| stevebaker | flwang: are you saying this didn't work? $STACK_ID: {get_param "OS::stack_id"} | 00:13 |
|---|---|---|
| stevebaker | flwang: that sounds like a heat bug | 00:13 |
| *** radeks has quit IRC | 01:15 | |
| *** hongbin has joined #heat | 01:44 | |
| flwang | stevebaker: it doesn't work for me | 01:56 |
| flwang | i got an error saying there is circular reference, i will test it again with a simpler HOT and let you know | 01:56 |
| flwang | stevebaker: many thanks for your help | 01:56 |
| *** ramishra has joined #heat | 02:27 | |
| flwang | stevebaker: https://github.com/coreos/ignition/issues/849 | 02:53 |
| *** maddtux has joined #heat | 03:03 | |
| *** skramaja has joined #heat | 03:18 | |
| *** hongbin has quit IRC | 03:26 | |
| *** hongbin has joined #heat | 03:28 | |
| *** e0ne has joined #heat | 03:47 | |
| *** gkadam has joined #heat | 03:50 | |
| *** hongbin has quit IRC | 04:29 | |
| *** e0ne has quit IRC | 04:49 | |
| *** hongbin has joined #heat | 04:49 | |
| *** hongbin has quit IRC | 04:49 | |
| *** e0ne has joined #heat | 04:50 | |
| *** e0ne has quit IRC | 04:51 | |
| *** e0ne has joined #heat | 05:08 | |
| openstackgerrit | gugug proposed openstack/python-heatclient master: Sync Sphinx requirement https://review.opendev.org/666061 | 05:12 |
| *** e0ne has quit IRC | 05:16 | |
| *** ivve has joined #heat | 05:35 | |
| *** e0ne has joined #heat | 06:00 | |
| *** e0ne has quit IRC | 06:03 | |
| ricolin | gregwork, Hi just saw your messages, get_file with URL only supported with orchestration client right now (will check with others on why we didn't add that for API on first place), If you're using commends like openstack stack create/heat stack-create, you should get warning on didn't fetch url content and forbid you to run the create process | 06:12 |
| *** e0ne has joined #heat | 06:12 | |
| gregwork | ricolin: alright .. i did eventually figure it out :) | 06:13 |
| gregwork | side question .. is there any way to refernce a key name in a map for the purposes of a comparison | 06:13 |
| ricolin | what map? | 06:15 |
| ricolin | zaneb, I'm wondering if we should directly support get_file with URL for API call. Know any history about that?:) | 06:16 |
| *** e0ne has quit IRC | 06:20 | |
| gregwork | im doing something like this: https://pastebin.com/p4GBnDTH | 06:20 |
| gregwork | so that all works but i had to change my ipa_enrollment environment from a key: password map | 06:21 |
| gregwork | to nameofhostkey: { name: nameofhost, otp: password } | 06:21 |
| gregwork | because i couldnt reference the keyname | 06:21 |
| gregwork | there doesnt appear to be a way to do that, only get values | 06:21 |
| gregwork | btw conditions and if .. so good :) | 06:23 |
| gregwork | also OS::Heat::TestResource | 06:24 |
| gregwork | very useful | 06:24 |
| gregwork | cant imagine doing too much self education without that kind of stuff | 06:25 |
| gregwork | ricolin: i found zaneb's about that during my googling .. sad it got shutdown due to "lack of customer interest" | 06:27 |
| gregwork | *zaneb's bz | 06:27 |
| gregwork | heat templates need to live in scm, you need maximum flex to leverage things like git | 06:28 |
| * gregwork sighs .. will need to open some tickets on why heat falls down so hard when dealing with octavia stack deletes | 06:30 | |
| gregwork | 15-20% of the time i stack delete the lb fails to get cleaned up by heat :/ | 06:31 |
| *** radeks has joined #heat | 06:31 | |
| ricolin | getting key names can be an interesting cases, we usually assume user already knows about all keys like `get_param: [ param_map_1, key_1 ] ` | 06:40 |
| ricolin | maybe you can use yaql in template for that functionality https://docs.openstack.org/heat/latest/template_guide/hot_spec.html#yaql | 06:45 |
| ricolin | and for octavia stack delete issue, are you plan to create a story in heat for that? | 06:47 |
