ICN and VirtualViewer® HTML5 Deployment models

When deploying the VirtualViewer®, either a single server or a distributed server deployment model may be used. As a plugin to IBM Content Navigator, it is recommended that the VirtualViewer be deployed in a manner that will best support the deployment model used for ICN.

Single server deployment model

The single server deployment model utilizes the same application server that is running IBM Content Navigator for deployment and configuration of the VirtualViewer® server. The biggest benefit of this model is that it aids in a swift setup and configuration of the Virtual Viewer® server. The scalability however is limited to the hardware configuration expandability.

This is the simplest form of deployment, where the VirtualViewer is deployed on the same application server as IBM Content Navigator. The ease of deployment allows for a quick setup of Development and/or POC environments. When using this deployment model, it is critical that the dedicated server has adequate resources allocated to allow for running multiple JVM’s.

This architecture permits limited scalability of the application environment, by scaling up the dedicated server. The diagram below depicts what a scale up of a server looks like:

Distributed server deployment model

Deploying the VirtualViewer® in this model conforms to IBM’s clustering architecture and allows the application to scale for future growth. The Distributed Server Deployment Model, consequently, is the preferred deployment model for Production environments.

This deployment model is implemented by deploying the VirtualViewer across multiple servers that each have ICN running. It is expected, as with any clustered environment, that the resource utilization for the applications running behind load balancers are optimally distributed across the infrastructure.

Conforming to IBM Content Navigator’s design practices, the VirtualViewer scales up alongside ICN.