*** dbecker has joined #heat | 00:40 | |
*** liyi has joined #heat | 00:47 | |
*** dbecker has quit IRC | 00:54 | |
*** sshnaidm|rover is now known as sshnaidm|afk | 00:56 | |
*** liyi has quit IRC | 00:59 | |
*** liyi has joined #heat | 01:00 | |
*** ricolin_ has joined #heat | 01:10 | |
*** edmondsw has joined #heat | 01:13 | |
openstackgerrit | Hongbin Lu proposed openstack/heat master: zun: use containers.update for renaming https://review.openstack.org/574099 | 01:15 |
---|---|---|
openstackgerrit | Hongbin Lu proposed openstack/heat master: zun: add property 'networks' to container https://review.openstack.org/533426 | 01:15 |
*** yangyapeng has joined #heat | 01:15 | |
*** hongbin has joined #heat | 01:16 | |
*** edmondsw has quit IRC | 01:18 | |
*** gmoro has joined #heat | 01:36 | |
*** gmoro has quit IRC | 01:39 | |
*** gmoro has joined #heat | 01:40 | |
*** liyi has quit IRC | 01:55 | |
*** liyi has joined #heat | 02:01 | |
*** liyi_ has joined #heat | 02:23 | |
*** liyi has quit IRC | 02:27 | |
*** skramaja has joined #heat | 02:55 | |
*** edmondsw has joined #heat | 03:02 | |
*** edmondsw has quit IRC | 03:06 | |
*** jmlowe has quit IRC | 03:10 | |
*** neatherweb has quit IRC | 03:12 | |
*** ramishra has joined #heat | 03:39 | |
*** gmoro has quit IRC | 04:03 | |
*** liyi_ has quit IRC | 04:32 | |
openstackgerrit | Merged openstack/heat master: Restore auth-less version negotiation https://review.openstack.org/574697 | 04:33 |
*** liyi has joined #heat | 04:33 | |
*** hongbin has quit IRC | 04:38 | |
*** mdnadeem has joined #heat | 04:38 | |
*** zzzeek has quit IRC | 04:40 | |
*** zzzeek has joined #heat | 04:43 | |
*** edmondsw has joined #heat | 04:51 | |
*** ysandeep has joined #heat | 04:51 | |
*** edmondsw has quit IRC | 04:55 | |
*** mdnadeem has quit IRC | 05:02 | |
*** zzzeek has quit IRC | 05:10 | |
*** zzzeek has joined #heat | 05:11 | |
*** liyi has quit IRC | 05:19 | |
*** mdnadeem has joined #heat | 05:23 | |
openstackgerrit | Merged openstack/python-heatclient master: Fix log_http_request function in http module https://review.openstack.org/539834 | 05:45 |
*** liyi has joined #heat | 05:45 | |
*** nicolasbock has joined #heat | 05:49 | |
*** liyi has quit IRC | 05:55 | |
*** liyi has joined #heat | 05:56 | |
*** Guest38444 has quit IRC | 05:57 | |
openstackgerrit | Rico Lin proposed openstack/heat master: Remove unrequired migrate check in test https://review.openstack.org/577998 | 06:06 |
openstackgerrit | Merged openstack/heat master: Change NovaClientPlugin to use MicroversionMixin https://review.openstack.org/556810 | 06:10 |
*** armaan has joined #heat | 06:13 | |
*** armaan has quit IRC | 06:15 | |
*** armaan has joined #heat | 06:16 | |
*** armaan has quit IRC | 06:21 | |
*** neatherweb has joined #heat | 06:24 | |
openstackgerrit | Rico Lin proposed openstack/heat master: Support py36 for test https://review.openstack.org/580298 | 06:31 |
*** dyasny has quit IRC | 06:35 | |
openstackgerrit | Merged openstack/heat master: Refactor resource plugins for microversion usage https://review.openstack.org/502961 | 06:42 |
*** dbecker has joined #heat | 06:43 | |
openstackgerrit | tianhui proposed openstack/heat master: Change deprecated url to a correct https://review.openstack.org/580303 | 06:53 |
*** dyasny has joined #heat | 06:54 | |
*** dbecker has quit IRC | 07:05 | |
*** yangyapeng has quit IRC | 07:10 | |
*** gkadam has joined #heat | 07:11 | |
*** sshnaidm|afk has quit IRC | 07:14 | |
*** ysandeep has quit IRC | 07:18 | |
*** sshnaidm|afk has joined #heat | 07:21 | |
*** peereb has joined #heat | 07:22 | |
*** rgold has joined #heat | 07:22 | |
*** rgold has quit IRC | 07:22 | |
*** yangyapeng has joined #heat | 07:32 | |
*** yangyapeng has quit IRC | 07:34 | |
openstackgerrit | Rico Lin proposed openstack/heat master: Add StoryBoard migration docs and releasenote https://review.openstack.org/566428 | 07:35 |
