Abstract: According to one general aspect, a method may include monitoring the execution of at least a portion of a software application. The method may also include collecting subroutine call information regarding a plurality of subroutine calls included by the portion of the software application, wherein one or more of the subroutine calls is selected for detailed data recording. The method may further include pruning, as the software application is being executed, a subroutine call tree to include only the subroutine calls selected for detailed data recording and one or more parent subroutine calls of each subroutine calls selected for detailed data recording.

Abstract: According to one general aspect, a method may include monitoring the execution or at least a portion of a software application. The method may also include collecting subroutine call information regarding a plurality of subroutine calls included by the portion of the software application, wherein one or more of the subroutine calls is selected for detailed data recording. The method may further include pruning, as the software application is being executed, a subroutine call tree to include only the subroutine calls selected for detailed data recording and one or more parent subroutine calls of each subroutine calls selected for detailed data recording.

Abstract: Analysis of execution traces to identify execution problems is described. Analysis of the execution trace allows the user to detect application execution patterns. Each pattern represents a sequence of operations performed by the application. Some patterns correspond to normal execution of the application. Some patterns correspond to typical operations such as file opening, file saving, site browsing, mail sending. Diagnostic classes include patterns associated with certain malfunctions. In one embodiment, the system includes a learning mode wherein the system accumulates patterns belonging to different classes and stores them in a pattern database. In one embodiment, the system also includes a recognition mode where the system matches the trace against the pattern database and assigns trace regions to specific classes such as normal, abnormal, classes of specific problems or user activities, etc.

Abstract: A tracing system that provides automated tuning of execution tracing by adjusting the collection of trace data is described. In one embodiment, the user sets an initial tracing profile for a tracing program. In addition, the user sets an upper limit for the tracing performance penalty. The auto-tuning system monitors the performance penalty induced by tracing and, when the performance impact is excessive, removes trace points that are causing the most impact on performance. Auto tuning is especially useful for performing software recording in mission-critical and/or time-critical applications, such as servers, real-time applications, etc. The system typically adjusts relatively quickly such that most users do not feel the influence of the tracer.

Abstract: A software system is disclosed which facilitates the process of tracing the execution paths of a program, called the client. The tracing is performed without requiring modifications to the executable or source code files of the client. Trace data collected during the tracing operation is collected according to instructions in a trace options file. At run time, the tracing library attaches to the memory image of the client. The tracing library is configured to monitor execution of the client and to collect trace data, based on selections in the trace options file. Conditional tracing, through the use of triggers and actions taken in response to the triggers, allows the developer to control the tracing operation. The triggers can be conditional triggers in which the corresponding action is taken only if a conditional expression is satisfied. The system can trace multiple threads and multiple processes. The tracing system provides a remote mode and an online mode.