Non-invasive time-based profiling tool

• US 10,127,129 B2
• Filed: 02/27/2007
• Issued: 11/13/2018
• Est. Priority Date: 02/27/2007
• Status: Active Grant

• Current Assignee: Red Hat, Inc. (International Business Machines Corporation)
• Original Assignee: Red Hat, Inc. (International Business Machines Corporation)
• Inventors: Parkinson, Steven W.
• Primary Examiner(s): Puente, Emerson
• Assistant Examiner(s): KRETZMER, ERIKA A
• Application Number:
  US11/712,338
• Publication Number:
  US 20080209402A1
• Time in Patent Office: 4,277 Days
• Field of Search: 714 45, 714 46, 714 51, 717125, 717127-133, 715700-867, 715961, 715963, 715965-970
• CPC Class Codes: G06F 11/323 Visualisation of programs or trace data
• CPC Class Codes: G06F 11/3495 for systems
• CPC Class Codes: G06F 11/3664 Environments for testing or debugging software
• CPC Class Codes: G06F 2201/835 Timestamp
• CPC Class Codes: G06F 2201/86 Event-based monitoring
• CPC Class Codes: H04L 41/064 involving time analysis

Abstract

A software execution profiling method comprising: collecting, from a plurality of subsystems of a client machine, information including an event report corresponding to each of a plurality of events occurring in the plurality of subsystems, each event report specifying an event identifier and a cause identifier; responsive to determining that the cause identifier includes an empty field, determining time values in the information associated with the first event and a second which caused the first event; determining to indicate a causation relationship between the first event and the second event in view of the time value of the second event being within the defined period of the time value of the first event; and presenting via a graphical user interface (GUI) the events and an indicator of the causation relationship.

Claims

1. A method comprising:

collecting information from a plurality of subsystems of a client machine in a system, the information comprising an event report corresponding to each of a plurality of events occurring in the plurality of subsystems, each event report indicating one of the plurality of subsystems on which a corresponding event occurred, a first identifier of the event and a cause identifier, wherein each event report is generated by a reporting client in each respective subsystem, and wherein the cause identifier associated with a first event comprises at least one of a unique identifier of a second event which caused the first event or an empty field;

responsive to determining that the cause identifier comprises an empty field, profiling, by a processing device, the system using the information collected from the plurality of subsystems to determine time values in the information associated with the first event and the second event of the plurality of events;

determining that a time value of the second event is within a defined period of time of a time value of the first event;

determining to indicate a causation relationship between the first event and the second event in view of the time value of the second event being within the defined period of the time value of the first event;

generating a timeline to display the plurality of events along with measurement data and a source subsystem of the plurality of subsystems corresponding to each of the plurality of events, wherein the plurality of events are plotted at positions along a first timeline axis relative to the determined time values corresponding to each event and at positions along a second timeline axis relative to the source subsystem of each event, the timeline comprising a visual indication of the causation relationship between the first event and the second event in view of the determining to indicate the causation relationship;

presenting the plurality of events on the timeline, wherein the presenting is performed in a graphical user interface (GUI); and

in response to a user request made via the GUI, presenting additional details of a portion of the timeline comprising the visual indication of the causation relationship and corresponding to a period of time specified in the user request to initiate debugging of at least one of the plurality of subsystems.

2. The method of claim 1, wherein profiling the system using the information comprises:

tracing the plurality of events using the information; and

analyzing the plurality of events on a time basis.

3. The method of claim 1, wherein profiling the system using the information comprises:

correlating the plurality of events using the information to identify causation relationships between the plurality of events.

4. The method of claim 1, further comprising:

storing the time values in a repository; and

allowing users to access the repository to retrieve the time values.

5. The method of claim 1, further comprising:

instrumenting the plurality of subsystems to collect the information during execution of an application program without modifying source code of the application program.

6. The method of claim 1, wherein the event report further comprises a second identifier for a corresponding client machine, a third identifier for a corresponding subsystem in the corresponding client machine, a fourth identifier for an event type of the corresponding event, and a timestamp for the corresponding event.

7. The method of claim 1, wherein the GUI comprises a directed acyclic graph showing the plurality of events along the timeline and a set of zero or more causation relationships between the plurality of events, wherein the measurement data comprises at least one of memory usage or processing device utilization data.

