| *** e0ne has joined #heat | 00:01 | |
| *** e0ne_ has joined #heat | 00:03 | |
| *** e0ne has quit IRC | 00:03 | |
| *** e0ne_ has quit IRC | 00:08 | |
| *** matsuhashi has joined #heat | 00:23 | |
| *** duncanjw has joined #heat | 00:32 | |
| *** duncanjw has quit IRC | 00:37 | |
| stevebaker | SpamapS: that is a fix from a recent migration regression, so it is at least improving the situation | 00:40 |
|---|---|---|
| *** pablosan is now known as zz_pablosan | 00:43 | |
| SpamapS | stevebaker: improving the situation is great. But without a regression test, it is likely to .. well.. regress | 00:43 |
| SpamapS | stevebaker: and testing that is really easy | 00:44 |
| *** alexpilotti has quit IRC | 00:46 | |
| stevebaker | SpamapS: yes, it should be easy to test. We should establish a pattern for testing these template format migrations which doesn't require full fake db migration setup <-- sdake | 00:47 |
| *** DandyPandy has quit IRC | 00:50 | |
| openstackgerrit | A change was merged to openstack/heat: Fix incorrect error msg in validate_template() https://review.openstack.org/83677 | 00:56 |
| *** DandyPandy has joined #heat | 00:57 | |
| *** e0ne has joined #heat | 01:01 | |
| *** funzo_ has joined #heat | 01:04 | |
| *** spzala has joined #heat | 01:05 | |
| *** andersonvom_ has joined #heat | 01:08 | |
| *** Slower_ has joined #heat | 01:09 | |
| *** blomquis` has joined #heat | 01:12 | |
| *** ruhe2 has joined #heat | 01:13 | |
| *** cody-somerville has quit IRC | 01:13 | |
| *** andersonvom has quit IRC | 01:13 | |
| *** dims has quit IRC | 01:13 | |
| *** dmakogon_ has quit IRC | 01:13 | |
| *** Darnoth has quit IRC | 01:13 | |
| *** pasquier-s has quit IRC | 01:13 | |
| *** ruhe has quit IRC | 01:13 | |
| *** DinaBelova has quit IRC | 01:13 | |
| *** Daviey has quit IRC | 01:13 | |
| *** e0ne has quit IRC | 01:13 | |
| *** killer_prince has quit IRC | 01:13 | |
| *** che-arne has quit IRC | 01:13 | |
| *** blomquisg_gone has quit IRC | 01:13 | |
| *** funzo has quit IRC | 01:13 | |
| *** dteselkin has quit IRC | 01:13 | |
| *** Slower has quit IRC | 01:13 | |
| *** ruhe2 is now known as ruhe | 01:13 | |
| *** DinaBelova has joined #heat | 01:14 | |
| *** dteselkin has joined #heat | 01:14 | |
| *** Daviey has joined #heat | 01:16 | |
| *** dmakogon_ has joined #heat | 01:20 | |
| *** cody-somerville has joined #heat | 01:20 | |
| *** pasquier-s has joined #heat | 01:21 | |
| *** killer_prince has joined #heat | 01:21 | |
| *** dims has joined #heat | 01:21 | |
| *** che-arne has joined #heat | 01:21 | |
| *** matsuhashi has quit IRC | 01:24 | |
| *** matsuhashi has joined #heat | 01:26 | |
| *** nosnos has joined #heat | 01:31 | |
| *** arbylee has joined #heat | 01:46 | |
| *** arbylee has quit IRC | 01:52 | |
| *** sdake has quit IRC | 01:54 | |
| *** Daviey has quit IRC | 02:00 | |
| *** e0ne has joined #heat | 02:01 | |
| *** andersonvom_ has quit IRC | 02:02 | |
| *** e0ne has quit IRC | 02:06 | |
| *** spzala has quit IRC | 02:13 | |
| *** Michalik- has joined #heat | 02:20 | |
| *** ramishra has joined #heat | 02:21 | |
| *** IlyaE has quit IRC | 02:30 | |
| *** zz_pablosan is now known as pablosan | 02:33 | |
| *** pablosan has quit IRC | 02:36 | |
| *** sdake has joined #heat | 02:45 | |
| *** liang has joined #heat | 02:50 | |
| *** Michalik- is now known as Michalik | 03:00 | |
| *** e0ne has joined #heat | 03:01 | |
| *** e0ne has quit IRC | 03:03 | |
| *** andersonvom has joined #heat | 03:17 | |
| *** matsuhashi has quit IRC | 03:19 | |
| *** andersonvom has quit IRC | 03:28 | |
| openstackgerrit | A change was merged to openstack/heat: Ensure that the NoCloud data source is loaded, part 2 https://review.openstack.org/78704 | 03:29 |
| openstackgerrit | A change was merged to openstack/heat: Add IP address to error message of failed stack https://review.openstack.org/77231 | 03:29 |
| openstackgerrit | Steve Baker proposed a change to openstack/heat: Credentials for native heat os-collect-config polling https://review.openstack.org/84046 | 03:29 |
| openstackgerrit | Steve Baker proposed a change to openstack/heat: Allow for stack users in _authorize_stack_user https://review.openstack.org/84047 | 03:29 |
| openstackgerrit | A change was merged to openstack/heat: Add events for automation statuses to Cloud Server https://review.openstack.org/80361 | 03:32 |
| stevebaker | SpamapS: when you're about I'd like to talk about the unit tests for my native heat occ collector | 03:33 |
| *** nosnos has quit IRC | 03:34 | |
| *** andersonvom has joined #heat | 03:44 | |
| *** matsuhashi has joined #heat | 03:57 | |
| *** e0ne has joined #heat | 04:01 | |
| *** nosnos has joined #heat | 04:02 | |
| *** e0ne has quit IRC | 04:06 | |
| *** ramishra_ has joined #heat | 04:16 | |
| *** ramishra has quit IRC | 04:17 | |
| *** nati_ueno has joined #heat | 04:23 | |
| *** nati_ueno has quit IRC | 04:23 | |
| *** IlyaE has joined #heat | 04:27 | |
| *** lazy_prince has joined #heat | 04:39 | |
| *** matsuhashi has quit IRC | 04:39 | |
| *** sdake has quit IRC | 04:39 | |
| *** pasquier-s has quit IRC | 04:39 | |
| *** killer_prince has quit IRC | 04:39 | |
| *** lazy_prince is now known as killer_prince | 04:39 | |
| *** gokrokve has joined #heat | 04:41 | |
| *** matsuhashi has joined #heat | 04:43 | |
| *** sdake has joined #heat | 04:46 | |
| *** pasquier-s has joined #heat | 04:46 | |
| *** andersonvom has quit IRC | 04:49 | |
| *** gokrokve has quit IRC | 04:52 | |
| *** gokrokve has joined #heat | 04:52 | |
| *** sdake has quit IRC | 04:54 | |
| *** akuznetsov has joined #heat | 04:55 | |
| *** e0ne has joined #heat | 05:01 | |
| *** e0ne has quit IRC | 05:04 | |
| *** alexheneveld has quit IRC | 05:07 | |
| skraynev | morning! | 05:09 |
| *** gokrokve_ has joined #heat | 05:10 | |
| *** gokrokve has quit IRC | 05:13 | |
| *** gokrokve_ has quit IRC | 05:15 | |
| *** gokrokve has joined #heat | 05:16 | |
| *** killer_prince has quit IRC | 05:19 | |
| *** gokrokve has quit IRC | 05:20 | |
| *** kevinbenton has left #heat | 05:25 | |
| *** e0ne has joined #heat | 05:32 | |
| *** killer_prince has joined #heat | 05:34 | |
| *** e0ne has quit IRC | 05:36 | |
| mattoliverau | Morning skraynev | 05:46 |
| *** gokrokve has joined #heat | 05:46 | |
| skraynev | mattoliverau: hi ;) | 05:47 |
| *** gokrokve_ has joined #heat | 05:49 | |
| *** chandan_kumar has joined #heat | 05:50 | |
| *** gokrokve has quit IRC | 05:51 | |
| *** nosnos_ has joined #heat | 05:53 | |
| *** nosnos has quit IRC | 05:53 | |
| *** matsuhashi has quit IRC | 05:54 | |
| *** matsuhashi has joined #heat | 05:55 | |
| *** gokrokve_ has quit IRC | 05:59 | |
| *** e0ne has joined #heat | 06:01 | |
| *** e0ne has quit IRC | 06:02 | |
| *** duncanjw has joined #heat | 06:08 | |
| *** pradipta has joined #heat | 06:09 | |
| *** bvandenh has joined #heat | 06:10 | |
| openstackgerrit | Jenkins proposed a change to openstack/heat: Imported Translations from Transifex https://review.openstack.org/83957 | 06:11 |
| *** tspatzier has joined #heat | 06:13 | |
| openstackgerrit | Sergey Kraynev proposed a change to openstack/heat: Using default value during update https://review.openstack.org/83709 | 06:15 |
| openstackgerrit | Sergey Kraynev proposed a change to openstack/heat: Fix incorrect logic during updating in unittest https://review.openstack.org/83708 | 06:15 |
| *** IlyaE has quit IRC | 06:16 | |
| skraynev | liang: ping | 06:17 |
| openstackgerrit | Jun Jie Nan proposed a change to openstack/heat: Adjust user_data_format for server with deployment https://review.openstack.org/82992 | 06:17 |
| *** mkollaro has joined #heat | 06:18 | |
| *** alexheneveld has joined #heat | 06:21 | |
| SpamapS | stevebaker: sorry, been a very family-focused Sunday. ;) We'll talk tomorrow. | 06:25 |
| openstackgerrit | Matthew Oliver proposed a change to openstack/heat: Standalone heat can't specify a region https://review.openstack.org/83669 | 06:27 |
| *** duncanjw has quit IRC | 06:29 | |
| *** duncanjw has joined #heat | 06:30 | |
| *** duncanjw has quit IRC | 06:30 | |
| *** duncanjw has joined #heat | 06:30 | |
| liang | skraynev, pong | 06:31 |
| *** alexheneveld has quit IRC | 06:31 | |
| *** duncanjw has quit IRC | 06:33 | |
| skraynev | liang: I wanted ask you here about my fix for unittests, but I did it through gerrit | 06:34 |
| skraynev | liang: please look my comment for https://review.openstack.org/#/c/83708/ | 06:34 |
| liang | skraynev, looking ;) | 06:35 |
| *** IlyaE has joined #heat | 06:36 | |
| liang | skraynev, so the next patch will use the default value for image_update_policy property when the new template doesn't specify one? | 06:37 |
| skraynev | liang: right | 06:38 |
| liang | skraynev, which effectively makes the old image_update_policy property not being used anymore. | 06:38 |
| skraynev | liang: not fully, when I did this change I thought same | 06:39 |
| liang | skraynev, Ok. perhaps you would like to associate that patch with https://bugs.launchpad.net/heat/+bug/1298850 ? | 06:40 |
| uvirtbot | Launchpad bug 1298850 in heat "Use default value during update" [Medium,In progress] | 06:40 |
