When you are implementing CA Service Catalog, it is critical to capture accurate business requirements and use them to formulate efficient technical architectures. Therefore, you start an implementation with a discovery phase per module to capture the business requirements (functional and non-functional). Next, you study the limitations and capabilities of the client operational infrastructure, and you formulate an efficient implementation architecture. With all the information you have obtained, you can formulate an initial sizing estimate.
Based on the many types of possible variations in usage and architecture, organizations should start with two instances of each component for fail-over and load balancing. Because components can be co-located, start with at least two server class machines in addition to the database server or servers. If computational engines (for example, from billing or CA Process Automation) are heavily used during peak activity times, they may require their own server or servers. Clients' usage-patterns typically change over time as they grow to understand a product's features and use them more heavily. As you discover usage patterns over time, you can dynamically add and distribute components as needed.
|
Copyright © 2013 CA.
All rights reserved.
|
|