Table of Contents |
---|
Implementation Overview
WSO2 Application Server (AS) 5.2.1 is 1 is provided as a service on WSO2 Private PaaS 4.1.0. This is implemented by creating a template module and a cartridge agent plugin to automate the configuration in run-time.
Template Module
The WSO2 Application Server (AS) 5.2.1 template module provides a set of templates that is used to configure the Application Server with the Configurator.
Info |
---|
You can find the WSO2 AS 5.2.1 template module here. |
...
The Server configurations are automated through the cartridge agent plugin, which was implemented for WSO2 AS 5.2.1.
The plugin configures the following and many more:
- server hostname
- cluster discovery parameters
- proxy ports
Info |
---|
You can find the cartridge agent plugin here. |
the following:
Template module
Panel borderColor #71CFF5 borderWidth 2 The template module configures the product in runtime, with a set of key value pairs known as the configuration parameters.
Cartridge agent plugin
Panel borderColor #71CFF5 borderWidth 2 The cartridge agent plugin, sets the runtime configuration parameters such as hostnames, IP addresses of dependent clusters, dynamic proxy ports and many more.
Docker image
Panel borderColor #71CFF5 borderWidth 2 The Docker image is built by packaging the following components:
Cartridge agent
Product distribution
Template module
Configurator
Java runtime
Puppet module
Panel borderColor #71CFF5 borderWidth 2 The puppet module includes the same set of component as in the Docker image. In the puppet image the template module is used to configure automation.
Deploying WSO2 AS as a Service
Deploy WSO2 AS as a service using one of the following methods.
- Deploy the application on Private PaaS when running on a Virtual Machine.
- Deploy the application on Private PaaS when running on Kubernetes.
Deployment Architecture
The diagram illustrates the deployment architecture of WSO2 AS on Private PaaS. In the deployment, two different Private PaaS clusters will be created for the AS manager and worker. These clusters can be scaled independently.