Technique for mapping goal violations to anamolies within a system

• US 7,673,189 B2
• Filed: 02/06/2006
• Issued: 03/02/2010
• Est. Priority Date: 02/06/2006
• Status: Active Grant

• Current Assignee: Twitter Inc (X Holdings Corp.)
• Original Assignee: International Business Machines Corporation
• Inventors: Gomez, Juan C., Duyanovich, Linda M., Pollack, Kristal T., Uttamchandani, Sandeep M.
• Primary Examiner(s): Baderman; Scott T
• Assistant Examiner(s): PATEL, KAMINI B
• Application Number:
  US11/347,999
• Publication Number:
  US 20070220371A1
• Time in Patent Office: 1,485 Days
• Field of Search: 714/47
• US Class Current: 714/47.1
• CPC Class Codes: G06F 11/3447 Performance evaluation by modeling
• CPC Class Codes: G06F 11/3485 for I/O devices
• CPC Class Codes: G06F 2201/86 Event-based monitoring

Abstract

A computer and method for problem detection and determination for automated system management in a system, wherein the method comprises monitoring system state, workload, and performance parameters of the system; comparing the monitored parameters against normal system performance behavior of the system, wherein the normal system performance behavior is maintained as a mapping of a system state and workload-to-performance parameters; summarizing performance abnormalities at a specified layer in the system as computation and data-processing attributes, wherein the performance abnormalities comprise deviations from the normal system performance behavior; correlating the performance abnormalities across multiple layers in the system using an attribute-based framework; and communicating a root-cause of the performance abnormalities.

Claims

1. A machine-implemented method for problem detection and determination for automated system management in a system, said method comprising:

monitoring, by a computing device, system state, workload, and performance parameters of said system;

comparing, by a computing device, the monitored parameters against normal system performance behavior of said system, wherein said normal system performance behavior is maintained as a mapping of a behavior of physical components as a function of load characteristics;

summarizing performance abnormalities, by a computing device, at a specified layer in said system as computer processing unit (CPU) and input/output (I/O) attributes, wherein said performance abnormalities comprise deviations from said normal system performance behavior;

correlating said performance abnormalities, by a computing device, across multiple layers in said system using a query-style interaction interface; and

communicating, by a computing device, a root-cause of said performance abnormalities based on a query index expressing a strategy of how an optimal performance behavior of said system can be adapted.

2. The method of claim 1, wherein said monitoring said system state, workload, and performance parameters comprises parsing said monitored data for values associated with said computer processing unit (CPU) and said input/output (I/O) work attributes, and wherein said system state, workload, and performance parameters are specific to a protocol of said specified layer in said system.

3. The method of claim 1, wherein said summarizing of performance abnormalities at a specified layer in said system as computer processing unit (CPU) and input/output (I/O) attributes comprises comparing said computer processing unit (CPU) and input/output (I/O) work attributes with said mapping of said normal performance behavior of said system.

4. The method of claim 1, further comprising:

communicating, by a computing device, said root-cause of said performance abnormalities in a query format by;

providing, by a computing device, a feasibility of implementing said strategy; and

optimizing, by a computing device, said strategy with respect to other performance strategies for running said system.

5. The method of claim 1, wherein said monitoring system state, workload, and performance parameters comprises measuring parameters related to any of a switch, storage controller, and computer disks of said system.

6. The method of claim 3, wherein said comparing CPU and I/O work attributes with a history of said normal performance behavior of said system comprises retrieving a current system state and workload in said mapping and comparing a performance from said mapping with a current performance of said system to determine if said system comprises said performance abnormalities.

7. The method of claim 6, further comprising analyzing said CPU and I/O work attributes using at least one of:

searching, by a computing device, said mapping for said load characteristics that are closest to a current load being applied to said system; and

determining, by a computing device, a change in said load characteristics of monitored data in comparison to a last n number of snapshots when performance goals of said system are being met.