*** Guest38444 has joined #heat | 07:38 | |
*** ysandeep has joined #heat | 07:39 | |
*** rcernin has quit IRC | 07:54 | |
*** liyi_ has joined #heat | 08:08 | |
*** liyi has quit IRC | 08:08 | |
*** sshnaidm|afk is now known as sshnaidm|rover | 08:16 | |
*** edmondsw has joined #heat | 08:28 | |
*** edmondsw has quit IRC | 08:32 | |
*** derekh has joined #heat | 08:36 | |
*** neatherweb has quit IRC | 08:37 | |
openstackgerrit | Merged openstack/heat master: Update docs bug link to storyboard https://review.openstack.org/570056 | 08:57 |
*** pbourke has quit IRC | 09:13 | |
*** pbourke has joined #heat | 09:15 | |
openstackgerrit | Rico Lin proposed openstack/heat master: Remove mox usage from test_heat_client https://review.openstack.org/557924 | 09:16 |
*** kiennt26 has joined #heat | 09:43 | |
*** armaan has joined #heat | 09:44 | |
*** ricolin_ has quit IRC | 09:57 | |
*** ricolin has joined #heat | 09:57 | |
*** ricolin_ has joined #heat | 09:58 | |
*** ricolin_ has quit IRC | 09:58 | |
*** armaan has quit IRC | 10:06 | |
*** armaan has joined #heat | 10:07 | |
*** armaan has quit IRC | 10:11 | |
*** mdnadeem has quit IRC | 10:34 | |
*** kiennt26 has quit IRC | 10:41 | |
*** mdnadeem has joined #heat | 10:46 | |
*** serlex has joined #heat | 10:53 | |
*** serlex has quit IRC | 10:54 | |
*** ricolin has quit IRC | 11:14 | |
*** serlex has joined #heat | 11:15 | |
*** armaan has joined #heat | 11:22 | |
*** edmondsw has joined #heat | 11:29 | |
*** neatherweb has joined #heat | 11:29 | |
*** sshnaidm|rover is now known as sshnaidm|rov|lnc | 11:33 | |
*** lifeless has quit IRC | 11:45 | |
*** dbecker has joined #heat | 11:49 | |
*** skramaja_ has joined #heat | 11:58 | |
*** skramaja has quit IRC | 11:59 | |
*** jmccarthy has joined #heat | 12:00 | |
jmccarthy | Hiya, anyone know where one might find a small boot-from-vol heat template example ? I've not found one yet that seems to actually work .. | 12:01 |
*** neatherweb has quit IRC | 12:05 | |
*** edmondsw has quit IRC | 12:07 | |
jmccarthy | I'm trying to use this one here http://download.cirros-cloud.net/0.4.0/cirros-0.4.0-x86_64-disk.img | 12:09 |
*** edmondsw has joined #heat | 12:13 | |
*** sshnaidm|rov|lnc is now known as sshnaidm|rover | 12:15 | |
*** edmondsw_ has joined #heat | 12:16 | |
*** neatherweb has joined #heat | 12:17 | |
*** edmondsw has quit IRC | 12:19 | |
therve | jmccarthy: That's not a template? :) | 12:23 |
jmccarthy | therve: lol true :) I meant using that image to create instance and boot it from vol | 12:25 |
therve | jmccarthy: So what are you trying? | 12:25 |
jmccarthy | I have something like this: https://paste.fedoraproject.org/paste/zdi3NftD48u1uOohbXfmiw | 12:26 |
therve | So that template doesn't work right? | 12:28 |
therve | You're mixing various functions | 12:28 |
*** jmlowe has joined #heat | 12:28 | |
jmccarthy | It seems to work ok, and there is avol attached to instance on /dev/vda but complaints about no bootable devices | 12:29 |
jmccarthy | therve: I guess I am, but I don't know what's missing ? | 12:30 |
therve | jmccarthy: Ref isn't a hot functions | 12:31 |
therve | You want to use get_param | 12:31 |
jmccarthy | therve: Ok I changed that to image : { get_param: ImageID } but still says no bootable device ? | 12:34 |
therve | jmccarthy: What do you pass as ImageID? | 12:35 |
jmccarthy | In my case --parameter "ImageID=cirros-0.4.0" which is the the image name in glancec | 12:35 |
