Mirth Newsletter

Deployment Scenarios

Deployment Scenarios
Mirth Appliances can be deployed in a number of ways, depending on the needs of your organization. They can be used for public facing, private, or hybrid messaging systems. The flexibility of the appliances allows them to be used as message gateways, data collection hubs, application integration systems, etc. They can also be deployed individually, as clustered pairs, or horizontally scaled behind a load balancer. The combinations are limitless.

The following examples present just a few possibilities for using Mirth Appliances within your organization.
Hub and Spoke Messaging System
The Mirth E1000 Enterprise Appliance is well suited to anchor the center of a hub and spoke messaging network. Its rack-mount form factor and high-availability features are designed for data center applications. The performance and processing capacity of the E1000 allows it to handle a large message volume from diverse spoke locations.

The example diagram shows an E1000 acting as a message hub for a variety of internal and external connections.
Scaling for High Message Volume
A single Mirth E1000 Enterprise Appliance can handle daily volumes well in excess of one million messages. When your performance requirements exceed the capabilities of a lone E1000, it is possible to scale horizontally using a network load balancer. With the exception of VPN traffic, all client-facing communication requires only one connection per session. This message traffic can successfully be routed through a network load balancer and directed to any number of appliances.

Administration of the appliances should be performed individually and not through the load balancer. Channel configurations should be kept the same across all appliances using the channel import and export features of the Mirth Administrator client.
High Availability Message Gateway
When high availability is required, any two identical Mirth Appliances can be configured as a fail-over cluster. For high volume applications, or to maximize availability of individual nodes, the E1000 model is recommended. For low volume solutions, even the entry level P1000 model supports clustering.

When a cluster is configured, channels are automatically synchronized from the primary node to the standby node. In the event of hardware or software failure, or if there is a loss of connectivity, the assigned cluster IP address will automatically direct traffic away from the failed primary node and deliver it to the backup device.
Multiple Environments Supporting Production
Best practices for deploying a production messaging system call for a strict change control process. By creating multiple environments, an organization can maintain the integrity of the production system while changes are being developed and tested. A recommended flow would include the following environments:

  • Development—environment for creating new channels and configurations.
  • QA—a separate quality assurance environment to test the changes as they would be assembled for production.
  • Staging—a functionally equivalent environment to test the deployment process and give outside users access to the pre-production channels.
  • Production—the environment handling live message traffic.

Since the entry level P1000 appliances support all of the same features as the larger E1000, it is reasonable to use P1000s for all pre-production environments. The only exception to this would be when load testing or performance benchmarks are part of the QA cycle. In these cases it will be necessary to use the E1000 to ensure accurate results.