HOT is a new template format meant to replace the CloudFormation-compatible format (CFN) as the native format supported by the Orchestration module over time. This guide is targeted towards template authors and explains how to write HOT templates based on examples. A detailed specification of HOT can be found at Heat Orchestration Template (HOT) specification.
This section gives an introduction on how to write HOT templates, starting from very basic steps and then going into more and more detail by means of examples.
The most basic template you can think of contains only a single resource definition using only predefined properties. For example, the template below could be used to deploy a single compute instance:
heat_template_version: 2015-04-30
description: Simple template to deploy a single compute instance
resources:
my_instance:
type: OS::Nova::Server
properties:
key_name: my_key
image: ubuntu-trusty-x86_64
flavor: m1.small
Each HOT template must include the heat_template_version key with the HOT version value, for example, 2013-05-23. A list of HOT template versions can be found at Heat Template Version file
The description key is optional, however it is good practice to include some useful text that describes what users can do with the template. In case you want to provide a longer description that does not fit on a single line, you can provide multi-line text in YAML, for example:
description: >
This is how you can provide a longer description
of your template that goes over several lines.
The resources section is required and must contain at least one resource definition. In the above example, a compute instance is defined with fixed values for the key_name, image and flavor properties.
Note
All the defined elements (key pair, image, flavor) have to exist in the OpenStack environment where the template is used.
Input parameters defined in the parameters section of a template allow users to customize a template during deployment. For example, this allows for providing custom key pair names or image IDs to be used for a deployment. From a template author’s perspective, this helps to make a template more easily reusable by avoiding hardcoded assumptions.
The following example extends the previous template to provide parameters for the key pair, image and flavor properties of the resource:
heat_template_version: 2015-04-30
description: Simple template to deploy a single compute instance
parameters:
key_name:
type: string
label: Key Name
description: Name of key-pair to be used for compute instance
image_id:
type: string
label: Image ID
description: Image to be used for compute instance
flavor:
type: string
label: Instance Type
description: Type of instance (flavor) to be used
resources:
my_instance:
type: OS::Nova::Server
properties:
key_name: { get_param: key_name }
image: { get_param: image_id }
flavor: { get_param: flavor }
Values for the three parameters must be defined by the template user during the deployment of a stack. The get_param intrinsic function retrieves a user-specified value for a given parameter and uses this value for the associated resource property.
For more information about intrinsic functions, see Intrinsic functions.
You can provide default values for parameters. If a user doesn’t define a value for a parameter, the default value is used during the stack deployment. The following example defines a default value m1.small for the flavor property:
parameters:
flavor:
type: string
label: Instance Type
description: Flavor to be used
default: m1.small
Note
If a template doesn’t define a default value for a parameter, then the user must define the value, otherwise the stack creation will fail.
The values that a user provides when deploying a stack are available in the stack details and can be accessed by any user in the same tenant. To hide the value of a parameter, use the hidden boolean attribute of the parameter:
parameters:
database_password:
type: string
label: Database Password
description: Password to be used for database
hidden: true
You can restrict the values of an input parameter to make sure that the user defines valid data for this parameter. The constraints property of an input parameter defines a list of constraints to apply for the parameter. The following example restricts the flavor parameter to a list of three possible values:
parameters:
flavor:
type: string
label: Instance Type
description: Type of instance (flavor) to be used
constraints:
- allowed_values: [ m1.medium, m1.large, m1.xlarge ]
description: Value must be one of m1.medium, m1.large or m1.xlarge.
The following example defines multiple constraints for a password definition:
parameters:
database_password:
type: string
label: Database Password
description: Password to be used for database
hidden: true
constraints:
- length: { min: 6, max: 8 }
description: Password length must be between 6 and 8 characters.
- allowed_pattern: "[a-zA-Z0-9]+"
description: Password must consist of characters and numbers only.
- allowed_pattern: "[A-Z]+[a-zA-Z0-9]*"
description: Password must start with an uppercase character.
The list of supported constraints is available in the Parameter Constraints section.
Note
You can define multiple constraints of the same type. Especially in the case of allowed patterns this not only allows for keeping regular expressions simple and maintainable, but also for keeping error messages to be presented to users precise.
In addition to template customization through input parameters, you can provide information about the resources created during the stack deployment to the users in the outputs section of a template. In the following example the output section provides the IP address of the my_instance resource:
outputs:
instance_ip:
description: The IP address of the deployed instance
value: { get_attr: [my_instance, first_address] }
Note
Output values are typically resolved using intrinsic function such as the get_attr. See Intrinsic functions for more information about intrinsic functions..
See Outputs section for more information about the outputs section.