8. An apparatus comprising:

a memory; and

a processing device, operatively coupled to the memory, the processing device to;

collect information from a plurality of subsystems of a client machine in a system, the information comprising an event report corresponding to each of a plurality of events occurring in the plurality of subsystems, each event report indicating one of the plurality of subsystems on which a corresponding event occurred, a first identifier of the event and a cause identifier, wherein each event report is generated by a reporting client in each respective subsystem, and wherein the cause identifier associated with a first event comprises at least one of a unique identifier of a second event which caused the first event or an empty field;

responsive to determining that the cause identifier comprises an empty field, profile the system using the information collected from the plurality of subsystems, the information being collected during execution of an application program in the system, to determine time values in the information associated with the first event and the second event of the plurality of events; and

determine that a time value of the second event is within a defined period of time of a time value of the first event;

determine to indicate a causation relationship between the first event and the second event in view of the time value of the second event being within the defined period of the time value of the first event;

generate a timeline to display the plurality of events along with measurement data and a source subsystem of the plurality of subsystems corresponding to each of the plurality of events, wherein the plurality of events are plotted at positions along a first timeline axis relative to the determined time values corresponding to each event and at positions along a second timeline axis relative to the source subsystem of each event, the timeline comprising a visual indication of the causation relationship between the first event and the second event in view of the determining to indicate the causation relationship;

present the plurality of events on the timeline, wherein the plurality of events are presented in a GUI; and

in response to a user request made via the GUI, present additional details of a portion of the timeline comprising the visual indication of the causation relationship and corresponding to a period of time specified in the user request to initiate debugging of at least one of the plurality of subsystems; and

a repository coupled to the processing device to store the time values.

9. The apparatus of claim 8, further comprising:

a network adapter communicatively coupled to the at least one client machine to receive the information.

10. The apparatus of claim 8, further comprising:

the processing device operatively coupled to the repository to retrieve the time values from the repository and to create a single GUI to present the time values.

11. The apparatus of claim 8, further comprising:

the processing device operatively coupled to the repository to generate a directed acyclic graph to present the plurality of events along the timeline and a set of zero or more causation relationships between the plurality of events, wherein the measurement data comprises at least one of memory usage or processing device utilization data.

12. The apparatus of claim 8, wherein the processing device further to trace the plurality of events and to analyze the plurality of events on a time basis.

13. The apparatus of claim 8, wherein the processing device to correlate the plurality of events to identify causation relationships between the plurality of events.

14. The apparatus of claim 8, wherein the event report further comprises a second identifier for a corresponding client machine, a third identifier for a corresponding subsystem in the corresponding client machine, a fourth identifier for an event type of the corresponding event, and a timestamp for the corresponding event.

15. The apparatus of claim 8, wherein the processing device to instrument the plurality of subsystems to collect the information without modifying source code of the application program.

16. The apparatus of claim 8, further comprising the at least one client machine.

17. A non-transitory machine-readable medium that provides instructions that, when executed by a processing device, will cause the processing device to:

collect information from a plurality of subsystems of a client machine in a system, the information comprising an event report corresponding to each of a plurality of events occurring in the plurality of subsystems, each event report indicating one of the plurality of subsystems on which a corresponding event occurred, a first identifier of the event and a cause identifier, wherein each event report is generated by a reporting client in each respective subsystem, and wherein the cause identifier associated with a first event comprises at least one of a unique identifier of a second event which caused the first event or an empty field;

responsive to determining that the cause identifier comprises an empty field, profile, by the processing device, the system using the information collected from the plurality of subsystems to determine time values in the information associated with the first event and the second event of the plurality of events to determine whether to indicate a causation relationship between the first event and the second event;

generate a timeline to display the plurality of events along with measurement data and a source subsystem of the plurality of subsystems corresponding to each of the plurality of events, wherein the plurality of events are plotted at positions along a first timeline axis relative to the determined time values corresponding to each event and at positions along a second timeline axis relative to the source subsystem of each event, the timeline comprising a visual indication of the causation relationship between the first event and the second event in view of the determining whether to indicate the causation relationship;

present the plurality of events on the timeline, wherein the presenting is performed in a graphical user interface (GUI); and

in response to a user request made via the GUI, present additional details of a portion of the timeline comprising the visual indication of the causation relationship and corresponding to a period of time specified in the user request to initiate debugging of at least one of the plurality of subsystems.

18. The non-transitory machine-readable medium of claim 17, wherein to profile the system using the information the instructions to cause the processing device further to:

trace the plurality of events using the information; and

analyze the plurality of events on a time basis.

19. The non-transitory machine-readable medium of claim 17, wherein to profile the system using the information the instructions to cause the processing device further to:

correlate the plurality of events using the information to identify causation relationships between the plurality of events.

20. The non-transitory machine-readable medium of claim 17, wherein the instructions further cause the processing device to:

store the time values in a repository; and

allow users to access the repository to retrieve the time values.

21. The non-transitory machine-readable medium of claim 17, wherein the instructions further cause the processing device to:

instrument the plurality of subsystems to collect the information during execution of an application program without modifying source code of the application program.

22. The non-transitory machine-readable medium of claim 17, wherein the event report further comprises a second identifier for a corresponding client machine, a third identifier for a corresponding subsystem in the corresponding client machine, a fourth identifier for an event type of the corresponding event, and a timestamp for the corresponding event.

23. The non-transitory machine-readable medium of claim 17, wherein the GUI comprises a directed acyclic graph showing the plurality of events along the timeline and a set of zero or more causation relationships between the plurality of events, wherein the measurement data comprises at least one of memory usage or processing device utilization data.