| skraynev | liang: old property will be used in case if image_update_policy was not changes at all. (like in this test) | 06:41 |
| skraynev | liang: sure, I will mark as Related-Bug. ok? | 06:41 |
| liang | skraynev, yes, to us. But to the end users, they will think only new property is used. | 06:43 |
| liang | skraynev, yeah, Related-Bug: #1298850 | 06:43 |
| liang | skraynev, thanks!! | 06:43 |
| skraynev | liang: thank you =) | 06:44 |
| liang | skraynev, you are welcome ;) | 06:44 |
| *** alexheneveld has joined #heat | 06:45 | |
| skraynev | liang: ;) | 06:45 |
| *** rpothier has joined #heat | 06:48 | |
| *** rpothier_ has quit IRC | 06:50 | |
| mattoliverau | Night all, time to go make dinner, for those of you who are on the other side of the world, enjoy your Monday :) | 06:50 |
| *** alexheneveld has quit IRC | 06:51 | |
| *** IlyaE has quit IRC | 06:52 | |
| therve | Good morning | 06:53 |
| *** alexheneveld has joined #heat | 06:55 | |
| *** gokrokve has joined #heat | 06:56 | |
| *** alexheneveld has quit IRC | 07:00 | |
| *** gokrokve has quit IRC | 07:01 | |
| *** chandan_kumar has quit IRC | 07:06 | |
| *** cmyster has joined #heat | 07:06 | |
| *** matsuhashi has quit IRC | 07:08 | |
| *** matsuhashi has joined #heat | 07:08 | |
| *** TonyBurn has joined #heat | 07:09 | |
| *** nosnos_ has quit IRC | 07:09 | |
| *** nosnos has joined #heat | 07:10 | |
| *** stannie has joined #heat | 07:11 | |
| *** tomek_adamczewsk has joined #heat | 07:14 | |
| *** pas-ha has joined #heat | 07:16 | |
| pas-ha | morning, heaters :) | 07:16 |
| *** ifarkas has joined #heat | 07:18 | |
| cmyster | morning | 07:20 |
| cmyster | I misread it as haters... | 07:20 |
| cmyster | coffee++ | 07:20 |
| tspatzier | cmyster: LOL | 07:20 |
| *** matsuhashi has quit IRC | 07:22 | |
| openstackgerrit | Thierry Carrez proposed a change to openstack/heat: Open Juno development https://review.openstack.org/84077 | 07:22 |
| *** yogesh has joined #heat | 07:26 | |
| *** mkollaro has quit IRC | 07:27 | |
| *** jstrachan has joined #heat | 07:28 | |
| *** matsuhashi has joined #heat | 07:28 | |
| *** sergmelikyan has joined #heat | 07:31 | |
| *** tomek_adamczewsk has quit IRC | 07:36 | |
| *** ramishra_ has quit IRC | 07:38 | |
| *** tomek_adamczewsk has joined #heat | 07:38 | |
| *** jistr has joined #heat | 07:40 | |
| *** julienvey has joined #heat | 07:42 | |
| *** jprovazn has joined #heat | 07:44 | |
| *** matsuhashi has quit IRC | 07:46 | |
| *** matsuhashi has joined #heat | 07:47 | |
| *** tomek_adamczewsk has quit IRC | 07:49 | |
| *** yogesh has quit IRC | 07:55 | |
| *** mkerrin has joined #heat | 07:59 | |
| *** akuznetsov has quit IRC | 07:59 | |
| *** akuznetsov has joined #heat | 07:59 | |
| *** shardy has joined #heat | 08:02 | |
| *** pas-ha has quit IRC | 08:02 | |
| shardy | morning all | 08:02 |
| *** akuznetsov has quit IRC | 08:06 | |
| shardy | huh? I'm I seeing things or has our stable/grizzly branch been deleted? | 08:08 |
| shardy | *Am* I | 08:09 |
| *** derekh has joined #heat | 08:09 | |
| * shardy needs moar coffee | 08:09 | |
| skraynev | shardy: really s/moar/more ;) | 08:09 |
| cmyster | no, actually moar | 08:10 |
| cmyster | MOAR even | 08:10 |
| cmyster | hi BTW | 08:10 |
| skraynev | Oh. you right MOAR. | 08:11 |
| *** ramishra has joined #heat | 08:12 | |
| cmyster | you just need moar coffee skraynev, all will look better after that | 08:12 |
| shardy | https://github.com/openstack/heat/releases/tag/grizzly-eol | 08:12 |
| shardy | I guess that explains it | 08:12 |
| shardy | I didn't expect the branch to actually disappear | 08:13 |
| cmyster | Hmmm | 08:13 |
| *** pas-ha has joined #heat | 08:14 | |
| skraynev | cmyster: possibly problem was, that I have not known this word. (and use dictionary to get meaning) | 08:14 |
| liang | shardy, so does that mean grizzly wouldn't be maintained anymore? | 08:15 |
| cmyster | and lack of coffee ;) | 08:15 |
| cmyster | liang: depends, no new features for sure, maybe only critical bug fixes for paying supported customers but I am not 100% | 08:16 |
| shardy | liang: It would appear so, yes | 08:16 |
| skraynev | cmyster: lol. it's possible always | 08:17 |
| *** e0ne has joined #heat | 08:17 | |
| shardy | liang: https://wiki.openstack.org/wiki/Releases | 08:18 |
| liang | cmyster, shardy, thanks ;) | 08:18 |
| shardy | That looks like it's pending an update, but I guess grizzly goes eol now, and havana becomes security-supported | 08:18 |
| shardy | cmyster: upstream, there can be no further updates at all, because the stable/grizzly branch has just been removed | 08:19 |
| shardy | what distributors do with their grizzly packages is a different thing :) | 08:19 |
| cmyster | shardy: ya I know that, but thanks for clarifying | 08:22 |
| *** ramishra has quit IRC | 08:22 | |
| *** ramishra has joined #heat | 08:24 | |
| *** akuznetsov has joined #heat | 08:32 | |
| *** Daviey has joined #heat | 08:35 | |
| *** liang has quit IRC | 08:40 | |
| *** tomek_adamczewsk has joined #heat | 08:40 | |
| *** yogesh has joined #heat | 08:43 | |
| *** yogesh has quit IRC | 08:44 | |
| *** RockKuo_TW_PC has joined #heat | 08:48 | |
| *** liang has joined #heat | 08:51 | |
| cmyster | shardy: I am looking for a way to programmatically return a list of keypair names. do we have something like that in heat or should I just use nova ? | 08:52 |
| *** RockKuo_TW has quit IRC | 08:52 | |
| shardy | cmyster: We have a Constraint plugin which will check a keypair name against the list from nova, but AFAIK we don't have anything which will allow you to directly manipulate the list | 08:54 |
| *** julienvey has quit IRC | 08:54 | |
| cmyster | not manipulate, just thinking about how to create a stack with keypair and the use the returned outputs | 08:54 |
| cmyster | hmmm | 08:54 |
| *** bvandenh has quit IRC | 08:55 | |
| cmyster | can I just use the returned data (rsa key) as the keypair to use or do I need to state the name of the key ? | 08:55 |
| cmyster | if so I'll need to call nova and I rather avoid it for now | 08:56 |
| shardy | cmyster: OS::Nova::KeyPair allows you to do either | 08:56 |
| shardy | So you either pass the name of an existing key into the template (which is passed into OS::Nova::Server) | 08:57 |
| *** alexheneveld has joined #heat | 08:57 | |
| cmyster | well, that stack I alreay created. the question is, if I 'remember' the output of that resource, can I use the rsa code to launch another resource, say a nova::server | 08:57 |
| shardy | or you create a KeyPair resource, which creates the key for you, optionally from an exisitng key string | 08:57 |
| cmyster | ya ok then | 08:57 |
| *** gokrokve has joined #heat | 08:57 | |
| shardy | Can't you just output the name of the keypair from stack1 and pass it into stack2? | 08:58 |
| cmyster | no? | 08:58 |
| cmyster | no | 08:58 |
| cmyster | all I have is outputs | 08:58 |
| cmyster | and for the key its the string | 08:58 |
| cmyster | the data itself not the name that is | 08:58 |
| shardy | well you could output either | 08:59 |
| cmyster | right, thanks | 08:59 |
| shardy | {get_resource: my_keypair} | 09:00 |
| cmyster | hmmm | 09:00 |
| cmyster | even if on a differnt stack ? | 09:00 |
| shardy | cmyster: yes, although it would be best to make the second stack nested so you can delete them together | 09:01 |
| shardy | the get_resource returns the nova keypair ID | 09:01 |
| cmyster | not sure I want to, I am playing with a set of 'independent' stacks that can talk with one another | 09:01 |
| *** bvandenh has joined #heat | 09:02 | |
| shardy | well they're not really independent if one depends on a keypair in the other | 09:02 |
| *** gokrokve has quit IRC | 09:02 | |
| shardy | maybe create the keypair outside and pass it as a parameter to all stacks instead | 09:02 |
| cmyster | thats what I am working on, henceforth the original question | 09:03 |
| shardy | Sure, I'm just explaining what is possible, up to you to decide what's best ;) | 09:03 |
| shardy | obviously creating the keypair via a heat resource only makes sense if the stacks are related and share the same lifecycle | 09:04 |
| cmyster | will let you know in a jiffy, just need to clear that pesky 'u' from those utf8 strings else nothing is equal :) | 09:04 |
| cmyster | in this test they will. I'm creating, testing and destroying | 09:05 |
| *** e0ne_ has joined #heat | 09:08 | |
| *** e0ne has quit IRC | 09:12 | |
| cmyster | there is already an open bug about failing to felete stacks right ? | 09:36 |
| cmyster | wow, failing to delete even | 09:36 |
| *** julienvey has joined #heat | 09:47 | |
| *** tomek_adamczewsk has quit IRC | 09:52 | |
| *** tomek_adamczewsk has joined #heat | 09:53 | |
| *** gokrokve has joined #heat | 09:58 | |
| *** gokrokve has quit IRC | 10:03 | |
| *** liang has quit IRC | 10:15 | |
| *** alexpilotti has joined #heat | 10:22 | |
| *** alexpilotti has quit IRC | 10:26 | |
| *** liang has joined #heat | 10:39 | |
| *** liang has quit IRC | 10:49 | |
| *** e0ne_ has quit IRC | 10:54 | |
| *** alexpilotti has joined #heat | 10:57 | |
