Planning › System Architecture › Key Terms and Concepts

Key Terms and Concepts

Availability refers to the ability of an organization to deliver consistent, predictable access to applications and data.

The number of catalog requests per minute is the key criterion for determining the size of the architecture as small, medium, or large. This criterion is much more important than the number of users in the system or other criteria. The higher the number of catalog requests per minute, the larger the recommended architecture.

A cluster is two or more interconnected computers that create a solution to provide higher availability, higher scalability, or both.

Clustering for CA Service Catalog and CA Service Accounting is accomplished through Service View clustering, because all user communications with catalog and accounting functions are accomplished through Service View. Thus, the term Service View clustering includes both CA Service Catalog clustering and CA Service Accounting clustering.

Load balancing means dividing the amount of work between two or more computers, instead of a single computer. As a result, the system accomplishes more work in the same amount of time and typically serves all users faster.

In horizontal clustering, different physical computers comprise a cluster. As a result, the load balancer forwards requests to different computers having different IP addresses. All Service View installations on all computers must use the same cluster of the MDB. To implement clustering, you make several related configuration changes manually. Horizontal clustering enables a single application to span several computers while presenting a single system image. Horizontal clustering can also provide increased throughput and high availability. Horizontal clustering applies to both Service View and CA Workflow.

In vertical clustering, all servers are on the same computers and thus use different ports. Avoid port conflicts. To implement clustering, you run an ant script that makes several related configuration changes automatically.

A single application server implemented by a single Java virtual machine (JVM) process cannot always maximize the power and RAM of a large multiprocessor computer. Vertical clustering provides a straightforward mechanism to create multiple JVM processes that together can maximize all the processing power and memory available. Vertical clustering applies to Service View only.

Note: If you are using CA Process Automation as your process automation tool, see your CA Process Automation documentation for information about using clustering with CA Process Automation.