |
| rloo | adam_g: ohhhhh. hmmm. i have to go now. Set the priority on that bug. if it is high, i'll try to look at it tonight. otherwise, will fix it tomorrow. thx! | 22:16 |
| adam_g | jroll, you mean the client? ideally it should be handled by global-requirements but nova apparently opposed it going into their requirements? yeah, i gues stable/juno devstack could install it with a version constraint | 22:17 |
| jroll | adam_g: yeah, that's what I mean, sorry, words are hard :) | 22:18 |
| JayF | adam_g: echo "XXX: $OS_CACERT" leftover from debugging? I don't see it anyone on the original and the XXX: looks suspcious :) | 22:19 |
| adam_g | JayF, jeez, thanks again | 22:19 |
| JayF | that's what reviews are for, lol | 22:20 |
| *** igordcard_ has joined #openstack-ironic | 22:21 | |
| *** romcheg has quit IRC | 22:25 | |
| *** ijw_ has quit IRC | 22:31 | |
| *** ijw has joined #openstack-ironic | 22:32 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: Fixup post-merge comments on cleaning document https://review.openstack.org/171372 | 22:32 |
| *** Marga_ has quit IRC | 22:32 | |
| *** Marga_ has joined #openstack-ironic | 22:33 | |
| JayF | rloo: ^ comments fixed, ty | 22:33 |
| *** alexpilotti has joined #openstack-ironic | 22:43 | |
| *** igordcard_ has quit IRC | 22:46 | |
| *** igordcard_ has joined #openstack-ironic | 22:47 | |
| *** Marga_ has quit IRC | 22:49 | |
| *** jgrimm has quit IRC | 22:51 | |
| *** maurosr has quit IRC | 22:51 | |
| *** jgrimm has joined #openstack-ironic | 22:52 | |
| *** maurosr has joined #openstack-ironic | 22:52 | |
| adam_g | rloo, fyi https://bugs.launchpad.net/python-ironicclient/+bug/1441374 | 22:53 |
| openstack | Launchpad bug 1441374 in python-ironicclient "'ironic port-create' 'node' vs 'node_uuid' is not backward compat" [Undecided,New] | 22:53 |
| *** jgrimm is now known as zz_jgrimm | 22:59 | |
| *** stendulker has joined #openstack-ironic | 23:01 | |
| stendulker | NobodyCam: Hi | 23:03 |
| devananda | back | 23:03 |
| devananda | adam_g: ugh. thanks | 23:04 |
| stendulker | NobodyCam: Did you get chance to look at this review https://review.openstack.org/#/c/167952/ ? | 23:05 |
| *** lucas-dinner has quit IRC | 23:05 | |
| *** achanda has quit IRC | 23:05 | |
| *** igordcard_ has quit IRC | 23:07 | |
| *** Sukhdev has joined #openstack-ironic | 23:08 | |
| NobodyCam | stendulker: I have not ... let me have a look | 23:09 |
| stendulker | NobodyCam: ok. Thank you. last day, so felt bit worried. all comments received are addressed | 23:10 |
| TheJulia | NobodyCam: was the node pingable on your second deploy attempt? | 23:10 |
| rloo | thx adam_g | 23:10 |
| NobodyCam | TheJulia: why YES it is | 23:11 |
| TheJulia | excellent! | 23:11 |
| BadCub | stendulker: we have ^ on the review list as well :-) | 23:11 |
| stendulker | BadCub: I saw that and happy to see that :) | 23:13 |
| BadCub | :) | 23:13 |
| openstackgerrit | Devananda van der Veen proposed openstack/python-ironicclient: Fix node_uuid option is required for port-create https://review.openstack.org/171429 | 23:13 |
| devananda | adam_g: ^ | 23:14 |
| stendulker | BadCub: Thank you | 23:14 |
| BadCub | stendulker: YW :) | 23:14 |
| BadCub | devananda: you want me to put ^ on the list? | 23:15 |
| devananda | BadCub: yah. that needs to land in the client soon | 23:15 |
| devananda | another way that the 0.5.0 release broke clients :( | 23:15 |
| devananda | we clearly aren't testing the client well enough ... | 23:15 |
| jroll | devananda: s/enough/at all/ | 23:16 |
| jroll | see also https://review.openstack.org/#/c/171426/ | 23:16 |
| devananda | jroll: well, we have some tests on it ... | 23:16 |
| jroll | we have unit tests | 23:16 |
| jroll | which don't test the shell because arguments are mocked | 23:17 |
| BadCub | devananda: added | 23:17 |
| devananda | anyone picked up 171334 ? | 23:17 |
| devananda | jroll: oh, right. and we have nova, which doesn't test the shell because it uses the lib | 23:17 |
| BadCub | not that I am aware of | 23:17 |
| devananda | and we have tempest, which doesn't test the client at all | 23:17 |
| devananda | jroll: you're right. we have NO tests for the CLI | 23:17 |
| jroll | devananda: right. but devstack tests the shell to some extent. | 23:17 |
| devananda | oh | 23:17 |
| jroll | it does some creates and stuff | 23:18 |
| devananda | yay! | 23:18 |
| jroll | but we weren't running that in the gate | 23:18 |
| jroll | we were running last release | 23:18 |
| jroll | (still are) | 23:18 |
| jroll | so we weren't testing via nova at all either | 23:18 |
| devananda | i thought that was fixed? | 23:18 |
| jroll | it's not voting yet, idk | 23:18 |
| devananda | no, seroiusly, i thought the client's gate now tests with tip of client | 23:18 |
| devananda | did that not land? | 23:18 |
| jroll | https://review.openstack.org/#/c/171426/ | 23:18 |
| devananda | omg | 23:19 |
| jroll | adam_g: left a comment there | 23:19 |