*** armaan has quit IRC | 12:46 | |
*** armaan has joined #heat | 12:47 | |
therve | jmccarthy: The volume is created properly? | 12:47 |
jmccarthy | therve: looks to be, it's there, and says it's attached to the instance on /dev/vda | 12:48 |
*** liyi_ has quit IRC | 12:50 | |
therve | Mok | 12:50 |
therve | You could try bdm_v2 and pass the image directly | 12:50 |
*** liyi has joined #heat | 12:51 | |
*** jmlowe has quit IRC | 12:51 | |
jmccarthy | therve: Is there an example of this somewhere ? My searches seem to be finding old/broken stuff | 12:51 |
therve | jmccarthy: https://support.rackspace.com/how-to/building-cloud-orchestration-templates/#boot-a-server-from-a-volume maybe | 12:53 |
jmccarthy | therve: Ok, I think I did something like this earlier .. lemme try that | 12:53 |
*** liyi has quit IRC | 12:56 | |
jmccarthy | therve: Hmm failed due to "stack_status_reason | Resource CREATE failed: ValueError: resources.my_server: nics are required after microversion 2.36" | 12:57 |
therve | Ah. You need to pass some network info I guess | 12:58 |
jmccarthy | therve: Ok, that seems like a step in the right direction, thanks :) | 12:59 |
therve | Yeah not sure about that :) | 12:59 |
*** jmlowe has joined #heat | 13:00 | |
*** armaan has quit IRC | 13:16 | |
*** armaan has joined #heat | 13:17 | |
*** armaan has quit IRC | 13:21 | |
*** ricolin has joined #heat | 13:22 | |
*** armaan has joined #heat | 13:24 | |
*** mdnadeem has quit IRC | 13:38 | |
*** skramaja_ has quit IRC | 13:43 | |
*** mriedem has joined #heat | 13:43 | |
jmccarthy | mriedem: Hiya ! | 13:44 |
mriedem | (8:42:58 AM) mriedem: question about compute api microversions and heat, specifically with OSC | 13:44 |
mriedem | (8:43:18 AM) mriedem: is it possible to create a heat template to create a nova server which has parameters that rely on a specific compute api microversion? | 13:44 |
mriedem | (8:43:23 AM) mriedem: and then pass that through heat to nova? | 13:44 |
mriedem | (8:43:43 AM) mriedem: specifically, the instance resource in the template specifies network: auto for compute api microversion >= 2.37 | 13:44 |
mriedem | zaneb: ^? | 13:44 |
mriedem | would: openstack --os-compute-api-version 2.37 stack create ... | 13:44 |
mriedem | pass that compute api version through? | 13:44 |
mriedem | or is it possible to specify the compute api version in the template itself? | 13:45 |
jmccarthy | mreidem: I was asking about that error here earlier, before I found that bug/starting asking you (fyi) | 13:45 |
zaneb | mriedem: for particular parameters that require a certain microversion, we explicitly use that microversion to handle those parameters | 13:47 |
zaneb | otherwise everything goes through the default (really old) microversion | 13:48 |
zaneb | there's no way to specify the microversion in the template | 13:48 |
zaneb | ramishra is currently working on changes in this area, so he can probably give a more detailed answer | 13:48 |
mriedem | hmm, ok, so if network: auto or network: none show up in the template, heat would have to check for those specific enums and pass >= 2.37 to nova | 13:49 |
mriedem | hopefully no one would ever have an actual network named auto or none | 13:49 |
jmccarthy | zaneb: I'm trying https://support.rackspace.com/how-to/building-cloud-orchestration-templates/#boot-a-server-from-a-volume but getting stack_status_reason | Resource CREATE failed: ValueError: resources.my_server: nics are required after microversion 2.36 - where I am going wrong ? | 13:49 |
ramishra | so https://review.openstack.org/#/c/556810/ merged today, so we use the latest mircoversion supported by the deployment now | 13:49 |
mriedem | jmccarthy: you'll have to add - networks: - network: auto | 13:50 |
