Heat
Heat is a template driven service that automates the management of the entire lifecycle of your application on the NeCTAR cloud.
A 'template driven service' simply means that you define your application's requirements in a human readable text file - the template. In this file you to describe both the infrastructure and its relationships that your application will need to run on the NeCTAR cloud.
Heat then uses this template to provision the required infrastructure and manage the lifecycle of your application from start to finish. This template, and the infrastructure that it has created, is termed a 'stack'.
As part of the life cycle management, the Heat service supports both scaling on demand and the freeing up of infrastructure once the application is finished.
Heat integrates well with configuration management tools, such as Chef and Puppet. Thus the Heat service offers executable documentation of your application's deployment and lifecycle, making your deployments repeatable and reliable. The net effect is to limit human error and to save you time. Thus saving you money.
The stack template format(s)
Heat is modelled after Amazon's CloudFormation service, and endeavours to maintain some degree of compatibility with this service. Hence Heat supports two different template formats.
- The first is a JSON based implementation that mimics the Amazon specification.
- The second is a YAML based native OpenStack implementation termed 'HOT'.
We strongly recommend that you use the HOT format.
The stack lifecycle
A template is created, using a standard text editor (such as Brackets). It is then uploaded into the OpenStack Heat service, either by means of the Heat Command line client, or the Nectar dashboard.
If uploaded via the command line client, the engine expects any mandatory parameters to be provided as arguments added at the time the template was uploaded.
If, however, uploaded via the dashboard, then the dashboard will create an input wizard that will step the person who uploaded the template through the process of entering the required parameter values.
Once all the required data has been gathered the stack is then provisioned and launched.
The template and its associated parameters will remain in the Heat database until such time as the engine is instructed to destroy the stack.
At that point all the provisioned infrastructure will be destroyed, its resources released, and then the template and its parameters will be removed from the Heat database
For more on Heat, read the other articles here.