8. A computer for problem detection and determination for automated system management in a system, said computer comprising:

modules comprising system state, workload, and performance parameters of said system capable of being monitored;

a first processor element adapted to compare the monitored parameters against normal system performance behavior of said system, wherein said normal system performance behavior is maintained as a mapping of a behavior of physical components as a function of load characteristics;

a second processor element adapted to summarize performance abnormalities at a specified layer in said system as computer processing unit (CPU) and input/output (I/O) attributes, wherein said performance abnormalities comprise deviations from said normal system performance behavior;

a query-style interaction interface adapted to correlate said performance abnormalities across multiple layers in said system; and

an output unit adapted to communicate a root-cause of said performance abnormalities based on a query index expressing a strategy of how an optimal performance behavior of said system may be adapted.

9. The computer of claim 8, further comprising:

a parser adapted to parse said monitored data for values associated with computer processing unit (CPU) and input/output (I/O) work attributes, and wherein said behavior of physical components as a function of load characteristics are specific to a protocol of said specified layer in said system.

10. The computer of claim 8, further comprising:

a comparator adapted to compare computer processing unit (CPU) and input/output (I/O) attributes with said mapping of said normal performance behavior of said system.

11. The computer of claim 8, further comprising:

said query index adapted to communicate said root-cause of said performance by;

providing a feasibility of implementing said strategy; and

optimizing said strategy with respect to other performance strategies for running said system.

12. The computer of claim 8, wherein said system state, workload, and performance parameters comprise parameters related to any of a switch, storage controller, and computer disks of said system.

13. The computer of claim 10, wherein said comparator is adapted to retrieve a current system state and workload in said mapping and comparing a performance from said mapping with a current performance of said system to determine if said system comprises said performance abnormalities.

14. The computer of claim 13, further comprising:

a cache searcher adapted to search said mapping for said load characteristics that are closest to a current load being applied to said system; and

a calculator adapted to determine a change in said load characteristics of monitored data in comparison to a last n number of snapshots when performance goals of said system are being met.

15. A program storage device readable by computer, tangibly embodying a program of instructions executable by said computer to perform a method for problem detection and determination for automated system management in a system, said method comprising:

monitoring system state, workload, and performance parameters of said system;

comparing the monitored parameters against normal system performance behavior of said system, wherein said normal system performance behavior is maintained as a mapping of a behavior of physical components as a function of load characteristics;

summarizing performance abnormalities at a specified layer in said system as computer processing unit (CPU) and input/output (I/O) attributes, wherein said performance abnormalities comprise deviations from said normal system performance behavior;

correlating said performance abnormalities across multiple layers in said system using a query-style interaction interface; and

communicating a root-cause of said performance abnormalities based on a query index expressing a strategy of how an optimal performance behavior of said system can be adapted.

16. The program storage device of claim 15, wherein said monitoring of said system state, workload, and performance parameters comprises parsing said monitored data for values associated with said computer processing unit (CPU) and said input/output (I/O) work attributes, and wherein said system state, workload, and performance parameters are specific to a protocol of said specified layer in said system.

17. The program storage device of claim 15, wherein said summarizing of performance abnormalities at a specified layer in said system as computation and data-processing attributes comprises comparing said computer processing unit (CPU) and said input/output (I/O) work attributes with said mapping of said normal performance behavior of said system.

18. The program storage device of claim 15, wherein said method further comprises:

communicating said root-cause of said performance abnormalities in a query format by;

providing a feasibility of implementing said strategy; and

optimizing said strategy with respect to other performance strategies for running said system.

19. The program storage device of claim 15, wherein said monitoring system state, workload, and performance parameters comprises measuring parameters related to any of a switch, storage controller, and computer disks of said system.

20. The program storage device of claim 17, wherein said comparing CPU and I/O work attributes with a history of said normal performance behavior of said system comprises retrieving a current system state and workload in said mapping and comparing a performance from said mapping with a current performance of said system to determine if said system comprises said performance abnormalities.

21. The program storage device of claim 20, wherein said method further comprises analyzing said CPU and I/O work attributes using at least one of:

searching said mapping for said load characteristics that are closest to a current load being applied to said system; and

determining a change in said load characteristics of monitored data in comparison to a last n number of snapshots when performance goals of said system are being met.