| US 7,552,212 B2 | ||
| Intelligent performance monitoring based on user transactions | ||
| Bryan Christopher Chagoly, Austin, Tex. (US); Xiaoping Chen, Austin, Tex. (US); Andrew Jason Lavery, Austin, Tex. (US); Howard Milton McKinney, Austin, Tex. (US); and Kirk Malcolm Sexton, Austin, Tex. (US) | ||
| Assigned to International Business Machines Corporation, Armonk, N.Y. (US) | ||
| Filed on Oct. 22, 2004, as Appl. No. 10/971,472. | ||
| Prior Publication US 2006/0107255 A1, May 18, 2006 | ||
| Int. Cl. G06F 15/16 (2006.01); G06F 15/173 (2006.01); G06F 15/177 (2006.01) | ||
| U.S. Cl. 709—224 [709/203; 709/217] | 3 Claims |

| 1. A method for monitoring an application operating on a computer system, the method comprising:
monitoring an entry point of the application for a transaction to selectively monitor the transaction within a plurality of
transactions in the application in order to minimize computational overhead and data recording associated with monitoring
all of the plurality of transactions in the application, wherein the entry point is a location in the application where the
transaction is recorded by a monitoring application, and wherein each transaction in the plurality of transactions is associated
with a corresponding token that includes data that uniquely identifies an associated transaction and contains configuration
information required by a probe that monitors an entry point in the application where the associated transaction occurs to
determine an appropriate level of data recording for the associated transaction;
responsive to the entry point being reached by the application for the transaction, acquiring a token associated with the
transaction, wherein the token is acquired by looking up the transaction in a policy mapper, and wherein the policy mapper
maps a monitoring policy to the transaction from a list of monitoring policies;
monitoring the transaction according to the monitoring policy of the token associated with the transaction, wherein the monitoring
policy contains a host name, a user name, an application name, a transaction name, scheduling information that defines when
the transaction is recorded, a threshold that defines an execution time limit for the transaction, and the configuration information
that determines the appropriate level of data recording for the transaction, and wherein the monitoring policy of the token
determines what specific information is gathered about the transaction, and wherein the transaction includes a plurality of
sub-transactions, and wherein each sub-transaction within the plurality of sub-transactions is associated with the monitoring
policy of the token;
collecting the appropriate level of data recording for the transaction defined by the configuration information; and
sending the collected appropriate level of data recording for the transaction to the monitoring application, wherein the monitoring
application records the collected appropriate level of data recording for the transaction for use by a user of the monitoring
application.
|