mriedem | but heat isn't passing the correct microversion on your older deployment it sounds like | 13:50 |
ramishra | of there is a property that needs a microversion higher than what's supported in your deployment then it would give you a validation error | 13:50 |
ramishra | s/of/if | 13:50 |
jmccarthy | mriedem: I got an error with 'auto' tho ? Let me try again | 13:50 |
mriedem | 2.37 was added in mitaka | 13:50 |
mriedem | jmccarthy: b/c heat isn't passing 2.37 | 13:50 |
mriedem | jmccarthy: it's passing 2.1 because it doesn't realize that network: auto is a special thing | 13:51 |
*** cliffparsons has joined #heat | 13:51 | |
mriedem | so, it sounds like one way to fix this is to add logic to heat such that if any network in the instance resource is 'auto' or 'none' enums, specify at least compute api version 2.37 | 13:51 |
jmccarthy | mreidem: I'm using queens | 13:51 |
ramishra | mriedem: that was the case earlier, now we've fixed it | 13:51 |
mriedem | which could be a backportable change | 13:51 |
mriedem | ramishra: i'm assuming your change isn't backportable? | 13:51 |
jmccarthy | ERROR: Property error: : resources.my_server.properties.networks[0].network: : Error validating value 'auto': Unable to find network with name or id 'auto' | 13:52 |
mriedem | jmccarthy: right because nova < 2.37 doesn't recognize auto or none as enums | 13:52 |
ramishra | mriedem: We can as it's linked to a bug. It was in review for long time though | 13:52 |
mriedem | oh....and the template format needs to change... | 13:52 |
jmccarthy | mriedem: Sorry if I'm not following, is there something I can do so it'll work with what I have at the moment with queens ? | 13:52 |
mriedem | jmccarthy: the template has networks as a list | 13:52 |
mriedem | so you're doing something like: | 13:53 |
mriedem | networks: | 13:53 |
mriedem | - network: auto | 13:53 |
mriedem | that ends up passing {'server': {'networks': ['auto']}} to nova | 13:53 |
mriedem | but auto is an enum not for the list, so it should be {'server': {'networks': 'auto'}} | 13:53 |
mriedem | which means the template would have to have: | 13:54 |
mriedem | networks: auto | 13:54 |
mriedem | or something like that, but heat probably doesn't allow that format today | 13:54 |
mriedem | that could be the indication, however, to heat that it needs to pass that format for >= 2.37 | 13:54 |
mriedem | jmccarthy: so i guess trying changing your template to "networks: auto" and see what happens | 13:55 |
mriedem | does heat reject that? | 13:56 |
mriedem | or does nova reject it? | 13:56 |
mriedem | ramishra: passing the latest microversion available in the server seems dangerous | 13:56 |
jmccarthy | mriedem: I tried setting the actual name of the network, which seems to create the stack ok (I'm not sure if the instance is ok or not yet) | 13:56 |
mriedem | and is going to be inconsistent across clouds at different versions | 13:56 |
ramishra | mriedem: it's not the latest, it checks for what's available in the cloud | 13:57 |
mriedem | jmccarthy: ok yeah is essentially what auto would do under the covers, or if you don't specify a network pre-2.37 - nova will find the single private tenant network or the single public network and use that | 13:57 |
mriedem | ramishra: right, it's the max version available in the cloud | 13:57 |
mriedem | but still, | 13:57 |
mriedem | my template is going to potentially behave differently between newton and queens clouds | 13:58 |
ramishra | mriedem: probably not, as we've property specific checks, we may be missing some though | 13:59 |
jmccarthy | mriedem: trying changing your template to "networks: auto" and see what happen, gives ERROR: Internal Error | 13:59 |
jmccarthy | mriedem: I'll go back to specifiying that actual name | 13:59 |