| *** gokrokve has joined #heat | 10:59 | |
| *** tomek_adamczewsk has quit IRC | 11:01 | |
| *** gokrokve has quit IRC | 11:03 | |
| *** liang has joined #heat | 11:07 | |
| *** tomek_adamczewsk has joined #heat | 11:08 | |
| *** liang has quit IRC | 11:12 | |
| *** tomek_adamczewsk has quit IRC | 11:14 | |
| *** tomek_adamczewsk has joined #heat | 11:16 | |
| *** rpothier has quit IRC | 11:26 | |
| *** matsuhashi has quit IRC | 11:26 | |
| *** nosnos has quit IRC | 11:27 | |
| lifeless | derekh: last thought - you might ask here if anyone knows about heat stack-list hangs etc :) | 11:30 |
| *** alexheneveld has quit IRC | 11:31 | |
| derekh | ok, so heat-list on our seed (deployed a few months ago) is hanging, has anybody seen this before and pointers as to why it might be happening ? | 11:32 |
| *** e0ne has joined #heat | 11:34 | |
| shardy | derekh: The only reason I can think of is the heat-api can't connect to the heat-engin via RPC, or the heat-engine is not running | 11:36 |
| derekh | shardy: ok, heat-engine is running , trying to verify RPC | 11:37 |
| shardy | derekh: e.g there was that reconnect issue recently, where I had to periodically restart qpidd to get the API reconnected to the engine | 11:38 |
| shardy | I think that's fixed in master and havana now though via an oslo patch | 11:39 |
| derekh | shardy: ok, I think lifeless tried that but will double check | 11:39 |
| cmyster | back | 11:40 |
| shardy | derekh: failing that pls enable debug in the heat.conf and we can take a look at the logs | 11:40 |
| derekh | shardy: ok, rabbitmq restart didn't work enabling debug | 11:41 |
| openstackgerrit | Pavlo Shchelokovskyy proposed a change to openstack/heat: Add handle_update to VolumeAttachment https://review.openstack.org/72681 | 11:42 |
| *** jprovazn has quit IRC | 11:42 | |
| cmyster | derekh: does horizon show you a list of stacks ? | 11:42 |
| cmyster | not exactly the same thing, but interesting to know | 11:43 |
| derekh | cmyster: we don't have horizon | 11:43 |
| cmyster | ok | 11:43 |
| *** e0ne has quit IRC | 11:44 | |
| *** e0ne has joined #heat | 11:44 | |
| *** saju_m has joined #heat | 11:51 | |
| derekh | shardy: dubug log produced by heat-api when I do a stack list http://paste.openstack.org/show/74655/ no corrisponding logs on heat engine, so looks like its not getting hit at all | 11:54 |
| derekh | the heat-engine logs don't produce anything after | 11:55 |
| derekh | 2014-03-31 11:42:17.857 1936 INFO heat.openstack.common.rpc.common [-] Connected to AMQP server on 127.0.0.1:5672 | 11:55 |
| derekh | 2014-03-31 11:42:17.857 1936 DEBUG heat.openstack.common.rpc.service [-] Creating Consumer connection for Service engine start /opt/stack/venvs/heat/local/lib/python2.7/site-packages/heat/openstack/common/rpc/service.py:50 | 11:55 |
| *** spzala has joined #heat | 11:57 | |
| shardy | derekh: hmm, interesting, could heat.conf be misconfigured? | 11:58 |
| shardy | e.g rabbit_host/qpid_hostname? | 11:58 |
| shardy | or is it only stack-list which hangs? | 11:58 |
| *** gokrokve has joined #heat | 12:00 | |
| *** saju_m has quit IRC | 12:00 | |
| derekh | rabbit_host = 127.0.0.1 # looks ok | 12:00 |
| derekh | "heat stack-show garbage" also hangs | 12:00 |
| derekh | gotta pop away for a bit will be back | 12:02 |
| *** spzala has quit IRC | 12:04 | |
| *** gokrokve has quit IRC | 12:05 | |
| *** saju_m has joined #heat | 12:06 | |
| shardy | derekh: I don't really have an answer for you unfortunately - if you restart heat-engine, rabbit and heat-api, and the config is correct, then it should work | 12:08 |
| *** saju_m has quit IRC | 12:08 | |
| *** ramishra_ has joined #heat | 12:10 | |
| *** ramishra has quit IRC | 12:13 | |
| *** blomquisg has joined #heat | 12:15 | |
| *** sergmelikyan has quit IRC | 12:17 | |
| *** jpeeler has joined #heat | 12:19 | |
| openstackgerrit | Chmouel Boudjnah proposed a change to openstack/heat: Skip tests when loading tests https://review.openstack.org/84117 | 12:25 |
| *** rpothier has joined #heat | 12:25 | |
| therve | pas-ha, Please don't do recheck no bug | 12:25 |
| pas-ha | tried to search for the bug on launchpad but found nothing.. currently looks like some specific volume tests are failing in many check jobs | 12:26 |
| pas-ha | but I can't produce a meaningful bug report on that :( | 12:27 |
| pas-ha | your ideas? | 12:28 |
| shardy | pas-ha: If you can't find a bug matching, raise a new one and recheck against that | 12:28 |
| therve | pas-ha, https://bugs.launchpad.net/bugs/1300161 I think | 12:28 |
| uvirtbot | Launchpad bug 1300161 in cinder "Error import taskflow Failure" [Undecided,New] | 12:28 |
| shardy | If everyone did recheck no bug when they found a new bug, we'd never fix any bugs ;) | 12:28 |
| therve | I'm surprised nobody else complained to be honest | 12:28 |
| pas-ha | therve: thanks for the link, wasn't able to find it by myself :( | 12:29 |
| therve | pas-ha, Don't give up! :) | 12:30 |
| therve | You can always ask here too | 12:30 |
| shardy | pas-ha: Are you aware of http://status.openstack.org/rechecks/? | 12:30 |
| shardy | That's my start point when looking for recheck bugs, followed by google normally | 12:30 |
| pas-ha | nope, that one is new for me, thanks for pointing | 12:30 |
| * pas-ha remembers to use google instead of launchpad for searching launchpad bugs | 12:31 | |
| shardy | pas-ha: You can search direct in launchpad too, but personally I get better results via google with part of the error logged and some additional keywords like "tempest" and "bug" | 12:32 |
| pas-ha | yes, I've already found that out too - the launchpad' search on its own content is quite inferior to google's | 12:33 |
| openstackgerrit | Bartosz Górski proposed a change to openstack/heat: Multi-region support with context concept https://review.openstack.org/53313 | 12:33 |
| therve | pas-ha, Another trick is that creating a bug uses a better search | 12:38 |
| therve | But yeah "test" site:launchpad.net works pretty well | 12:38 |
| *** dims has quit IRC | 12:38 | |
| *** sgordon has joined #heat | 12:43 | |
| *** sgordon has quit IRC | 12:43 | |
| *** sgordon has joined #heat | 12:43 | |
| *** jdob has joined #heat | 12:44 | |
| derekh | shardy: ok, thanks for looking, I'll keep poking at it, will let you knwo if I find anything | 12:46 |
| *** radez_g0n3 is now known as radez | 12:48 | |
| *** akuznetsov has quit IRC | 12:52 | |
| *** dims has joined #heat | 12:52 | |
| openstackgerrit | Thomas Spatzier proposed a change to openstack/heat: Move some template validation to Template class https://review.openstack.org/83984 | 12:53 |
| openstackgerrit | Thomas Spatzier proposed a change to openstack/heat: Check top-level sections when parsing a template https://review.openstack.org/83758 | 12:53 |
| *** dmakogon_ is now known as denis_makogon | 12:55 | |
| *** sballe has joined #heat | 12:57 | |
| *** gokrokve has joined #heat | 13:01 | |
| *** TonyBurn has quit IRC | 13:01 | |
| *** TonyBurn has joined #heat | 13:01 | |
| *** akuznetsov has joined #heat | 13:03 | |
| *** e0ne has quit IRC | 13:03 | |
| *** e0ne has joined #heat | 13:04 | |
| openstackgerrit | Chmouel Boudjnah proposed a change to openstack/heat: Skip tests when loading plugins https://review.openstack.org/84117 | 13:06 |
| *** gokrokve has quit IRC | 13:06 | |
| *** jdob has quit IRC | 13:06 | |
| *** jdob has joined #heat | 13:06 | |
| pas-ha | so, how this should be done practically for the new bug on the gate? "recheck bug XXXX" and the gate will ignore the failure with said symptoms, or we can only wait for the bug to be fixed? | 13:08 |
| shardy | pas-ha: Normally the gate bugs are transient so recheck may work, nothing is ignored, it's just a workaround for bugs which aren't repeatable 100% of the time | 13:10 |
| *** spzala has joined #heat | 13:10 | |
| pas-ha | but in this last case cinder has to be fixed / brought in line with taskflow, right? | 13:11 |
| therve | Yeah, the otherway around now :) | 13:11 |
| therve | Currently it's useless to recheck until the bug is fixed | 13:11 |
| pas-ha | now got it, thanks :) | 13:12 |
| pas-ha | pity we have only 1 patch left and stuck due to this gate bug to able to unfreeze.. | 13:12 |
| *** pradipta is now known as pradipta_away | 13:13 | |
| *** duncanjw has joined #heat | 13:21 | |
| *** pafuent has joined #heat | 13:21 | |
| *** ramishra_ has quit IRC | 13:28 | |
| *** ramishra_ has joined #heat | 13:30 | |
| *** aweiteka has joined #heat | 13:31 | |
| derekh | shardy: found the problem, rabbitmq blocks connections from doing anything if free disk space falls below a certain amount | 13:32 |
| shardy | derekh: aha, glad you found the issue :) | 13:32 |
| derekh | shardy: thanks for the ideas | 13:32 |
| *** alexheneveld has joined #heat | 13:32 | |
| *** jdob_ has joined #heat | 13:35 | |
| *** funzo_ is now known as funzo | 13:36 | |
| sdague | anyone up for helping me understand how incredibly I fubarred the ha template for the tempest test? - https://review.openstack.org/#/c/83924/ | 13:40 |
| shardy | sdague: I can try :) | 13:40 |