| ricolin | gregwork, ^^^ | 06:47 |
| gregwork | ricolin: you see what im trying to do there tho right .. if the guest_name matches the key of the same name in the ipa_enrollment dictionary | 06:49 |
| gregwork | but that cant be done without maybe yaql magic so i had to adjust my dictionary and add placeholders | 06:49 |
| gregwork | im almost done my test case which ill post on my github, in context this might make a bit more sense. the goal is to accomplish the following mostly in heat: 1) stand up the cloudforms project with quotas, 2) deploy a 8 node cluster ( bastion, 2db, 2web, 3 worker ) with a load balancer, proper network isolation by cluster role, security groups, neutron rbac for the workers to talk to the undercloud so they can enroll | 06:52 |
| gregwork | the | 06:52 |
| gregwork | director | 06:52 |
| gregwork | there just nothing out there ive seen that does that, and i think there should be something that does a proper prod like deployment of cloudforms in openstack | 06:53 |
| gregwork | can heat work ansible? i saw something about software configuration via puppet .. not sure if there is ansible capability | 06:54 |
| gregwork | i wrote roles in ansible to configure the cfme appliances in their respective tasks .. ive been thinking about how to do that in the absence of ansible with softwareconfig | 06:55 |
| gregwork | i was thinking about maybe using a wait condition on the bastion and once the cluster nodes are all up perhaps having the bastion do the cfme appliance configuration | 06:56 |
| *** gfidente has joined #heat | 07:10 | |
| *** rcernin has quit IRC | 07:23 | |
| ricolin | gregwork, you can use ansible with software config https://github.com/openstack/heat-templates/blob/master/hot/software-config/example-templates/example-ansible-template.yaml | 07:30 |
| gregwork | nice | 07:32 |
| gregwork | is there a heat method for generic http ? | 07:35 |
| gregwork | that will need context | 07:35 |
| gregwork | so we also have awx/tower in our environment | 07:35 |
| gregwork | i can expose a job template in tower and if you have creds to perform a get/post you can have the ansible workflow kick off | 07:35 |
| gregwork | i was thinking for a simple get i could use OS::Heat::TestResource and get_file | 07:36 |
| gregwork | but that wouldnt work for anything real, espeically if i needed to pass login / post | 07:37 |
| *** zhurong has quit IRC | 07:56 | |
| *** mlycka has joined #heat | 07:57 | |
| *** tkajinam has quit IRC | 08:02 | |
| *** rubasov has quit IRC | 08:04 | |
| *** rubasov has joined #heat | 08:05 | |
| *** rubasov has quit IRC | 08:08 | |
| *** rubasov has joined #heat | 08:08 | |
| *** e0ne has joined #heat | 08:09 | |
| *** rubasov has quit IRC | 08:14 | |
| *** rubasov has joined #heat | 08:14 | |
| *** k_mouza has joined #heat | 08:18 | |
| *** mikecmpbll has joined #heat | 08:32 | |
| *** rubasov has quit IRC | 08:33 | |
| openstackgerrit | Thomas Bechtold proposed openstack/heat-tempest-plugin master: Drop "docs" tox target and the docs integration https://review.opendev.org/679724 | 08:49 |
| *** rcernin has joined #heat | 09:00 | |
| ricolin | I think we better handing http fetch in a place that can deal with more sophisticate logic like use heat-agents with script or ansible | 09:08 |
| ricolin | but totally fine if any new function added in stack template:) | 09:09 |
| *** SmearedBeard has joined #heat | 09:31 | |
| *** tkajinam has joined #heat | 09:48 | |
| *** k_mouza has quit IRC | 09:50 | |
| *** k_mouza has joined #heat | 10:03 | |
| *** k_mouza has quit IRC | 10:14 | |
| *** k_mouza has joined #heat | 10:15 | |
| *** tkajinam has quit IRC | 10:18 | |
| *** SmearedBeard has quit IRC | 10:31 | |
| *** SmearedBeard has joined #heat | 10:33 | |
| *** hoonetorg has quit IRC | 10:44 | |
| *** hoonetorg has joined #heat | 10:57 | |
| *** SmearedBeard has quit IRC | 11:02 | |