| devananda | adam_g: was there a similar review about a month back? | 23:19 |
| devananda | oh -- i see. this just makes the job voting now | 23:20 |
| jroll | still not in the gate either way | 23:20 |
| devananda | jroll: check still runs during gate, iiuc | 23:20 |
| * BadCub adds ^ to list as well | 23:21 | |
| jroll | devananda: I don't think so https://review.openstack.org/#/c/169036/ | 23:21 |
| *** achanda has joined #openstack-ironic | 23:21 | |
| devananda | oh | 23:22 |
| devananda | gnarrrr | 23:22 |
| devananda | unrelated to RC1, but important and related to versions - https://review.openstack.org/#/c/166386/ | 23:26 |
| NobodyCam | stendulker: +2'd | 23:28 |
| NobodyCam | BadCub: ^^^ | 23:29 |
| stendulker | NobodyCam: Thank you :) | 23:29 |
| * BadCub looks | 23:29 | |
| BadCub | devananda: added it to the microversions section on pad | 23:30 |
| NobodyCam | devananda: there wasn't a test for 171429 :( | 23:33 |
| *** davideagnello has quit IRC | 23:35 | |
| devananda | NobodyCam: see previous comments with jroll. we have (basically) no tests of the CLI | 23:35 |
| NobodyCam | :( ya was reading | 23:36 |
| devananda | did anyone file a bug for the "new client yields a 406 error talking to mid-kilo versions of the server" ? | 23:36 |
| *** Marga_ has joined #openstack-ironic | 23:37 | |
| jroll | devananda: https://bugs.launchpad.net/tripleo/+bug/1441170 | 23:38 |
| openstack | Launchpad bug 1441170 in tripleo "ironicclient "Not Acceptable" error" [Critical,Confirmed] - Assigned to Dan Prince (dan-prince) | 23:38 |
| NobodyCam | devananda: BadCub: 171429 +2'd | 23:40 |
| *** stendulker has quit IRC | 23:40 | |
| BadCub | NobodyCam: TY! :) | 23:41 |
| openstackgerrit | Devananda van der Veen proposed openstack/python-ironicclient: Client should fall back to the lower versions if necessary https://review.openstack.org/171334 | 23:42 |
| NobodyCam | devananda: https://review.openstack.org/#/c/171334 has several faliures is that a wip? | 23:42 |
| NobodyCam | lol | 23:42 |
| devananda | yes | 23:43 |
| BadCub | lol | 23:43 |
| devananda | it clearly says WIP i nthe commit message | 23:43 |
| BadCub | and on the pad lol | 23:43 |
| NobodyCam | oh the ** WIP ** is that what that means :-p | 23:43 |
| devananda | :P | 23:43 |
| NobodyCam | :) | 23:43 |
| BadCub | omg lol | 23:43 |
| jroll | NobodyCam asked if it was a 'wip', not a 'WIP' | 23:44 |
| *** Haomeng has joined #openstack-ironic | 23:44 | |
| jroll | you guys should know abuot case sensitivity by now | 23:44 |
| NobodyCam | lol | 23:44 |
| BadCub | lol | 23:44 |
| * BadCub reads urban dictionary to get definition of "abuot" | 23:45 | |
| BadCub | *ducks* | 23:45 |
| * NobodyCam will have glasses in about a week.. he'a hoping that helps make the world less fuzzy | 23:45 | |
| jroll | lmao | 23:45 |
| openstackgerrit | Devananda van der Veen proposed openstack/python-ironicclient: Client should fall back to the lower versions if necessary https://review.openstack.org/171334 | 23:45 |
| * TheJulia need to go to the eye doctor | 23:46 | |
| * BadCub goes to the eye tomrentor, er doctor on 16th | 23:46 | |
| NobodyCam | :) | 23:46 |
| *** Haomeng|2 has quit IRC | 23:47 | |
| TheJulia | BadCub: torturer? | 23:47 |
| BadCub | torturer, tormentor... Or just "the creep who wants to put things on my eyes" | 23:47 |
| devananda | so, I may need to drop off again shortly. i'll try to clean up ^ some more in the next few minutes | 23:47 |
| NobodyCam | devananda: are you on the bus coding? | 23:48 |
| devananda | in office. there's an evening thing I said I'd be at ... | 23:48 |
| NobodyCam | :) cloudbeers again? | 23:49 |
| TheJulia | BadCub: indeed :( | 23:50 |
| BadCub | TheJulia: hopefully they will get it right this time *frowns* | 23:51 |
| * NobodyCam thinks its time to call it a day. | 23:51 | |
| TheJulia | ditto | 23:52 |
| * BadCub does too | 23:52 | |
| NobodyCam | TheJulia: I'm going to look in to having bifrost disable cleaning if testing is set | 23:52 |
| NobodyCam | but not until tomorrpw | 23:52 |
| NobodyCam | tomorrow even | 23:53 |
| BadCub | Have a good night folks! | 23:53 |
| *** BadCub is now known as BadCub_Hiding | 23:53 | |
| TheJulia | NobodyCam: I swear it should be doing that, but I lost the last few hours digging into the ansible based issue you encountered | 23:53 |
| NobodyCam | TheJulia: it maybe and I just not looking in the right placces | 23:54 |
| TheJulia | I'll look at it in the morning | 23:54 |
| * TheJulia engages cloaking device | 23:54 | |
| NobodyCam | but now that it is working for me I dig in more tomorrow | 23:54 |
| *** rloo has quit IRC | 23:58 | |
| openstackgerrit | Merged openstack/ironic: Shorten time for unittest test_download_with_retries https://review.openstack.org/171002 | 23:59 |
| openstackgerrit | Devananda van der Veen proposed openstack/python-ironicclient: Client should fall back to the lower versions if necessary https://review.openstack.org/171334 | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!