| *** bgorski has joined #heat | 13:40 | |
| sdague | it's complaining about not defining ImageId | 13:41 |
| sdague | http://logs.openstack.org/24/83924/2/check/check-tempest-dsvm-neutron-heat-slow/021be46/console.html#_2014-03-31_12_55_53_949 | 13:41 |
| shardy | sdague: are you aware that we'd like to move towards deprecating and removing HARestarter? | 13:42 |
| shardy | There are design issues with it so we've discussed replacing it (or possibly fixing it I guess) | 13:42 |
| sdague | yes, however, it's in the code now | 13:42 |
| skraynev | sdague: did you use autoscaling scenario test as exmaple? | 13:42 |
| sdague | skraynev: I tried to, but apparently did it badly | 13:42 |
| shardy | sdague: Ok, cool just wanted to make sure you knew :) | 13:42 |
| sdake_ | morning | 13:43 |
| skraynev | sdague: AFAIK, we have not any scenario test which work (because autoscaling is skipped now) | 13:43 |
| shardy | sdague: The LaunchConfiguration is missing ImageId | 13:44 |
| skraynev | all these tests needed F20 image, which is not in devstack | 13:44 |
| therve | sdague, LaunchConfig can be removed AFAIU | 13:45 |
| shardy | therve: +1 | 13:45 |
| sdague | skraynev: it is during the heat slow tests | 13:45 |
| therve | AccessKeys as well I guess? | 13:45 |
| shardy | likewise the User/AccessKey | 13:45 |
| *** jmckind has joined #heat | 13:45 | |
| skraynev | Oops. So I suppose, that therve and shardy are right | 13:45 |
| shardy | skraynev: there is no requirement to use F20 unless you want cfntools | 13:46 |
| sdague | so if I remove launch config, how do I get cfn-push-stats to run on the guest? | 13:46 |
| sdague | because my understanding was that was how HARestarter made it's decision | 13:46 |
| *** samstav has joined #heat | 13:47 | |
| shardy | sdague: You leave the user/accesskey and create the cfn-credentials file in the myserver user_data | 13:47 |
| shardy | the LaunchConfig is not doing anything atm | 13:47 |
| shardy | LaunchConfig is only used with groups of instances, e.g AutoScalingGroup | 13:47 |
| sdague | ok | 13:47 |
| shardy | sdague: also, if you wish, you can trigger the HARestarter by just hitting the URL it exposes | 13:48 |
| *** samstav has quit IRC | 13:48 | |
| shardy | you don't *have* to setup an alarm if you're just testing that one resource | 13:48 |
| * shardy grabs an example | 13:48 | |
| sdague | ok, so my thinking about this test was that we'd start the guest with heat | 13:48 |
| sdague | make sure it was up | 13:48 |
| sdague | kill it with nova | 13:48 |
| sdague | then make sure heat restarts it | 13:48 |
| sdague | if there is a simpler way to provide health ping back from the guest, I'm all for it | 13:49 |
| shardy | sdague: You certainly could do it that way, you just need to move the cfn-credentials file creation into the user_data and that should work | 13:49 |
| shardy | http://paste.fedoraproject.org/90221/13962738 | 13:50 |
| *** zigo has quit IRC | 13:50 | |
| shardy | Or, you could start the stack, check the instance is ACTIVE, then hit AlarmUrl from the test | 13:51 |
| shardy | sdague: 5mins, I'll do a working version of your current template | 13:52 |
| *** samstav has joined #heat | 13:52 | |
| sdague | well, if I hit the alarm url from the test, it wouldn't test that HARestart was correctly detecting that the resource went way | 13:52 |
| sdague | away | 13:52 |
| sdague | shardy: cool, thank you | 13:52 |
| shardy | HARestarter isn't detecting anything, it just responds to something hitting that URL | 13:53 |
| shardy | which in most cases is an alarm, but you can trigger it manually | 13:53 |
| *** zaneb has joined #heat | 13:57 | |
| sdague | ok, so for a scenario test we should bake the alarm part into this. I could see us doing HARestarter hitting that url directly as an API test, especially if there was a super cheap resource to test | 13:58 |
| *** zigo has joined #heat | 13:58 | |
| sdague | like a flavor existing or a user, if we have access to those | 13:59 |
| *** zns has joined #heat | 14:00 | |
| *** ramishra has joined #heat | 14:01 | |
| *** gokrokve has joined #heat | 14:01 | |
| *** ramishra_ has quit IRC | 14:02 | |
| *** maishsk has joined #heat | 14:04 | |
| maishsk | good afternoon | 14:04 |
| maishsk | anyone up and about ? | 14:04 |
| sdake_ | 7am here | 14:06 |
| sdake_ | barely awake | 14:06 |
| maishsk | Hi | 14:06 |
| *** gokrokve has quit IRC | 14:06 | |
| *** rwsu has joined #heat | 14:06 | |
| maishsk | I was wondering - HEAT has the option of performing operations during the creation of a stack | 14:07 |
| sdake_ | define "performing operations" | 14:08 |
| maishsk | is there any option to perform operations when a stack is removed - (just for example - remove DNS records or remove puppet certficates) | 14:08 |
| *** ramishra_ has joined #heat | 14:08 | |
| sdake_ | we call that quiescing, and we plan to tackle that in Juno | 14:08 |
| sdake_ | we need it for stack update and autoscaling use cases | 14:08 |
| sdake_ | tripleo and tusker cats need it | 14:08 |
| sdake_ | so its going to happen one way or another :) | 14:08 |
| *** ramishra has quit IRC | 14:09 | |
| *** cfriesen has joined #heat | 14:09 | |
| maishsk | is there a blueprint currently available? | 14:09 |
| *** bnemec has joined #heat | 14:09 | |
| *** lipinski has joined #heat | 14:10 | |
| sdake_ | i am not sure on blueprints | 14:10 |
| sdake_ | usually with a blueprint someone proposes HOW not WHAT | 14:11 |
| sdake_ | we haven't figured out HOW yet :) | 14:11 |
| sdake_ | but we got when figured out :) | 14:12 |
| maishsk | that is funny... :) | 14:12 |
| *** jordanP has joined #heat | 14:12 | |
| maishsk | any discussion topics that I look at / chime in? | 14:12 |
| *** cmyster has quit IRC | 14:13 | |
| jordanP | hi guys, I've juste filled a bug report " Possible confusion between auth_uri and auth_port" (https://bugs.launchpad.net/heat/+bug/1300246 ) | 14:13 |
| uvirtbot | Launchpad bug 1300246 in heat "Possible confusion between auth_uri and auth_port" [Undecided,New] | 14:13 |
| jordanP | can someone confirm my saying :) | 14:13 |
| jordanP | plz | 14:13 |
| *** sjmc7 has joined #heat | 14:19 | |
| *** jcru has joined #heat | 14:23 | |
| *** david-lyle has joined #heat | 14:25 | |
| *** Darnoth has joined #heat | 14:26 | |
| *** sballe has quit IRC | 14:29 | |
| *** duncanjw has quit IRC | 14:31 | |
| *** ramishra_ has quit IRC | 14:36 | |
| *** sergmelikyan has joined #heat | 14:37 | |
| *** jdob_ has quit IRC | 14:38 | |
| openstackgerrit | Bartosz Górski proposed a change to openstack/heat: Ensure parameter timeout_mins available in update https://review.openstack.org/76211 | 14:40 |
| *** jdob has quit IRC | 14:41 | |
| *** TonyBurn has quit IRC | 14:43 | |
| *** alexheneveld has quit IRC | 14:45 | |
| *** ramishra has joined #heat | 14:45 | |
| *** rcleere has joined #heat | 14:48 | |
| *** mkerrin has quit IRC | 14:53 | |
| *** vijendar has joined #heat | 14:54 | |
| *** alexheneveld has joined #heat | 14:55 | |
| *** alexheneveld_ has joined #heat | 14:56 | |
| *** maishsk has left #heat | 14:58 | |
| *** IlyaE has joined #heat | 14:58 | |
| *** alexheneveld has quit IRC | 14:59 | |
| *** alexheneveld_ is now known as alexheneveld | 14:59 | |
| openstackgerrit | A change was merged to openstack/heat: Properly delete OS::Nova::FloatingIPAssociation https://review.openstack.org/84021 | 14:59 |
| *** pas-ha has quit IRC | 15:02 | |
| *** gokrokve has joined #heat | 15:02 | |
| *** ramishra_ has joined #heat | 15:04 | |
| *** ramishra has quit IRC | 15:05 | |
| *** gokrokve has quit IRC | 15:06 | |
| *** tspatzier has quit IRC | 15:06 | |
| openstackgerrit | Thomas Herve proposed a change to openstack/heat: Pass and use the environment in validate_template https://review.openstack.org/83442 | 15:07 |
| *** alexpilotti has quit IRC | 15:07 | |
| *** blamar has quit IRC | 15:07 | |
| *** mkerrin has joined #heat | 15:11 | |
| *** ramishra_ has quit IRC | 15:11 | |
| *** ramishra has joined #heat | 15:12 | |
| therve | shardy, So do you think #1300246 is valid? It seems we should be using auth_host/port instead of auth_uri | 15:15 |
| *** ramishra has quit IRC | 15:16 | |
| *** gokrokve has joined #heat | 15:17 | |
| *** gokrokve_ has joined #heat | 15:20 | |
| *** gokrokve has quit IRC | 15:21 | |
| *** ramishra has joined #heat | 15:21 | |
| shardy | therve: I'm not sure tbh, AFAICS heat doesn't use auth_uri at all except for ec2tokens, where probably using the public endpoint is the right thing | 15:26 |
| therve | shardy, It's used by our KeystoneClientV3 it seems | 15:27 |
| therve | self.v3_endpoint = cfg.CONF.keystone_authtoken.auth_uri.replace( 'v2.0', 'v3') | 15:27 |
| openstackgerrit | Bartosz Górski proposed a change to openstack/heat: Fix timeout for stack actions https://review.openstack.org/77902 | 15:28 |
| shardy | therve: Hmm, maybe we need to change that, but my original expectation was that we'd want to make requests to the public keystone API, as they're on behalf of a user | 15:29 |
| *** kgriffs|afk is now known as kgriffs | 15:29 | |