jmccarthy | s/that/the | 13:59 |
mriedem | jmccarthy: ok that seems like a gap in heat then, but i'm not sure how easy that is to fix | 14:00 |
mriedem | but basically heat doesn't support the get me a network feature | 14:00 |
mriedem | jmccarthy: want to open a heat bug that i could dump some words into? | 14:00 |
jmccarthy | mriedem: Sure one sec | 14:01 |
jmccarthy | um 'OpenStack Heat must be configured in order for Launchpad to forward bugs to the project's developers' where are these raised ? | 14:02 |
mriedem | oh they use storyboard now | 14:02 |
mriedem | https://storyboard.openstack.org/#!/project/989 | 14:03 |
ricolin | jmccarthy, https://storyboard.openstack.org/#!/project/989 | 14:03 |
mriedem | so create new > story > project = openstack/heat | 14:03 |
mriedem | search doesn't work in storyboard so i can't see if there is an existing issue for this | 14:03 |
jmccarthy | mriedem: Ok, not sure what to put here for Story Title, Task title hmm | 14:05 |
jmccarthy | mreidem: Can whatever I put in be edited afterwards ? | 14:06 |
mriedem | want me to open it and you can fill in error stuff? | 14:06 |
ricolin | mriedem, yeah, storyboard still working on that, one way to search it is through typing your keyword in `Jump to` bar on top right, that will help you find stories contain that keyword in title | 14:06 |
jmccarthy | mriedem: Issue with network definition in heat template - https://storyboard.openstack.org/#!/story/2002875 I stuck in .. | 14:07 |
jmccarthy | ricolin: Thanks I logged in, haven't tried this yet ! | 14:08 |
ricolin | jmccarthy, NP:) | 14:09 |
mriedem | jmccarthy: updated | 14:13 |
mriedem | gdi, storyboard dropped all of my formatting | 14:13 |
jmccarthy | mriedem: Thanks ! | 14:25 |
*** armaan has quit IRC | 14:29 | |
therve | zaneb: OK so I'm mostly concerned about that change: https://review.openstack.org/#/c/578393/7/heat/api/openstack/v1/stacks.py | 14:36 |
therve | It adds a new argument to various stack operations | 14:36 |
therve | IIUC it's then used here: https://review.openstack.org/#/c/578392/5/heat/engine/service.py | 14:37 |
therve | When that parameter is passed, we load the creds from it, and create a context with those | 14:38 |
zaneb | therve: oh. that's weird. and probably unsafe. I don't know why we would ever do that. | 14:40 |
therve | Ah :) | 14:41 |
therve | zaneb: http://eavesdrop.openstack.org/irclogs/%23heat/%23heat.2018-07-04.log.html#t2018-07-04T14:52:18 | 14:41 |
zaneb | therve: that makes no sense to me | 14:42 |
therve | That reassures me somehow :) | 14:43 |
ricolin | therve, only create/update stack will get the args from https://review.openstack.org/#/c/578393/7/heat/api/openstack/v1/stacks.py | 14:43 |
zaneb | the whole new table thing does make more sense in that context though | 14:46 |
therve | Yes | 14:47 |
therve | That patch could become way simpler without all that | 14:47 |
therve | ricolin: Yeah but even then? | 14:48 |
*** ayoung has quit IRC | 14:52 | |
*** liyi has joined #heat | 14:52 | |
ricolin | can we put an extra UUID in context and bring it cross two OpenStack site(through API and client)? | 14:53 |
ricolin | therve, zaneb ^^^ | 14:53 |
*** liyi has quit IRC | 14:56 | |
*** ayoung has joined #heat | 15:03 | |
ricolin | therve, zaneb ramishra another question is we now got few ways to achieve remote stack, think it will be better if we can clarify which path we do desire to own. For now we got 3 properties, `provider_cred`, `cred` and `provider_name`, 1. with `cred` provided, the context will refresh in remote site, and remote site will keep using that `cred`(and read authN info from remote Barbican) for rest of stack action. 2. if `provider_cred` | 15:03 |
