Method for building enterprise scalability models from load test and trace test data
First Claim
1. A method of scaling a production network, the production network including a client computer connected to a plurality of servers on which a set of business functions are running, where each business function is a user task having a sequence of transactions between the client computer and the plurality of servers, comprising the steps of:
- initiating a single business function trace test on the production network for a single business function in the set of business functions;
gathering network flow trace data from the production network;
deriving a transaction summary from the network flow trace data wherein the transaction summary comprises a tabular listing of the sequence of transactions for each business function in the set of business functions, the tabular listing including information related to resource usage within the plurality of servers;
initiating a single business function load test on the production network;
gathering a set of resource data from the plurality of servers;
normalizing the resource data;
wherein the step of normalizing further comprises;
synchronizing and merging the set of resource data into a single table,converting the set of resource data to a set of consistent units,equalizing the set of resource data to a set of standard measurement intervals by extrapolating a set of utilization data within the set of resource data and, aggregating a set of servers, from the plurality of servers, into a mega server;
gathering web log data from at least one server from the plurality of servers for the single business function;
deriving a measured number of executions of the single business function on the production network during a predefined time period on the production network;
calculating a first consumption of resources used by the plurality of servers;
calculating a business function cost for the single business function by dividing the first consumption of resources by the measured number of executions of the single business function;
creating a business function profile from the transaction summary and the business function cost for the single business function;
building a discrete event simulation model from the business function profile;
optimizing the production network to provide a scaled production network, by performing a simulation with the discrete event simulation model in a simulation application running on an analysis computer; and
producing a report of the scaled production network on the analysis computer.
5 Assignments
0 Petitions
Accused Products
Abstract
A method is disclosed that preparing a discrete event simulation model of a production network is provided including a client and a plurality of servers connected by a network comprising the steps of initiating a single business function trace test on the production network; gathering network flow trace data from the network; deriving a transaction summary from the network flow trace data; initiating a single business function load test on the production network; gathering resource data from the plurality of servers; normalizing the resource data; gathering web log data from at least one server from the plurality of servers; calculating a number of business functions running on the production network; calculating a consumption of resources used by the plurality of servers; calculating a business function cost associated with the consumption of resources; creating a business function profile from the transaction summary and the business function cost; and building a discrete event simulation model from the business function profile.
34 Citations
16 Claims
-
1. A method of scaling a production network, the production network including a client computer connected to a plurality of servers on which a set of business functions are running, where each business function is a user task having a sequence of transactions between the client computer and the plurality of servers, comprising the steps of:
-
initiating a single business function trace test on the production network for a single business function in the set of business functions; gathering network flow trace data from the production network; deriving a transaction summary from the network flow trace data wherein the transaction summary comprises a tabular listing of the sequence of transactions for each business function in the set of business functions, the tabular listing including information related to resource usage within the plurality of servers; initiating a single business function load test on the production network; gathering a set of resource data from the plurality of servers; normalizing the resource data; wherein the step of normalizing further comprises; synchronizing and merging the set of resource data into a single table, converting the set of resource data to a set of consistent units, equalizing the set of resource data to a set of standard measurement intervals by extrapolating a set of utilization data within the set of resource data and, aggregating a set of servers, from the plurality of servers, into a mega server; gathering web log data from at least one server from the plurality of servers for the single business function; deriving a measured number of executions of the single business function on the production network during a predefined time period on the production network; calculating a first consumption of resources used by the plurality of servers; calculating a business function cost for the single business function by dividing the first consumption of resources by the measured number of executions of the single business function; creating a business function profile from the transaction summary and the business function cost for the single business function; building a discrete event simulation model from the business function profile; optimizing the production network to provide a scaled production network, by performing a simulation with the discrete event simulation model in a simulation application running on an analysis computer; and producing a report of the scaled production network on the analysis computer. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8)
-
-
9. A system for optimizing and scaling a production network, the production network comprising a plurality of servers which operates a set of business functions, where each business function is a user task having a sequence of transactions between a client computer and the plurality of servers, a collection of network sniffers deployed on the production network, an analysis computer connected to the production network and the collection of network sniffers, the analysis computer programmed to carry out the steps of:
-
loading a discrete event simulation tool; gathering resource data from a business function load test on the plurality of servers; wherein the resource data includes a reporting interval, a plurality of timestamps, and a plurality of resource utilization indicators for a single business function in the set of business functions; merging the resource data into a single table; equalizing at least one reporting interval in the resource data to a standard measurement interval by extrapolating a set of utilization data within the resource data; synchronizing at least one time stamp in the resource data; converting at least one set of units in the plurality of resource utilization indicators to a set of consistent units; aggregating a set of servers, from the plurality of servers, into a mega server; gathering a set of network data from a functional trace test carried out by the collection of network sniffers; deriving a transaction summary from the set of network data wherein the transaction summary comprises a tabular listing of the sequence of transactions for each business function in the set of business functions, the tabular listing including information related to resource usage within the plurality of servers; deriving a set of web log data from the function load test; calculating a number of business functions from the transaction summary; calculating a resource consumption for the plurality of servers; calculating a business function cost for the single business function by dividing the resource consumption by a measured number of executions of the single business function; deriving a single business function profile for the single business function from the resource data, the business function cost for the single business function and the number of business functions in the set of business functions; adopting a transaction flow from the transaction summary; building a discrete event simulation model from the single business function profile and the transaction flow; and
,performing a simulation of the production network with the discrete event simulation model in the discrete event simulation tool and producing a report displayed on the analysis computer which optimizes and scales the production network. - View Dependent Claims (10, 11, 12, 13, 14, 15, 16)
-
Specification