| shardy | My understanding was that the admin details in auth_host etc were only for getting a new token, e.g in keystoneclient's auth_token | 15:29 |
| therve | That's one way to see it | 15:30 |
| therve | shardy, Some deployment(s) don't have access to publicURL from the infra itself | 15:31 |
| shardy | Saying "Other Openstack projects deals fine with this" is kinda innacurate, because all other projects only use keystone for token validation | 15:31 |
| *** alexpilotti has joined #heat | 15:31 | |
| shardy | therve: But aren't there limitations on what you can do via the keystone admin endpoint? | 15:31 |
| shardy | I thought there were | 15:31 |
| therve | Good questions | 15:31 |
| *** mkerrin has quit IRC | 15:31 | |
| jordanP | shardy, I filled the bug report. I know that 'Other Openstack projects deals fine with this' is inaccurate, sorry | 15:32 |
| shardy | jordanP: Hi, no worries, just trying to figure out what the best solution is | 15:32 |
| jordanP | okay | 15:32 |
| jordanP | in my env, publicURL (auth_uri) is behind a SSL but internal URL is not. PublicURL is not accesible from the heat node itself | 15:33 |
| shardy | jordanP: so, to clarify, you can access the public endpoint of every service except keystone? | 15:33 |
| therve | Arguably, we'd use adminURL when using service admin creds | 15:33 |
| jordanP | shardy, I'va submitted a patch to use internalURL in engine/clients.py | 15:34 |
| jordanP | it was merged a few month ago | 15:34 |
| jordanP | I meant, to be able to configure on which endpoint type to connect | 15:34 |
| jordanP | it works well with all projects, but keystone, as keystone is "special" | 15:34 |
| *** TonyBurn has joined #heat | 15:35 | |
| shardy | therve: that argument could be correct, I'm just a little worried about the implications of routing all user-initiated requests to the admin endpoint | 15:35 |
| therve | shardy, Agreed, we'd need to check what happens then | 15:36 |
| *** alexpilotti has quit IRC | 15:36 | |
| jordanP | do you route requests "as is" or you initiate new request to keystone ? | 15:36 |
| *** ramishra has quit IRC | 15:36 | |
| therve | jordanP, Both | 15:37 |
| shardy | jordanP: we initiate new requests, either directly using the credentials in the request, or using heat administrative credentials, depending on how you have configured heat | 15:37 |
| shardy | but, for example, are extensions like OS-TRUST available via the admin endpoint? | 15:38 |
| *** mkerrin has joined #heat | 15:38 | |
| jordanP | at the moment, the work arround for my env is to configure auth_uri to point to admin/private endpoint, but I shouldn't be so I think, as far as I understand the diff between auth_uri and auth_host | 15:38 |
| jordanP | shardy, I believe that's why there's a difference between auth_host and auth_uri, in other OS I have to configure both | 15:39 |
| jordanP | but heat only uses auth_uri | 15:39 |
| jordanP | *other OS projects | 15:39 |
| therve | jordanP, The point is that adminURL ought to be functionally different. We (heat) don't know the differences | 15:40 |
| openstackgerrit | A change was merged to openstack/heat: version migration fails if version is an integer https://review.openstack.org/83869 | 15:40 |
| shardy | jordanP: I think maybe the issues stem from our using auth_uri as a basis for the endpoint passed to keystoneclient | 15:41 |
| shardy | jordanP: that was a workaround for the fact that everyone still had hard-coded v2.0 endpoints | 15:41 |
| jordanP | I agree but I am not sure | 15:41 |
| shardy | perhaps we can look at removing that as AIUI keystoneclient now has version discovery | 15:42 |
| shardy | then we can let the client work out what endpoint to use via the catalog | 15:42 |
| shardy | which is the "right way" to do it, but until now didn't work for v3 keystone | 15:42 |
| shardy | jordanP: do you have versioned keystone endpoints in the catalog? | 15:44 |
| jordanP | I have to look how other project does it. But I feel that heat should no try to contact keystone using auth_uri | 15:44 |
| jordanP | shardy, yes I have. But ATM i tries to make heat work with keystone v2... | 15:45 |
| therve | Well that doesn't really tell us anything. Heat has some unique requirements | 15:45 |
| jordanP | so i am using the plugin in contrib/ | 15:45 |
| jordanP | therve, true. | 15:45 |
| shardy | jordanP: That would have been valuable information to disclose when reporting the issue | 15:46 |
| shardy | jordanP: the non-contrib heat_keystoneclient is v3 only | 15:46 |
| jordanP | the pb is the same, v2 and v3 tries to contact keystone trough auth_uri which is unreachable from my heat node | 15:47 |
| jordanP | I can make auth_uri to point to my internal endpoint, which should fixe the issue | 15:48 |
| jordanP | but I still think it's wrong and it's not what the keystone middleware expects, IMO | 15:49 |
| *** ramishra has joined #heat | 15:49 | |
| shardy | jordanP: Ok, so can you provide an example of another project which creates a keystoneclient Client, and passes in an auth_url derived from the admin endpoint details? | 15:50 |
| jordanP | yep ! | 15:51 |
| jordanP | that's what I should do now :) | 15:51 |
| shardy | jordanP: I'm just trying to decouple the "uses keystoneclient directly" requirement from "has auth_token int the API pipeline" use-case which all projects have | 15:51 |
| therve | jordanP, It seems auth_uri is only used by the middleware to do a redirect in case of error | 15:52 |
| jordanP | shardy, I don"t understand your last point, but I must say I am not familiar with all these mechanisms | 15:53 |
| shardy | jordanP: The configuration section in question is primarily for the API paste filter, provided by keystoneclient, the auth_token middleware | 15:53 |
| shardy | arguably we shouldn't be (ab)using the API configuration in heat-engine | 15:54 |
| jordanP | therve, auth_uri seems to be used to connect to keystone directly, I mean there's point to be somewhere the URL to talk to keystone, and to Heat, it's auth_uri | 15:54 |
| shardy | perhaps the simplest fix is to provide a heat-specific auth_uri which defaults to the keystonclient one if not set | 15:54 |
| *** tomek_adamczewsk has quit IRC | 15:54 | |
| therve | jordanP, I'm differentiating how heat uses auth_uri and how the keystone middleware uses it | 15:55 |
| jordanP | ah ok | 15:55 |
| therve | jordanP, You said "it's not what the keystone middleware expects" | 15:55 |
| therve | But it doesn't really uses it for anything useful | 15:55 |
| jordanP | indeed the middleware only uses auth_uri when talking with the client, to set an httpheader named WWW-something | 15:56 |
| therve | It does a redirect in case of 401 AFAIU | 15:57 |
| *** andrew_plunk has joined #heat | 15:58 | |
| *** e0ne_ has joined #heat | 16:01 | |
| *** tango has joined #heat | 16:01 | |
| *** ramishra has quit IRC | 16:04 | |
| *** e0ne has quit IRC | 16:04 | |
| *** IlyaE has quit IRC | 16:09 | |
| *** e0ne has joined #heat | 16:12 | |
| *** sdake has joined #heat | 16:14 | |
| *** sdake has quit IRC | 16:14 | |
| *** sdake has joined #heat | 16:14 | |
| *** e0ne_ has quit IRC | 16:15 | |
| *** jordanP has quit IRC | 16:21 | |
| *** pvaneck has joined #heat | 16:22 | |
| openstackgerrit | A change was merged to openstack/python-heatclient: Resolve files in resource types https://review.openstack.org/80528 | 16:24 |
| *** che-arne has quit IRC | 16:27 | |
| *** IlyaE has joined #heat | 16:30 | |
| *** e0ne has quit IRC | 16:39 | |
| *** pafuent has left #heat | 16:42 | |
| *** duncanjw has joined #heat | 16:45 | |
| *** RockKuo_TW_PC has quit IRC | 16:46 | |
| *** pafuent has joined #heat | 16:47 | |
| *** harlowja_away is now known as harlowja | 16:48 | |
| shardy | sdake: Sorry for the delay - I think this example basically does what you want: | 16:49 |
| shardy | oops sdague ^^ | 16:49 |
| shardy | http://paste.fedoraproject.org/90310/28451913 | 16:49 |
| shardy | tab-completion fail ;) | 16:49 |
| sdague | shardy: awesome, I'll give it a shot in a few. Thanks much. | 16:50 |
| *** jdob has joined #heat | 16:50 | |
| shardy | sdague: np! | 16:50 |
| *** RockKuo_iPad has joined #heat | 16:53 | |
| *** RockKuo_iPad has quit IRC | 16:57 | |
| *** andersonvom has joined #heat | 16:59 | |
| *** TravT has joined #heat | 17:03 | |
| *** derekh has quit IRC | 17:04 | |
| *** TravT has quit IRC | 17:07 | |
| *** TravT has joined #heat | 17:07 | |
| *** alexpilotti has joined #heat | 17:07 | |
| *** asalkeld has joined #heat | 17:07 | |
| *** rwsu has quit IRC | 17:11 | |
| *** jistr has quit IRC | 17:14 | |
| *** gokrokve has joined #heat | 17:15 | |
| SpamapS | stevebaker: I'm around for a while, so whenever you want to chat. | 17:16 |
| *** pablosan has joined #heat | 17:16 | |
| *** gokrokve_ has quit IRC | 17:18 | |
| *** cmyster has joined #heat | 17:21 | |
| cmyster | evening | 17:24 |
| *** bvandenh has quit IRC | 17:24 | |
| *** tspatzier has joined #heat | 17:25 | |
| *** arbylee has joined #heat | 17:30 | |
| *** gokrokve has quit IRC | 17:31 | |
| *** gokrokve has joined #heat | 17:32 | |
| *** gokrokve_ has joined #heat | 17:34 | |
| *** gokrokve has quit IRC | 17:36 | |