ricolin | provided, It will read authN info (including auth_url) and use it to generate context to access remote stack( and will use it to create remote stack if `cred` not provided). and 3. If `provider_name` provided, will read authN info from clouds.yaml and use it to access to remote stack(and create if `cred` not provided) | 15:03 |
*** mriedem has left #heat | 15:04 | |
ricolin | option 3 might not be a desired way, since it's kind of share concern like ssl. It allow user to access to a local file for information just by provied name | 15:05 |
jmccarthy | How do things in storyboard get reviewed, will https://storyboard.openstack.org/#!/story/2002875 be triaged and assigned, or I should assign it to someone ? | 15:06 |
*** hongbin has joined #heat | 15:08 | |
therve | ricolin: Why would we need an extra UUID in context? | 15:08 |
*** armaan has joined #heat | 15:08 | |
zaneb | ricolin: 2 is the only one we _need_ | 15:13 |
ricolin | therve, that's for option 1 above | 15:14 |
therve | Yes what Zane says | 15:15 |
ricolin | zaneb, we also need to figure out what we need to deal with ssl for remote stack | 15:16 |
therve | Well ideally, nothing :) | 15:17 |
ricolin | therve, I mean this patch https://review.openstack.org/#/c/480923/ | 15:18 |
zaneb | ricolin: how does clouds.yaml deal with it? | 15:18 |
ricolin | zaneb, something like this I think https://docs.openstack.org/os-client-config/latest/user/configuration.html#ssl-settings | 15:19 |
zaneb | ugh, so it's still a reference to a local file :( | 15:20 |
zaneb | so I'd say we do probably want to do something like https://review.openstack.org/#/c/480923/ except passing the actual cert (using get_file or whatever) rather than looking for the filename locally | 15:21 |
zaneb | that'd work for both multi-region and multi-cloud | 15:21 |
*** peereb has quit IRC | 15:22 | |
ricolin | one why is to keep local file, but we try to find way to record which region, provider get to use which cert file, another way is to directly ask user to pass down cert by their own(which is what you proposed) | 15:24 |
ricolin | zaneb, | 15:24 |
*** dyasny has quit IRC | 15:25 | |
*** bnemec has quit IRC | 15:25 | |
*** dyasny has joined #heat | 15:26 | |
zaneb | ricolin: it's not a bad idea to use a local clouds.yaml that already has the most commonly used stuff, so you only need to pass your own if you want to use something different | 15:27 |
zaneb | maybe we could even look up the auth_url in that file instead of indexing by cloud 'name', so that we wouldn't have to add a separate property | 15:28 |
zaneb | I do have some slight interoperability concerns | 15:28 |
ricolin | isn't that kind of lead us to option 3 above? | 15:29 |
*** rgold has joined #heat | 15:30 | |
*** crag_ has joined #heat | 15:35 | |
zaneb | yeah, that's effectively 2 + 3 at that point | 15:36 |
zaneb | we _need_ 2. 3 would make it easier to use in some common cases, at a cost of some interoperability | 15:37 |
* ricolin try to figure out if there is a third way to deal with this | 15:39 | |
*** crag_ has quit IRC | 15:44 | |
*** bnemec has joined #heat | 15:56 | |
*** ysandeep has quit IRC | 15:56 | |
*** jmlowe has quit IRC | 16:01 | |
openstackgerrit | wu.chunyang proposed openstack/heat-dashboard master: Add release note link in README https://review.openstack.org/580439 | 16:12 |
*** dyasny has quit IRC | 16:12 | |
*** dyasny has joined #heat | 16:13 | |
*** Guest38444 has quit IRC | 16:13 | |
*** ricolin has quit IRC | 16:43 | |
*** armaan_ has joined #heat | 16:45 | |
*** armaan has quit IRC | 16:45 | |
*** serlex has quit IRC | 16:49 | |
*** armaan_ has quit IRC | 16:50 | |
*** bnemec has quit IRC | 16:51 | |
*** liyi has joined #heat | 16:53 | |
*** liyi has quit IRC | 16:57 | |
flwang1 | zaneb: any lucky you're around? | 17:02 |
