Method to identify performance and capacity bottlenecks of complex systems
First Claim
Patent Images
1. A method for determining capacity in complex systems by setting capacity values for service transaction types that assist in identifying performance bottlenecks comprising:
- listing the service transaction types;
identifying Workflow Manager (WFM) data where invocation data records for each service transaction type are located;
retrieving the WFM data;
parsing the WFM data to acquire the invocation data records for each service transaction type;
for a given transaction type;
determining average response time RTh aggregation levels and volume counts counth, wherein the volume counts counth is the traffic load and is a count of successful transactions within a predetermined measurement interval;
performing a linear regression of data pairs (counth,RTh);
determining an average Response Time Model RTM(x);
estimating response time model z0, z1 and z2 parameters from the average response time RTh and volume counts counth data;
calculating a capacity warning valve that the complex system can handle;
calculating a capacity warning value L1;
calculating a capacity limit value L2; and
analyzing a time trend of volume counts counth to determine when the capacity warning value L1 and capacity limit value L2 will be reached if current volume counts counth trends continue.
1 Assignment
0 Petitions
Accused Products
Abstract
Systems and methods are described that analyze complex systems and identify potential performance bottlenecks that may affect capacity and response time. The bottlenecks are identified to resolve problems originating at a specific subsystem(s).
-
Citations
21 Claims
-
1. A method for determining capacity in complex systems by setting capacity values for service transaction types that assist in identifying performance bottlenecks comprising:
-
listing the service transaction types; identifying Workflow Manager (WFM) data where invocation data records for each service transaction type are located; retrieving the WFM data; parsing the WFM data to acquire the invocation data records for each service transaction type; for a given transaction type; determining average response time RTh aggregation levels and volume counts counth, wherein the volume counts counth is the traffic load and is a count of successful transactions within a predetermined measurement interval; performing a linear regression of data pairs (counth,RTh); determining an average Response Time Model RTM(x); estimating response time model z0, z1 and z2 parameters from the average response time RTh and volume counts counth data; calculating a capacity warning valve that the complex system can handle; calculating a capacity warning value L1; calculating a capacity limit value L2; and analyzing a time trend of volume counts counth to determine when the capacity warning value L1 and capacity limit value L2 will be reached if current volume counts counth trends continue. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21)
-
-
13. The method according to claim 12 further comprising if the least squares regression slope s is negative, stopping further analysis.
-
14. The method according to claim 13 further comprising examining the invocation data records producing a negative slope s.
-
15. The method according to claim 1 wherein estimating response time model z0, z1 and z2 parameters further comprises modeling the RTh of a specific function RTM(x) for a given load x as the sum of the sojourn time in each service node required to process the function as
-
x , wherein z0 is the expected amount of time to traverse non-bottlenecked nodes, z1 is the expected amount of time to traverse a bottleneck node under no transaction load and z2 is the average amount of time of a bottleneck node per transaction type.
-
-
16. The method according to claim 12 further comprising using the data pairs (counth,RTh) to estimate the z0, z1 and z2 parameters.
-
17. The method of claim 16 wherein to estimate the z0, z1 and z2 parameters further comprises:
-
letting the z0, z1 and z2 parameters be constraint values that minimize the sum of (log(RTh)−
log(RTM(counth|z)))2; andusing a conjugate-gradient method to find z0≧
0, z1 and z2>
0.
-
-
18. The method according to claim 17 wherein when z0 is not equal to zero, the capacity warning value L1 is
-
z 1 + z 1 2 ) 1 2 z 0 z 2 and when z0 is equal to zero, the capacity warning value L1 is
-
-
19. The method according to claim 17 wherein when z0 is not equal to zero, the capacity limit value L2 is
-
z 0 + 3 z 1 - ( 8 z 0 z 1 + 9 z 1 2 ) 1 2 4 z 0 z 2 and when z0 is equal to zero, the capacity limit value L2 is
-
-
20. The method according to claim 17 wherein z0+z1 is the best expected response time that can be achieved.
-
21. The method according to claim 1 wherein the Response Time Model RTM(x) is
-
x .
-
Specification