| *** bvandenh has joined #heat | 17:36 | |
| *** gokrokve has joined #heat | 17:37 | |
| *** gokrokve_ has quit IRC | 17:38 | |
| *** e0ne has joined #heat | 17:41 | |
| openstackgerrit | Thomas Spatzier proposed a change to openstack/heat: Move some template validation to Template class https://review.openstack.org/83984 | 17:42 |
| openstackgerrit | Thomas Spatzier proposed a change to openstack/heat: Check top-level sections when parsing a template https://review.openstack.org/83758 | 17:42 |
| *** arbylee has quit IRC | 17:45 | |
| *** arbylee has joined #heat | 17:45 | |
| *** rustlebee is now known as russellb | 17:47 | |
| *** gokrokve has quit IRC | 17:49 | |
| *** gokrokve has joined #heat | 17:49 | |
| *** packet has joined #heat | 17:50 | |
| *** gokrokve has quit IRC | 17:53 | |
| *** cmyster has quit IRC | 17:54 | |
| *** tspatzier has quit IRC | 17:54 | |
| *** TonyBurn has quit IRC | 17:56 | |
| *** yogesh has joined #heat | 17:56 | |
| *** cmyster has joined #heat | 17:56 | |
| openstackgerrit | Bartosz Górski proposed a change to openstack/heat: Fix timeout for stack actions https://review.openstack.org/77902 | 17:56 |
| *** denis_makogon has quit IRC | 17:59 | |
| *** dmakogon_ has joined #heat | 17:59 | |
| *** IlyaE has quit IRC | 18:07 | |
| *** jmckind has quit IRC | 18:08 | |
| *** IlyaE has joined #heat | 18:08 | |
| openstackgerrit | Bartosz Górski proposed a change to openstack/heat: Fix timeout for stack actions https://review.openstack.org/77902 | 18:11 |
| openstackgerrit | Cyril Roelandt proposed a change to openstack/python-heatclient: Python 3: fix test_template_utils.py https://review.openstack.org/84198 | 18:15 |
| *** tspatzier has joined #heat | 18:16 | |
| *** IlyaE has quit IRC | 18:17 | |
| *** kevinbenton has joined #heat | 18:17 | |
| *** IlyaE has joined #heat | 18:27 | |
| *** lipinski has quit IRC | 18:28 | |
| *** gokrokve has joined #heat | 18:29 | |
| *** gokrokve has quit IRC | 18:34 | |
| *** gokrokve has joined #heat | 18:37 | |
| *** gokrokve_ has joined #heat | 18:38 | |
| *** lipinski has joined #heat | 18:40 | |
| openstackgerrit | Jason Dunsmore proposed a change to openstack/heat: Add show_deleted parameter to stack-list API https://review.openstack.org/80312 | 18:40 |
| *** gokrokve has quit IRC | 18:41 | |
| *** gokrokve_ has quit IRC | 18:43 | |
| *** duncanjw has quit IRC | 18:46 | |
| *** gokrokve has joined #heat | 18:51 | |
| *** jcru has quit IRC | 18:55 | |
| yogesh | hello, i need help on a basic template question | 18:57 |
| yogesh | can i create the resources conditionally? | 18:57 |
| yogesh | basically if a supplied parameter has value 1 then create single compute resource and if 2 then create 2 of them | 18:57 |
| *** jcru has joined #heat | 18:58 | |
| *** jstrachan has quit IRC | 19:00 | |
| *** david-lyle has quit IRC | 19:00 | |
| sdake_ | yogesh for that case you can using a scaling group and set the minimum scaling group size to the parameter of the # of nodes | 19:01 |
| yogesh | sdake_: thanks, lemme try it out | 19:04 |
| yogesh | is there a way to have something like a "case" or "if else" in the template...i guess no, but just to ensure | 19:05 |
| sdake_ | there are no conditionals, hot is declarative not imperative | 19:05 |
| *** e0ne has quit IRC | 19:05 | |
| yogesh | ok... | 19:06 |
| yogesh | lemme check out on scaling gorups | 19:06 |
| yogesh | groups* | 19:06 |
| yogesh | thanks much | 19:06 |
| *** e0ne has joined #heat | 19:06 | |
| sjmc7 | there is a CFN conditional construct | 19:06 |
| sjmc7 | no plans to add something similar? | 19:06 |
| sdake_ | sjmc7 there is no analog in heat for the cfn conditional construction | 19:06 |
| sdake_ | sjmc7 hard to predict - I'd not be keen for it but some others in the core team are interested in such a construct | 19:07 |
| yogesh | it surely would be helpful... | 19:07 |
| sdake_ | I see conditionals as a slippery slope which must be avoided :) | 19:07 |
| sjmc7 | :) | 19:07 |
| sjmc7 | can't fight progress? :) | 19:08 |
| yogesh | adds to flexibility | 19:08 |
| sdake_ | i could see how that could lead to a "ok now we want python as the HOT dsl" | 19:08 |
| *** e0ne has quit IRC | 19:08 | |
| sdake_ | I understand its useful, but its not declarative :) | 19:08 |
| sjmc7 | i think while it's in CFN there'll continue to be arguments for it in heat but i can see why you're wary | 19:08 |
| *** tspatzier has quit IRC | 19:09 | |
| sdake_ | hot is tidy and simple to learn - the intrinsics are easy to understand | 19:10 |
| sdake_ | this is what makes hot very appealing | 19:10 |
| sdake_ | adding a bunch of language features makes it less attractive from the "how long does it take to learn" POV | 19:10 |
| *** john-n-seattle has quit IRC | 19:11 | |
| sdake_ | conditionals imo belong in a higher level component (workflow perhaps?) | 19:12 |
| *** john-n-s- has quit IRC | 19:13 | |
| yogesh | +1 | 19:13 |
| yogesh | they sure do | 19:13 |
| *** john-n-seattle has joined #heat | 19:16 | |
| *** andrew_plunk has quit IRC | 19:16 | |
| *** john-n-s- has joined #heat | 19:18 | |
| *** Ajaeger1 has joined #heat | 19:22 | |
| Ajaeger1 | Hi, the translation jobs currently are broken since we have broken po files. Could you review and approve this, please? https://review.openstack.org/#/c/83960/ | 19:22 |
| *** bgorski has quit IRC | 19:27 | |
| stevebaker | Ajaeger1: done | 19:28 |
| Ajaeger1 | Thanks! | 19:30 |
| asalkeld | jasond, you got a moment? | 19:32 |
| jasond | asalkeld: sure | 19:34 |
| *** sballe has joined #heat | 19:34 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack/heat: Check that all po/pot files are valid https://review.openstack.org/84226 | 19:40 |
| *** gokrokve has quit IRC | 19:40 | |
| Ajaeger1 | This patch ^ gates heat against introducing the breakage from 83960 again. | 19:40 |
| *** gokrokve has joined #heat | 19:40 | |
| *** gokrokve_ has joined #heat | 19:42 | |
| *** e0ne has joined #heat | 19:44 | |
| *** gokrokve has quit IRC | 19:45 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack/heat: Check that all po/pot files are valid https://review.openstack.org/84226 | 19:45 |
| *** david-lyle has joined #heat | 19:47 | |
| *** radez is now known as radez_g0n3 | 19:49 | |
| *** Ajaeger1 has quit IRC | 19:59 | |
| stevebaker | SpamapS: so I have a working collector which depends on heatclient and keystoneclient, but I'm having trouble integrating that with test_collect since that assumes that all collectors can be mocked out by passing in an alternate requests_impl | 20:00 |
| *** pablosan is now known as zz_pablosan | 20:01 | |
| *** zns_ has joined #heat | 20:06 | |
| *** Tross has left #heat | 20:08 | |
| *** zns has quit IRC | 20:09 | |
| *** Tross has joined #heat | 20:09 | |
| *** Tross has left #heat | 20:09 | |
| *** akuznetsov has quit IRC | 20:13 | |
| lifeless | uhm | 20:15 |
| lifeless | what does heat stack-delete testenv | 20:15 |
| lifeless | ERROR: Failed to stop stack (testenv) on other engine (9bb3ee47-79af-44c7-8d6b-49110c9fea5f) | 20:15 |
| lifeless | mean when I only have one engine running | 20:15 |
| lifeless | ps fax | grep heat-engine | 20:16 |
| lifeless | 4097 pts/2 S+ 0:00 \_ grep --color=auto heat-engine | 20:16 |
| lifeless | 32302 ? Ss 0:32 /opt/stack/venvs/heat/bin/python /opt/stack/venvs/heat/bin/heat-engine | 20:16 |
| lifeless | root@undercloud-notcompute-jws3awlsb2kh:/opt/stack/tripleo-image-elements/elements/tripleo-cd# | 20:16 |
| lifeless | stevebaker: ^ | 20:16 |
| stevebaker | lifeless: what does heat stack-list show for the state of that stack> | 20:17 |
| stevebaker | ? | 20:17 |
| lifeless | | b7ff4cb2-0e1c-46a4-8808-542cfe7d3a2d | testenv | CREATE_IN_PROGRESS | 2014-03-31T19:26:35Z | | 20:18 |
| lifeless | I have restarted heat-engine since that timestamp | 20:18 |
| lifeless | stevebaker: ^ | 20:21 |
| stevebaker | lifeless: I think you will need to capture the heat-engine log during the stack-delete call to diagnose this. | 20:21 |
| *** rwsu has joined #heat | 20:22 | |
| lifeless | stevebaker: http://paste.openstack.org/show/74695/http://paste.openstack.org/show/74695/ | 20:22 |
| lifeless | http://paste.openstack.org/show/74695/ | 20:22 |
| *** sballe has quit IRC | 20:24 | |
| *** zz_pablosan is now known as pablosan | 20:24 | |
| zaneb | lifeless: it's _supposed_ to send a ping to your old (dead) heat-engine, and go ahead and delete anyway if (when) it times out | 20:25 |
| lifeless | zaneb: so, I can't get tripleo-ci back up again until I can get past this | 20:25 |
| lifeless | zaneb: is there a knob to permit sysadmins to fix things? | 20:25 |
| zaneb | lifeless: delete the entry in the lock table in the DB? | 20:26 |
| zaneb | jasond: about? ^ | 20:26 |
| jasond | zaneb: what's up? | 20:27 |
| zaneb | jasond: see lifeless's question above | 20:27 |
| lifeless | wow that table grows cruft | 20:27 |
| lifeless | rows dating back to early feb | 20:28 |
| lifeless | truncate table ftw | 20:28 |
| zaneb | lifeless: do you restart Heat in the middle of operations a lot? | 20:28 |
| lifeless | zaneb: of course | 20:28 |
