Method for building enterprise scalability models from production data

• US 8,271,643 B2
• Filed: 08/05/2009
• Issued: 09/18/2012
• Est. Priority Date: 02/01/2006
• Status: Active Grant

• Current Assignee: CA, Inc. (d/b/a CA Technologies) (Broadcom, Inc.)
• Original Assignee: CA, Inc. (d/b/a CA Technologies) (Broadcom, Inc.)
• Inventors: Wise, Timothy E., Clarke, Allan Drew, Gimarc, Richard, Reynolds, James, Spellmann, Amy, Barnett, Paul T.
• Primary Examiner(s): Moise, Emmanuel L
• Assistant Examiner(s): MURPHY, CHARLES C
• Application Number:
  US12/462,547
• Publication Number:
  US 20100106478A1
• Time in Patent Office: 1,140 Days
• Field of Search: 709/238, 709/231, 709/224, 709/223, 709/201, 709/206, 703/22, 703/21, 703/19, 703/17, 703/13, 719/328, 719/315, 719/311
• US Class Current: 709/224
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling

Abstract

The invention provides a method for preparing a discrete event simulation model of a production network including a client and a plurality of servers on a plurality of network tiers connected by a network comprising the steps of executing a business function from the client on the production network, gathering resource data from the plurality of servers, normalizing the resource data, calculating a consumption of resources used by the plurality of servers, estimating a transaction cost associated with the consumption of resources and building a discrete event simulation model from the transaction cost.

Claims

1. A non-transitory computer readable medium containing a computer program product storing a set of computer executable instructions, the set of computer executable instructions operable on a host computer system, including a CPU, that when executed creates a discrete event simulation model of a production network including a client and a plurality of servers on a plurality of network tiers connected by a network, the plurality of servers executing a plurality of applications, and running a business function, the set of computer executable instructions including the steps of:

executing the business function, including accessing a plurality of applications running on the production network, from the client on the production network;

gathering resource utilization data from the plurality of servers in a set of reporting intervals;

normalizing the resource utilization data by performing the substeps of;

synchronizing the resource utilization data by correlating a set of timestamp data, adjusting the resource utilization data for consistent physical units, extrapolating the resource utilization data in the set of reporting intervals into a set of equalized time intervals, aggregating the resource utilization data for the plurality of servers into an aggregated resource utilization for each network tier in the plurality of network tiers;

calculating an aggregated resource consumption in a measurement time period by multiplying the aggregated resource utilization by the measurement time period for each network tier in the plurality of network tiers;

estimating a set of transaction cost from the aggregated resource consumption;

building a discrete event simulation model from the set of transaction cost; and

,

storing the discrete event simulation model on the non-transitory computer readable medium.

2. The non-transitory computer readable medium of claim 1 wherein the step of gathering further comprises the steps of:

deploying one of a plurality of resource monitors on each server resident on the production network;

collecting the resource utilization data from the plurality of resource monitors including collecting the set of time stamp data; and

merging the resource utilization data into a single file for analysis.

3. The non-transitory computer readable medium of claim 2 wherein the step of building a discrete event simulation model further comprises the steps of:

assembling a set of simulation instructions incorporating the set of transaction costs;

approximating a flow of a transaction associated with the business function;

calculating data throughput on a network interface of the production network; and

assembling the approximated flow of a transaction and the calculated data throughput into the set of simulation instructions.

4. The non-transitory computer readable medium of claim 1 wherein the plurality of network tiers includes a first network tier comprising a set of web servers, a second network tier comprising a set of application servers and a third network tier comprising a set of database servers.

5. The non-transitory computer readable medium of claim 4 wherein the step of approximating a flow of a transaction further comprises the steps of:

aggregating a first set of servers into a first combined tier of servers in the plurality of servers;

aggregating a second set of servers into a second combined tier of servers in the plurality of servers; and

,

approximating the flow of a transaction between the first combined tier of servers and the second combined tier of servers.

6. The non-transitory computer readable medium of claim 4 wherein the step of calculating data throughput further comprises the step of:

repeating the step of calculating the aggregated resource consumption for a set of measurement time periods and accumulating a running total of consumption of resources for each network tier in the plurality of network tiers.

7. The non-transitory computer readable medium of claim 1 wherein the step of estimating a set of transaction cost further comprises the step of:

counting a number of transactions performed during the measurement time period;

dividing the aggregated resource consumption by the counted number of transactions.

8. The non-transitory computer readable medium of claim 1 comprising the further step of:

running the discrete event simulation model on a discrete event simulation modeler.

