Sun Java Communications Suite 5 Deployment Planning Guide

Planning for Growth

In addition to determining what capacity you need today, assess what capacity you need in the future, within a time frame that you can plan for. Typically, a growth time line is in the range of 12 to 18 months. Growth expectations and changes in usage characteristics are factors that you need to take into account to accommodate growth.

As the number of users and messages increase, you should outline successful guidelines for capacity planning. You need to plan for increases in message traffic for the various servers, a larger volume of users, larger mailbox sizes, more calendar appointments, and so forth. As growth occurs in the user population, usage characteristics change over time. Your deployment goals (and therefore deployment design) must respond accordingly to be viable into the future.

Ideally, you should design your architecture to easily accommodate future growth. For example, use logical names for the Communications Suite services themselves. See Using Logical Service Names for more information. Monitoring the deployment, once it enters its production phase, is also crucial to being able to understand when and by how much a deployment needs to grow.

Understanding Total Cost of Ownership

Total Cost of Ownership (TCO) is another factor that affects capacity planning. This includes choosing the hardware upon which to deploy your Communications Suite. The following table presents some factors to consider as to whether to deploy more smaller hardware systems or fewer larger hardware systems.

Table 2–1 Considerations for Total Cost of Ownership

Hardware Choice  



More, smaller hardware systems 

  • Smaller hardware systems generally cost less.

  • More, smaller hardware systems can be deployed across many locations to support a distributed business environment.

  • More, smaller hardware systems can mean less down time for system maintenance, upgrade, and migration because traffic can be routed to other servers that are still online while others are being maintained.

  • Smaller hardware systems have a more limited capacity, so more of them are needed. Management, administration, and maintenance costs go up as the number of hardware systems goes up.

  • More, smaller hardware systems require more system maintenance because there are more of them to maintain.

Fewer, larger hardware systems 

  • Fewer hardware systems means fewer fixed management costs per server. If your management costs are a recurring monthly bill, whether internal or from an ISP, costs will be lower, because you have fewer hardware systems to manage.

  • Fewer hardware systems can also mean easier system maintenance, upgrade, and migration because there are fewer systems to maintain.

  • Larger hardware systems generally cost more initially.

  • Fewer hardware systems can mean a greater system down-time for maintenance, upgrade and migration.