*** derekh has quit IRC | 17:02 | |
*** jmccarthy has quit IRC | 17:10 | |
*** ramishra has quit IRC | 17:13 | |
*** bnemec has joined #heat | 17:18 | |
*** ysandeep has joined #heat | 17:21 | |
*** serlex has joined #heat | 17:23 | |
*** sshnaidm|rover has quit IRC | 17:43 | |
*** jmlowe has joined #heat | 17:43 | |
*** FL1SK has joined #heat | 17:44 | |
*** sshnaidm|rover has joined #heat | 17:52 | |
*** serlex has quit IRC | 17:59 | |
*** ysandeep has quit IRC | 18:03 | |
zaneb | flwang1: o/ | 18:12 |
flwang1 | zaneb: mind revisiting this https://review.openstack.org/579669 ? | 18:12 |
flwang1 | we do need it for queens and especially magnum | 18:13 |
flwang1 | zaneb: do I need a change at here https://github.com/openstack/heat-agents/blob/master/heat-config/os-apply-config/etc/os-collect-config.conf#L37 and https://github.com/openstack/heat-agents/blob/master/heat-config/os-apply-config/etc/os-collect-config.conf#L47 ? Thanks | 18:13 |
zaneb | oh yeah, I had that tab open somewhere | 18:13 |
openstackgerrit | Merged openstack/heat master: Handle resource plugins without default_client_name https://review.openstack.org/537708 | 18:13 |
zaneb | flwang1: I suspect you might. your change will help with signalling back to the server, but you'll probably need to do something there to make sure it can get the data in the first place | 18:14 |
zaneb | there might be another change required in Heat to populate that in the cloud-init data as well | 18:15 |
flwang1 | zaneb: yep, that's my main concern | 18:15 |
flwang1 | ah? another change in heat? | 18:15 |
flwang1 | where? | 18:15 |
flwang1 | that's shocking me, TBH, why the bug is there for years? :( | 18:16 |
zaneb | because ~nobody uses multiple regions? | 18:17 |
zaneb | (looking for the place now) | 18:17 |
*** schaney has joined #heat | 18:19 | |
flwang1 | zaneb: but that doesn't make sense saying nobody using multi regions for heat :( | 18:20 |
flwang1 | based on my understanding, with current code, the heat agent will try to use the url_for to get the heat URL, but because there is no region given, so keystone will return one randomly | 18:20 |
zaneb | flwang1: it would be in this method here: http://git.openstack.org/cgit/openstack/heat/tree/heat/engine/resources/server_base.py#n64 | 18:28 |
flwang1 | zaneb: ah, yep | 18:29 |
flwang1 | zaneb: btw, is there any easy way to test it? | 18:29 |
zaneb | no :) | 18:29 |
flwang1 | :( | 18:29 |
flwang1 | ok, thank you so much, Zane. i will propose another patch for heat | 18:32 |
zaneb | if there were, that bug probably wouldn't have been around for so long :/ | 18:32 |
*** armaan has joined #heat | 18:32 | |
flwang1 | that does make sense for me ;) | 18:32 |
zaneb | :D | 18:32 |
zaneb | flwang1: I +2d your heat-agents patch and added a comment summarising what we just discussed | 18:32 |
flwang1 | zaneb: for the heat-agents patch, shouldn't we wait to add the change for https://github.com/openstack/heat-agents/blob/master/heat-config/os-apply-config/etc/os-collect-config.conf#L37 and https://github.com/openstack/heat-agents/blob/master/heat-config/os-apply-config/etc/os-collect-config.conf#L47 ? | 18:33 |
flwang1 | or you think we better to fix it in heat first? | 18:33 |
zaneb | I don't think it matters | 18:33 |
zaneb | you have to _get_ the software config data before you can even think about signalling back, so that change won't even really take effect until we do the other ones | 18:35 |