9. The non-transitory computer readable medium of claim 8 wherein the number of transactions is derived from a transaction log.

10. The non-transitory computer readable medium of claim 8 wherein the number of transactions is derived from a load test.

11. The non-transitory computer readable medium of claim 3 wherein the step of approximating a flow of a transaction further comprises the steps of:

assuming an iteration count between at least two alternate paths on the non-linear production network;

assigning weights to each of the at least two alternate paths on the non-linear production network based on the iteration count; and

within a simulation, allocating the aggregated resource consumption among the at least two alternate paths according to the assigned weights.

12. A non-transitory computer readable medium containing a computer program product storing a set of computer executable instructions, the set of computer executable instructions operable on a host computer system, including a CPU, that when executed creates a discrete event simulation model comprising the program steps of:

gathering resource data from a plurality of resource monitors operatively disposed on a plurality of servers resident on a multi-tiered computer system executing a business function;

the business function accessing a plurality of applications running on the multi-tiered computer system;

wherein the resource data includes a set of reporting intervals, a set of timestamps, a resource utilization, a received packet count and a sent packet count;

equalizing at least one reporting interval in the resource data by extrapolating the resource data in the set of reporting intervals into a set of equalized time intervals;

synchronizing at least one time stamp in the set of timestamps;

normalizing at least one set of units in the resource data;

selecting a subset of the plurality of servers;

calculating a resource consumption for the plurality of servers as a combined resource consumption for the subset of the plurality of servers wherein the combined resource consumption is calculated by multiplying the resource utilization by a measured time period;

calculating a transaction cost from the combined resource consumption;

approximating a transaction flow for a plurality of transactions related to the business function;

calculating a network throughput between the plurality of servers; and

building a discrete event simulation model from the transaction cost, the transaction flow and the network throughput; and

storing the discrete event simulation model on the non-transitory computer readable medium.

13. The non-transitory computer readable medium of claim 12 wherein the steps of selecting a subset of the plurality of servers, calculating a resource consumption, calculating a transaction cost, approximating a transaction flow, calculating a network throughput, building a discrete event simulation model and storing the discrete event simulation model are repeated for a set of web servers, a set of application servers and a set of database servers in the plurality of servers.

14. The non-transitory computer readable medium of claim 12 wherein the program step of approximating a transaction flow comprises the further program step of:

estimating the transaction flow of a linear multi-tiered system.

15. The non-transitory computer readable medium of claim 12 wherein the program step of approximating a transaction flow comprises the further program step of:

estimating the transaction flow of a non-linear multi-tiered system.

16. The non-transitory computer readable medium of claim 15 wherein the program step of estimating the transaction flow of a non-linear multi-tiered system comprises the further program step of:

performing a discrete event simulation by weighting a set of alternate paths on the multi-tiered system based on a percentage of usage of each path in the set of alternate paths.

17. A system for creating a discrete event simulation model, comprising an analysis computer communicatively connected to a multi-tiered computer system, the multi-tiered computer system executing a business function, the analysis computer programmed to perform the steps of:

gathering a set of resource data from a plurality of resource monitors operatively disposed on a plurality of servers resident on each tier of the multi-tiered computer system;

the business function accessing a plurality of applications running on the multi-tiered computer system;

wherein the set of resource data includes a set of reporting intervals, a set of timestamps, a resource utilization, a received packet count and a sent packet count;

equalizing at least one reporting interval in the resource data by extrapolating the resource data in the set of reporting intervals into a set of equalized time intervals;

synchronizing at least one time stamp in the set of timestamps;

normalizing at least one set of units in the resource data;

calculating a resource consumption for each tier of the multi-tiered computer system by multiplying the resource utilization by a reporting time interval in the set of reporting intervals;

calculating a transaction cost from the resource consumption;

approximating a transaction flow for a plurality of transactions related to the business function;

calculating a network throughput between the plurality of servers; and

building a discrete event simulation model from the transaction cost, the transaction flow and the network throughput; and

executing the discrete event simulation model on the analysis computer.

19. The computer system of claim 17 wherein the step of normalizing further comprises aggregating resource data from the plurality of servers, residing on each tier of the multi-tiered computer system, into a single server model.

20. The computer system of claim 17 including the steps of selecting a set of web servers as a first tier of the multi-tiered computer system, selecting a set of application servers as a second tier in the multi-tiered computer system and selecting a set of database servers as a third tier in the multi-tiered computer system.

18. The computer system of 17 wherein the step of calculating the resource consumption for the plurality of servers further comprises interpolating a at least one resource utilization value from the set of resource data for the at least one reporting interval.