Self-evolving distributed system

• US 20060262726A1
• Filed: 04/28/2006
• Published: 11/23/2006
• Est. Priority Date: 03/25/2005
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Lieuallen, Brian R., Singhal, Sandeep Kishan, Manion, Todd R., Samuel, Geogy A., Norton, Noah
• Granted Patent: US 7,698,239 B2
• Time in Patent Office: Days
• US Class Current: 370/241
• CPC Class Codes: H04L 41/145 involving simulating, designing, planning or modelling of a network
• CPC Class Codes: H04L 41/5003 Managing SLA; Interaction between SLA and QoS
• CPC Class Codes: H04L 43/0817 by checking functioning
• CPC Class Codes: H04L 43/50 Testing arrangements
• CPC Class Codes: H04L 63/0272 Virtual private networks

Abstract

In a distributed computing system, an artificial intelligence system may be employed to configure the network variables. A metric describing the overall system performance may be derived during network operation or simulation and compared to an ideal metric describing the same distributed system performance. The difference between the derived metric and the ideal metric may then be used with an artificial intelligence system to modify the network variables to evolve the system toward the ideal performance standard.

Claims

1. A method of improving the performance of a distributed computing system comprising:

determining a first system health index, the first system health index comprising a first measurement of system statistics;

propagating the first measurement of system statistics through at least one artificial intelligence system;

determining at least one improved system behavior variable;

determining a second system health index, the second system health index comprising a second measurement of system statistics related to the at least one improved system behavior variable; and

determining if the at least one improved system behavior variable enhanced the performance of the distributed system.

2. The method of claim 1, wherein determining the at least one improved system behavior variable comprises configuring the artificial intelligence system based on a comparison of the first system health index and the second system health index.

3. The method of claim 2, further comprising:

applying a learning algorithm to the artificial intelligence system; and

deriving the at least one improved system behavior variable from the artificial intelligence system.

4. The method of claim 3, wherein the learning algorithm is a back propagation algorithm.

5. The method of claim 1, wherein the at least one improved system behavior variable comprises a system timeout value, a distributed hash table keyspace size, a routing table size, a maximum number of distributed system nodes, or a maximum number of allowable packet hops.

6. The method of claim 1, wherein the first measurement of system statistics and the second measurement of system statistics comprises a number of distributed system nodes, an average distributed system latency, an average distributed system packet loss, an operation frequency, or a distributed system node cache chum rate.

7. The method of claim 1, wherein the distributed system is one of a live distributed system or a simulated distributed system.

8. The method of claim 1, wherein the at least one artificial intelligence system is a neural network.

9. The method of claim 8, wherein the neural network comprises at least one hidden layer node, the at least one hidden layer node comprising a function, the function representing an interaction between a plurality of system behavior variables.

10. A method of improving the performance of a distributed computing system comprising:

determining a first system health index, the first system health index comprising at least one system performance parameter, the at least one system performance parameter comprising system statistics related to at least one system behavior variable;

propagating the at least one system performance parameter through at least one neural network;

determining at least one improved system behavior variable from the at least one system performance parameter;

saving the at least one improved system behavior variable to the distributed system;

determining a second system health index, the second system health index comprising the at least one system performance parameter, the at least one system performance parameter comprising system statistics related to the at least one configured system behavior variable;

calculating a system health index error, the health index error comprising a comparison of the first system health index and the second system health index;

determining if the improved system behavior variable enhanced the performance of the distributed system; and

if the improved system behavior variable did not enhance the performance of the distributed system, modifying the at least one neural network according to a neural network learning algorithm, the neural network learning algorithm comprising the system health index error;

wherein the neural network learning algorithm comprises a back propagation algorithm; and

wherein the artificial intelligence system comprises an input, a hidden layer, and an output, the input comprising the at least one least one system performance parameter, the output comprising the improved system behavior variable.

11. The method of claim 10, wherein the distributed system is one of a live distributed system or a simulated distributed system.

12. A computer-readable medium comprising computer-executable instructions for improving the performance of a distributed computing system comprising:

computer executable instructions for determining a first system health index, the first system health index comprising at least one system performance parameter, the at least one system performance parameter comprising system statistics related to at least one system behavior variable;

computer executable instructions for propagating the at least one system performance parameter through at least one artificial intelligence system;

computer executable instructions for determining at least one improved system behavior variable from the at least one system performance variable;

computer executable instructions for determining a second system health index, the second system health index comprising the at least one system performance parameter, the at least one system performance parameter comprising system statistics related to the improved system behavior variable;

computer executable instructions for comparing the first system health index and the second system health index; and

computer executable instructions for determining if the improved system behavior variable enhanced the performance of the distributed system.

13. The computer-readable medium of claim 12, wherein the computer executable instructions for determining at least one improved system behavior variable from the at least one system performance parameter comprises configuring the artificial intelligence system based on the result of the computer executable instructions for comparing the first system health index and the second system health index.

14. The computer-readable medium of claim 13, wherein the computer executable instructions for determining at least one improved system behavior variable from the at least one system performance parameter further comprises applying a learning algorithm to the artificial intelligence system.

15. The computer-readable medium of claim 14, wherein the learning algorithm is a back propagation algorithm.

16. The computer-readable medium of claim 12, wherein the at least one system behavior variable comprises a system timeout value, a distributed hash table keyspace size, a routing table size, a maximum number of distributed system nodes, or a maximum number of allowable packet hops.

17. The computer-readable medium of claim 12, wherein the at least one system performance parameter comprises a number of distributed system nodes, an average distributed system latency, an average distributed system packet loss, an operation frequency, or a distributed system node cache churn rate.

18. The computer-readable medium of claim 12, wherein the distributed system is one of a live distributed system or a simulated distributed system.

19. The computer-readable medium of claim 12, wherein the at least one artificial intelligence system is a neural network.

20. The computer-readable medium of claim 19, wherein the neural network comprises at least one hidden layer node, the at least one hidden layer node comprising a function, the function representing an interaction between multiple system behavior variables.