| *** SmearedBeard has joined #heat | 11:03 | |
| *** rubasov has joined #heat | 11:04 | |
| *** SmearedBeard has quit IRC | 11:05 | |
| *** ricolin_ has joined #heat | 11:51 | |
| *** ricolin has quit IRC | 11:54 | |
| *** k_mouza_ has joined #heat | 12:19 | |
| *** k_mouza has quit IRC | 12:22 | |
| *** jcoufal has joined #heat | 12:29 | |
| *** maddtux has quit IRC | 12:36 | |
| *** gkadam has quit IRC | 12:37 | |
| *** jcoufal has quit IRC | 12:39 | |
| *** jcoufal has joined #heat | 13:06 | |
| *** k_mouza has joined #heat | 13:07 | |
| *** k_mouza_ has quit IRC | 13:11 | |
| *** k_mouza_ has joined #heat | 13:35 | |
| *** k_mouza has quit IRC | 13:38 | |
| *** k_mouza has joined #heat | 14:02 | |
| *** k_mouza_ has quit IRC | 14:05 | |
| *** rcernin has quit IRC | 14:32 | |
| *** ramishra_ has joined #heat | 14:36 | |
| *** ramishra has quit IRC | 14:36 | |
| *** mlycka has quit IRC | 14:50 | |
| *** skramaja has quit IRC | 15:10 | |
| *** beekneemech is now known as bnemec | 15:51 | |
| *** jmlowe has quit IRC | 15:59 | |
| *** e0ne has quit IRC | 16:00 | |
| *** jcoufal has quit IRC | 16:15 | |
| *** jcoufal has joined #heat | 16:16 | |
| *** ricolin_ has quit IRC | 16:25 | |
| *** k_mouza_ has joined #heat | 16:49 | |
| *** k_mouza_ has quit IRC | 16:51 | |
| *** k_mouza has quit IRC | 16:52 | |
| *** jmlowe has joined #heat | 17:00 | |
| *** jtomasek has quit IRC | 17:01 | |
| *** mikecmpbll has quit IRC | 17:03 | |
| *** e0ne has joined #heat | 17:12 | |
| *** e0ne has quit IRC | 17:23 | |
| *** e0ne has joined #heat | 17:23 | |
| *** ramishra_ has quit IRC | 17:30 | |
| *** e0ne has quit IRC | 17:30 | |
| *** jtomasek has joined #heat | 17:46 | |
| *** jtomasek has quit IRC | 17:50 | |
| *** jtomasek has joined #heat | 17:53 | |
| *** e0ne has joined #heat | 18:15 | |
| *** gfidente is now known as gfidente|afk | 18:43 | |
| *** e0ne_ has joined #heat | 19:24 | |
| *** e0ne_ has quit IRC | 19:26 | |
| *** e0ne has quit IRC | 19:27 | |
| *** jmlowe has quit IRC | 19:29 | |
| *** jmlowe has joined #heat | 19:47 | |
| *** radeks has quit IRC | 20:37 | |
| *** jtomasek has quit IRC | 20:42 | |
| *** jcoufal has quit IRC | 20:42 | |
| *** iokiwi has joined #heat | 22:47 | |
| *** tkajinam has joined #heat | 22:55 | |
| *** rcernin has joined #heat | 23:06 | |
| flwang | stevebaker: around? | 23:48 |
| stevebaker | flwang: hey | 23:48 |
| flwang | stevebaker: i tried vnd.coreos.ignition+json and hacked heat to support that, but still no luck | 23:49 |
| stevebaker | flwang: did the multipart resource actually get created this time though? | 23:49 |
| flwang | stevebaker: i added a 'or' condition at https://github.com/openstack/heat/blob/master/heat/engine/resources/openstack/heat/multi_part.py#L120 to support the coreos mime | 23:50 |
| flwang | i can't ssh into the node, could i check that from heat? | 23:50 |
| stevebaker | flwang: oh yeah, you'll need to add 'application' type support https://github.com/openstack/heat/blob/master/heat/engine/resources/openstack/heat/multi_part.py#L56 | 23:53 |
| flwang | is 'application' the accurate value? | 23:54 |
| flwang | or i should use 'vnd.coreos.ignition+json'? | 23:54 |
| flwang | http://paste.openstack.org/show/770468/ | 23:55 |
| flwang | this is my code | 23:55 |
| stevebaker | the type is application, the subtype is vnd.coreos.ignition+json. I think the resource will need a python change to support 'application' as a type, to be treated exactly like 'text' | 23:56 |
| flwang | ok | 23:57 |
| flwang | let me try now | 23:57 |
| openstackgerrit | Feilong Wang proposed openstack/heat master: [WIP] Support CoreOS ignition https://review.opendev.org/679882 | 23:58 |
| flwang | stevebaker: is this https://review.opendev.org/#/c/679882/1/heat/engine/resources/openstack/heat/multi_part.py enough? | 23:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!