| lifeless | zaneb: deployment of engines cannot in any sane world be interlocked with users using heat | 20:28 |
| shardy | what happens when you restart the nova scheduler, or compute service in the middle of spawning vm's, out of interest? | 20:29 |
| shardy | I'm interested what their recovery strategy is for such a scenario, if it's common | 20:30 |
| jasond | lifeless: how many times did you encounter that? | 20:30 |
| lifeless | shardy: compute service - vms deploying but not deployed *to that node* go to error. Workaround is to disable the compute engine via the api, wait a short while, then deploy, then reenable | 20:30 |
| stevebaker | I thought there would be some sort of timeout on locks for dead engines, but I can't see anything obvious | 20:30 |
| shardy | lifeless: interesting, thanks | 20:30 |
| openstackgerrit | A change was merged to openstack/heat: Open Juno development https://review.openstack.org/84077 | 20:31 |
| lifeless | shardy: ideally nova would just deal, but since there is a reasonable thing that doesn't involve waiting forever, its ok. | 20:31 |
| lifeless | shardy: the scheduler is more of an issue, in principle you run two and just kill them serially. in practice... well, not sure yet. ask me when the HA control plane is well established in tripleo. | 20:32 |
| jasond | stevebaker: they're always confirmed to be alive | 20:32 |
| stevebaker | ok, the call to elif engine_alive looks suspect, since that method only returns None or False | 20:32 |
| stevebaker | service.py line 707 | 20:32 |
| stevebaker | oh, ignore me | 20:33 |
| zaneb | stevebaker: look here: https://github.com/openstack/heat/blob/master/heat/engine/stack_lock.py#L75 | 20:33 |
| sdake | stevebaker I've been all over that part of the code - imo multiengine is in solid shape except it suffers from spit brain | 20:33 |
| stevebaker | brain spit | 20:34 |
| sdake | my brain has split multiple times! | 20:34 |
| zaneb | stevebaker: actually, I'm wrong because it is calling try_acquire(), not acquire | 20:37 |
| zaneb | but the code looks equivalent | 20:38 |
| *** andrew_plunk has joined #heat | 20:39 | |
| zaneb | lifeless: it appears that the other heat-engine is alive enough to respond to pings, but not enough to stop doing whatever it's doing | 20:40 |
| lifeless | zaneb: its dead jim | 20:40 |
| lifeless | zaneb: see my ps listing above | 20:40 |
| zaneb | hmm | 20:41 |
| zaneb | lifeless: if you retry the delete a second time will it work? | 20:41 |
| lifeless | unless ps is *entirely* lying to me | 20:41 |
| lifeless | zaneb: it did not | 20:41 |
| lifeless | zaneb: I tried a few times | 20:41 |
| lifeless | zaneb: then asked here | 20:41 |
| lifeless | zaneb: then truncated the lock table once pointed at that | 20:42 |
| stevebaker | or its not doing anything and it is wedged IN_PROGRESS for some other reason | 20:42 |
| sdake | did the update involve a floating ip | 20:42 |
| lifeless | stevebaker: no, process was entirely gone | 20:42 |
| lifeless | sdake: no | 20:42 |
| sdake | lifeless split brain a possibility? | 20:43 |
| lifeless | sdake: one server | 20:43 |
| sdake | how doe smultiengine play int oit? | 20:43 |
| zaneb | so somehow we're getting True returned from here: https://github.com/openstack/heat/blob/master/heat/engine/stack_lock.py#L43 | 20:43 |
| sdake | I guess I missed the probelm report :) | 20:43 |
| lifeless | sdake: ps fax full | grep heat-engine output is above | 20:43 |
| lifeless | sdake: I restarted the engnie | 20:43 |
| sdake | the dreaded restart | 20:44 |
| zaneb | sdake: the engine was restarted in the middle of a create operation, now the stack cannot be deleted | 20:44 |
| lifeless | well I deleted it now, thanks to deleting the stack lock row | 20:45 |
| sdake | zaneb stacklock only listens on the topic of the engine_id | 20:45 |
| stevebaker | I do that frequently, and it works for me | 20:46 |
| sdake | would make sense a restart would respond with a listen request to the restarted server | 20:46 |
| sdake | the engine_id does not change between runs | 20:46 |
| zaneb | sdake: yes it does | 20:46 |
| lifeless | it doesn't? | 20:46 |
| lifeless | how is the id persisted ? | 20:47 |
| sdake | moment let me double check the code | 20:47 |
| *** sgordon has quit IRC | 20:47 | |
| stevebaker | generate_engine_id | 20:47 |
| jasond | sdake: yes, every engine process has a unique id | 20:47 |
| sdake | yup every engine is unique | 20:48 |
| sdake | my bad | 20:48 |
| jasond | if it's restarted, it will have a new uuid | 20:48 |
| sdake | one possibility is somehow the exchanges got confused | 20:48 |
| sdake | eg, amqp problem | 20:48 |
| stevebaker | maybe "listening" should assert a matching engine id before returning True | 20:49 |
| andersonvom | hey folks, is it possible to use functions (e.g. str_replace) in parameters? | 20:49 |
| lifeless | it was, it broke | 20:49 |
| andersonvom | lifeless: as in "it needs to be fixed" or "we don't really want that anymore"? | 20:50 |
| *** e0ne has quit IRC | 20:50 | |
| *** e0ne has joined #heat | 20:51 | |
| stevebaker | andersonvom: it used to work by accident, then it stopped working as a side-effect of other changes. It could be enabled again once we make the concious decision to make it part of the HOT spec | 20:51 |
| zaneb | extremely curious what y'all are using that for | 20:52 |
| * zaneb had not heard of this | 20:52 | |
| jasond | lifeless: if you see this happen again, please see which engine or engines are subscribed to the amqp queue named for the engine ID referenced in the StopActionFailed message. so in the error case above, it would've been a queue named "9bb3ee47-79af-44c7-8d6b-49110c9fea5f" | 20:52 |
| andersonvom | stevebaker, lifeless: thanks! =) | 20:53 |
| lifeless | jasond: how can a dead engine be subscribed? | 20:54 |
| lifeless | andersonvom: I believe it was considered ok to add back, but needs a) work and b) tests. | 20:55 |
| lifeless | zaneb: we used it to centralise calculated values - basically as variables | 20:55 |
| jasond | lifeless: it wouldn't be, but maybe another engine was subscribed to the dead engine's queue. not sure how that would happen, but i can't think of another explanation for what you saw | 20:55 |
| *** e0ne has quit IRC | 20:55 | |
| *** jdob has quit IRC | 20:56 | |
| andersonvom | lifeless: fair enough! =P | 20:56 |
| *** pablosan is now known as zz_pablosan | 20:57 | |
| *** harlowja is now known as harlowja_away | 21:00 | |
| openstackgerrit | Chmouel Boudjnah proposed a change to openstack/heat: Skip tests when loading plugins https://review.openstack.org/84117 | 21:01 |
| lifeless | jasond: well, I can check. what output do you want? | 21:01 |
| jasond | lifeless: are there any consumers for queue "9bb3ee47-79af-44c7-8d6b-49110c9fea5f"? | 21:03 |
| stevebaker | http://tarballs.openstack.org/heat/heat-2014.1.rc1.tar.gz | 21:03 |
| stevebaker | juno is open | 21:03 |
| stevebaker | keep an eye out for fixes which would justify an rc2 | 21:04 |
| *** sgordon has joined #heat | 21:05 | |
| SpamapS | stevebaker: I see. | 21:06 |
| stevebaker | SpamapS: I could: | 21:06 |
| stevebaker | 1. do the keystone and heat calls with raw requests | 21:06 |
| SpamapS | I like that for one simple reason: less dependencies to install. :) | 21:07 |
| SpamapS | but | 21:07 |
| SpamapS | I have a better idea | 21:07 |
| stevebaker | 2. refactor test_collect to use a factory method from test_* instead of assuming it needs a requests_impl | 21:07 |
| SpamapS | why don't we use one-off or signed urls? | 21:07 |
| SpamapS | (in juno.. I get that we don't have that option right now ;) | 21:07 |
| SpamapS | I've always been a bit confused why we don't just pass in a secret UUID that the agent would then use to establish secure comms. | 21:08 |
| shardy | SpamapS: because all requests are authenticated via keystone | 21:08 |
| shardy | SpamapS: essentially we are doing that via a random uuid password | 21:09 |
| stevebaker | but with real keystone users | 21:09 |
| shardy | If oauth support ever lands in keystoneclient, we could add the option to use oauth signed urls | 21:09 |
| SpamapS | shardy: right, so I'm saying that for an in-instance agent, not having to be able to reach keystone is a win. | 21:10 |
| SpamapS | oauth would be cool yeah | 21:10 |
| shardy | SpamapS: but then we have to worry about a whole load of stuff that keystone currently handles for us | 21:11 |
| SpamapS | shardy: this is super super super narrow focus | 21:11 |
| SpamapS | but anyway, keystoneclient and heatclient vs. building your own requests.. | 21:12 |
| SpamapS | I did raw requests for ec2 and cfn because "F boto". | 21:12 |
| SpamapS | But keystoneclient and heatclient are somewhat sane compared to boto :) | 21:12 |
| stevebaker | yeah, but its impractical to swap out their requests usage | 21:13 |
| *** pafuent has left #heat | 21:13 | |
| shardy | SpamapS: I agree that keystone currently lacks some features we'd like to improve the in-instance user use-case, but short-circuiting keystone and doing it all ourselves is not the answer, IMO | 21:14 |
| shardy | SpamapS: anyway, I proposed a summit session so we can discuss the roadmap for this stuff :) | 21:15 |
