Performance Tuning Terminology

Table 2-1 provides definitions of specific terms related to performance and tuning Siebel Business Applications. For definitions of performance and scalability, see "About Performance and Scalability".

For more information about some of these terms and concepts (including concurrent users and think time) in the context of tuning Siebel Application Object Manager components, see "Performance Factors for Siebel Application Object Manager Deployments".

Table 2-1 Performance Tuning Terminology

Term	Definition
Concurrent users	The number of application users actively using and accessing the Siebel application, or a particular element, such as a Siebel Application Object Manager process, at a particular time. It is useful to distinguish between concurrently connected users and concurrently active users: both sets of users consume memory, but only active users consume CPU (processor) resources.
Latency	Delay experienced in network transmissions as network packets traverse the network infrastructure.
Think time	The wait time between user operations. For example, if a user navigates to the Account screen and reviews data for 10 seconds before performing another operation, then the think time in this case is 10 seconds. Average think time is a critical element in performance and scalability tuning, particularly for Siebel Application Object Manager. When think time values are correctly forecasted, then actual load levels will be close to anticipated loads.
Process	An operating system (OS) process. For example, a Siebel Server component such as Siebel Application Object Manager consists of multiple OS processes, referred to as multithreaded processes.
Multithreaded process (or MT server)	A process running on a multithreaded Siebel Server component that supports multiple threads (tasks) per process. Siebel Application Object Manager components run multithreaded processes that support threads.
Task	A concept for Siebel applications of a unit of work that can be done by a Siebel Server component. Siebel tasks are typically implemented as threads.
Thread	An operating system feature for performing a given unit of work. Threads are used to implement tasks for most Siebel Server components. A multithreaded process supports running multiple threads to perform work such as to support user sessions.
Response time	Amount of time the Siebel application takes to respond to a user request, as experienced by the end user. Response time is an aggregate of time incurred by all server processing and transmission latency for an operation. Response time is based on processing related to the request and to processing for other requests that might affect this user request.
Throughput	Typically expressed in transactions per second (TPS), expresses how many operations or transactions can be processed in a set amount of time.