flwang1 | zaneb: does that mean even without the change for http://git.openstack.org/cgit/openstack/heat/tree/heat/engine/resources/server_base.py#n64 and https://github.com/openstack/heat-agents/blob/master/heat-config/os-apply-config/etc/os-collect-config.conf#L37 and https://github.com/openstack/heat-agents/blob/master/heat-config/os-apply-config/etc/os-collect-config.conf#L47 heat-agent can already get the "deploy_region_name" from the /var/run/ | 18:36 |
zaneb | no, I mean that until you make those changes any server in a non-default region won | 18:37 |
zaneb | won't be able to download its software deployments | 18:37 |
flwang1 | :( | 18:38 |
zaneb | so it doesn't matter what they might do after they did manage to download and run their sw deployments, because they won't | 18:38 |
flwang1 | ok, so i do still need to propose a fix for http://git.openstack.org/cgit/openstack/heat/tree/heat/engine/resources/server_base.py#n64 ? and the corresponding changes in heat-agent? | 18:39 |
flwang1 | but which most like a double checker? can i say that? | 18:41 |
zaneb | yes, you still need to propose a fix for that to resolve the bug | 18:42 |
flwang1 | zaneb: no problem | 18:43 |
flwang1 | i will get it done today, many thanks | 18:43 |
zaneb | awesome, cheers | 18:43 |
flwang1 | cheers | 18:43 |
*** armaan has quit IRC | 18:44 | |
*** armaan has joined #heat | 18:45 | |
*** sshnaidm|rover is now known as sshnaidm|off | 18:46 | |
openstackgerrit | Zane Bitter proposed openstack/heat master: Destroy mox forever https://review.openstack.org/560476 | 18:48 |
*** hongbin has quit IRC | 18:48 | |
*** hongbin has joined #heat | 18:48 | |
*** armaan has quit IRC | 18:49 | |
*** zaneb has left #heat | 18:50 | |
*** zaneb has joined #heat | 18:50 | |
*** liyi has joined #heat | 18:53 | |
*** liyi has quit IRC | 18:57 | |
openstackgerrit | Feilong Wang proposed openstack/heat master: Fix multi region issue for software deployment https://review.openstack.org/580470 | 18:59 |
flwang1 | zaneb: ^ | 18:59 |
zaneb | thanks! will take a look | 18:59 |
flwang1 | thank you | 19:00 |
*** cliffparsons has quit IRC | 19:27 | |
*** gkadam has quit IRC | 19:32 | |
zaneb | schaney: looking through the code, your understanding of the docs re PUT/PATCH appears to be correct | 19:36 |
zaneb | so I don't know why existing=True on PUT would not work | 19:37 |
zaneb | (I also don't know why anybody wouldn't just use PATCH for that) | 19:37 |
*** lifeless has joined #heat | 19:39 | |
*** jmlowe has quit IRC | 19:40 | |
*** jmlowe has joined #heat | 20:00 | |
*** flwang1 has quit IRC | 20:03 | |
*** crose has joined #heat | 20:07 | |
*** crose has quit IRC | 20:12 | |
schaney | ok cool, thanks for taking a look zaneb, PATCH will work just fine. | 20:24 |
*** stevebaker has quit IRC | 20:28 | |
*** dmellado has quit IRC | 20:28 | |
*** mchlumsky has quit IRC | 20:36 | |
*** jmlowe has quit IRC | 20:59 | |
*** jmlowe has joined #heat | 21:00 | |
*** flwang1 has joined #heat | 21:06 | |
*** armaan has joined #heat | 21:23 | |
*** stevebaker has joined #heat | 21:52 | |
*** ivve has quit IRC | 21:55 | |
*** rcernin has joined #heat | 22:00 | |
*** nicolasbock has quit IRC | 22:17 | |
*** armaan has quit IRC | 22:36 | |
*** armaan has joined #heat | 22:37 | |
*** edmondsw_ has quit IRC | 22:40 | |
*** edmondsw has joined #heat | 22:41 | |
*** sai- has joined #heat | 22:41 | |
*** armaan has quit IRC | 22:41 | |
*** sai- has left #heat | 22:41 | |
*** sai- has joined #heat | 22:42 | |
*** edmondsw has quit IRC | 22:45 | |
*** sai- has left #heat | 22:48 | |
*** hongbin has quit IRC | 22:50 | |
*** rcernin has quit IRC | 22:58 | |
*** rcernin has joined #heat | 23:01 | |
*** neatherweb has quit IRC | 23:02 | |
*** sai- has joined #heat | 23:41 | |
*** rgold has quit IRC | 23:56 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!