| SpamapS | shardy: it isn't doing it all ourselves, it is doing it in heat-engine to simplify in-instance tools. | 21:15 |
| SpamapS | much like the current cfn signals work | 21:15 |
| SpamapS | which I think is "tha bomb" | 21:15 |
| SpamapS | except the XML part ;) | 21:15 |
| shardy | And maintaining another special variety of auth middleware, and dealing with how to revoke the secrets if they are leaked and end up known outside the stack, doing our own RBAC to limit the API surface accessible etc etc | 21:17 |
| stevebaker | SpamapS: focus! ;) | 21:17 |
| SpamapS | I'd hope that it would just be oauth or something. | 21:17 |
| SpamapS | Not special. | 21:17 |
| SpamapS | Also for all I care we can shove them in swift and hand out tempurls :) | 21:17 |
| SpamapS | stevebaker: I think it would be fine to refactor the tests a bit to be able to pass an arbitrary "client impl" | 21:18 |
| shardy | SpamapS: Yeah, https://blueprints.launchpad.net/heat/+spec/oauth-credentials-resource | 21:18 |
| shardy | The client support has been blocked for $many cycles afaict | 21:18 |
| SpamapS | stevebaker: so instead of "requests" it is just "the thing that works like the collector expects a client to work" | 21:18 |
| *** harlowja_away is now known as harlowja | 21:19 | |
| sdake_ | \o/ rc1 released ;) | 21:19 |
| SpamapS | well done | 21:19 |
| SpamapS | stevebaker: ^5 | 21:19 |
| stevebaker | SpamapS: 5^ | 21:19 |
| sdake_ | too bad its not friday, i'd make a ^ joke | 21:20 |
| stevebaker | SpamapS: ok, I'll hack the tests | 21:20 |
| SpamapS | anyway, any scheme that requires instances to reach keystone will create extra problems for operators who run with tenant isolated networks. | 21:20 |
| SpamapS | They'll already be struggling to get heat-api traffic out | 21:20 |
| *** e0ne has joined #heat | 21:21 | |
| *** zz_pablosan is now known as pablosan | 21:22 | |
| SpamapS | Feels like neutron-metadata-agent will need to become a little bit smarter... otherwise networks with no gateway will have no way to even be routed to keystone and heat. | 21:22 |
| *** e0ne has quit IRC | 21:23 | |
| *** alexheneveld has quit IRC | 21:24 | |
| *** e0ne has joined #heat | 21:24 | |
| *** zns_ has quit IRC | 21:25 | |
| *** e0ne has quit IRC | 21:26 | |
| *** sgordon has quit IRC | 21:27 | |
| andersonvom | stevebaker: when will feature freeze be lifted? | 21:30 |
| stevebaker | andersonvom: now | 21:30 |
| SpamapS | +A the world!! | 21:33 |
| *** arosen1 has joined #heat | 21:34 | |
| *** pablosan is now known as zz_pablosan | 21:34 | |
| *** alexheneveld has joined #heat | 21:36 | |
| *** Slower_ has quit IRC | 21:38 | |
| *** rpothier has quit IRC | 21:39 | |
| *** Slower has joined #heat | 21:39 | |
| *** vijendar has quit IRC | 21:41 | |
| *** aweiteka has quit IRC | 21:46 | |
| andersonvom | stevebaker: oooh, that was quick! :P | 21:48 |
| *** duncanjw has joined #heat | 21:48 | |
| andersonvom | stevebaker: can we lift this -2 then? https://review.openstack.org/#/c/58641/ :D | 21:50 |
| *** andrew_plunk has quit IRC | 21:50 | |
| stevebaker | andersonvom: any abandoned changes will need to be un-abandoned before -2s can be raised | 21:51 |
| andersonvom | whoops | 21:53 |
| andersonvom | stevebaker: done | 21:53 |
| stevebaker | andersonvom: done | 21:54 |
| *** tnurlygayanov_ has quit IRC | 21:55 | |
| *** tnurlygayanov has joined #heat | 21:56 | |
| *** dims has quit IRC | 21:57 | |
| *** e0ne has joined #heat | 21:59 | |
| andersonvom | stevebaker: danke schon! =) | 22:02 |
| *** packet has quit IRC | 22:02 | |
| stevebaker | andersonvom: alstublieft | 22:03 |
| *** zz_pablosan is now known as pablosan | 22:04 | |
| *** zns has joined #heat | 22:10 | |
| *** dims has joined #heat | 22:13 | |
| *** e0ne has quit IRC | 22:13 | |
| arosen1 | does anyone have a heat template that does anything with docker off hand? | 22:14 |
| *** rcleere has quit IRC | 22:20 | |
| *** lipinski has quit IRC | 22:21 | |
| *** lipinski has joined #heat | 22:21 | |
| *** lipinski has quit IRC | 22:26 | |
| *** jcru has quit IRC | 22:27 | |
| *** yogesh has quit IRC | 22:28 | |
| openstackgerrit | Anderson Mesquita proposed a change to openstack/heat: Check resources are in expected states https://review.openstack.org/58641 | 22:36 |
| andersonvom | sweet, no merge conflicts! =) | 22:36 |
| arosen1 | I'm trying to use docker with heat. I set the following value in /etc/heat/heat.conf plugin_dirs=/opt/stack/heat/contrib/docker/docker/resources/ though when I load my docker template h-eng is returning: | 22:40 |
| arosen1 | 22:40 | |
| arosen1 | ERROR: Unknown resource Type : OS::Heat::Docker | 22:41 |
| arosen1 | Anyone know off hand where I might be going wrong? | 22:41 |
| stevebaker | arosen1: it may have been renamed recently | 22:41 |
| stevebaker | arosen1: DockerInc::Docker::Container | 22:42 |
| arosen1 | stevebaker: so OS::Heat::Docker --> OS::Heat::DockerInc::Docker::Container | 22:42 |
| stevebaker | arosen1: no, OS::Heat::Docker -> DockerInc::Docker::Container | 22:43 |
| *** asalkeld has quit IRC | 22:44 | |
| arosen1 | stevebaker: thanks | 22:45 |
| *** tango has quit IRC | 22:49 | |
| *** sjmc7 has quit IRC | 22:52 | |
| *** alexpilotti has quit IRC | 23:11 | |
| arosen1 | I'm playing around with heat and docker. When I try to run this wordpress_mysql.yml file http://codepad.org/Dwp7adAZ i get the following error. It looks like it's trying to launch docker on the node running heat-engine? This should'nt be something that someone is trying to do though right? | 23:12 |
| *** duncanjw has quit IRC | 23:18 | |
| *** TravT has quit IRC | 23:21 | |
| *** david-lyle has quit IRC | 23:30 | |
| *** zns has quit IRC | 23:32 | |
| stevebaker | arosen1: You need to specify the docker_endpoint property http://docs.openstack.org/developer/heat/template_guide/contrib.html#DockerInc::Docker::Container | 23:32 |
| stevebaker | arosen1: that resource is not at all integrated with keystone | 23:32 |
| *** kgriffs is now known as kgriffs|afk | 23:32 | |
| arosen1 | stevebaker: thanks. Another quick question. I'm using this heat template which is supposed to boot a vm then install docker on top of it then launch a cirros image in it: http://codepad.org/nhujso76 | 23:36 |
| arosen1 | though it seems to be failing on the user_data: section. I'm guessing that, #include https://get.docker.io doesn't install docker for you? I found that here: http://blog.docker.io/category/openstack/ | 23:37 |
| stevebaker | arosen1: try adding user_data_format=RAW | 23:37 |
| stevebaker | arosen1: so that #include will be given directly to cloud-init | 23:38 |
| arosen1 | stevebaker: for the docker endpoint property you're talking about does that endpoint just tie you down to one server? | 23:38 |
| stevebaker | arosen1: yes, until there is a container API which has some form of scheduler you will have to manually decide where every container gets created. Heat will not become a scheduler | 23:40 |
| *** pvaneck has quit IRC | 23:40 | |
| arosen1 | stevebaker: like this: http://codepad.org/N2HMS27R ? | 23:43 |
| stevebaker | arosen1: yes | 23:44 |
| stevebaker | arosen1: and I'd be interested to know if that works ;) | 23:44 |
| arosen1 | stevebaker: sure i'll keep you posted. I'm trying to integrate support for heat+docker with neutron though i'm just trying to figure out how things work today. | 23:45 |
| *** rpothier has joined #heat | 23:46 | |
| stevebaker | arosen1: ah, cool | 23:46 |
| arosen1 | stevebaker: no go: http://codepad.org/UeO4c7kP | 23:49 |
| arosen1 | I don't see anything about docker in the cloud-init-log | 23:49 |
| stevebaker | arosen1: ok, next attempt. try user_data: {get_file: https://get.docker.io/} | 23:50 |
| stevebaker | arosen1: this will require a recent python-heatclient | 23:50 |
| arosen1 | stevebaker: and delete user_data_format: RAW? | 23:50 |
| *** arbylee has quit IRC | 23:51 | |
| stevebaker | arosen1: you can leave that | 23:51 |
| arosen1 | stevebaker: the python-heatclient needs to be recent on my desktop or in the guest? | 23:51 |
| stevebaker | arosen1: on your desktop. It actually resolves the url on your local machine | 23:52 |
| arosen1 | stevebaker: i'm running from devstack from today so i should have pretty new stuff. | 23:53 |
| stevebaker | that should be fine | 23:53 |
| arosen1 | It doesn't like this line: user_data: {get_file: https://get.docker.io/} | 23:53 |
| *** david-lyle has joined #heat | 23:53 | |
| arosen1 | in "<unicode string>", line 13, column 29 found unexpected ':' | 23:54 |
| arosen1 | get_file no : maybe | 23:54 |
| *** andersonvom has quit IRC | 23:56 | |
| stevebaker | arosen1: try user_data: {get_file: 'https://get.docker.io/'} | 23:57 |
| *** RockKuo_iPad has joined #heat | 23:57 | |
| *** david-lyle has quit IRC | 23:58 | |
| *** RockKuo_iPad has quit IRC | 23:58 | |
| arosen1 | stevebaker: http://codepad.org/7IXzW6AN | 23:59 |
| *** gokrokve_ has quit IRC | 23:59 | |
| arosen1 | seems like it did something | 23:59 |
| *** gokrokve has joined #heat | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!