System and method for collecting response times for exception response applications

• US 5,331,574 A
• Filed: 08/06/1991
• Issued: 07/19/1994
• Est. Priority Date: 08/06/1991
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Weddington, Harry F. Jr., Temoshenko, Leo
• Primary Examiner(s): Harvey, Jack B.
• Assistant Examiner(s): CHOI, KYLE JAEHUN
• Application Number:
  US07/741,005
• Time in Patent Office: 1,078 Days
• Field of Search: 364/514, 364/551.01, 364/DIG. 1, 364/264, 364/264.4, 364/264.5, 364/569, 395/200, 395/275
• US Class Current: 702/186
• CPC Class Codes: G06F 11/3419 by assessing time
• CPC Class Codes: G06F 2201/87 Monitoring of transactions
• CPC Class Codes: H04L 43/0829 Packet loss

Abstract

A system and method for performance monitoring of transactions which involve an application which normally operates in exception response mode. A network performance monitor sets information in memory regarding applications and resources for which performance data is sought. The applications are modified so that they check the appropriate locations in memory during the processing for each transaction to check whether the mode should be switched to definite response mode for that transaction. If so, the application switches to definite response mode, thereby causing a definite response path information unit to be sent by the resource following the resource'"'"'s reception of the outbound path information unit from the application. The network performance monitor monitors the passage of these messages through a communication access means which operates the network. The network performance monitor uses the passage time of each message to obtain performance time for the given transaction relative to the given resource and at the application. Following the completion of each transaction, the application can return to exception response mode so that the definite response message will only be sent from the resource when performance monitoring is occurring, thereby reducing network overhead.

Claims

1. A system for obtaining performance data related to an application running in exception response mode, comprising:

means for communicating information to the application running in exception response to switch dynamically to definite response mode for a selected resource for at least one transaction;

means for monitoring for messages related to the at least one transaction; and

means for obtaining performance data based on data from said monitoring means.

2. The system for obtaining performance data according to claim 1, wherein the application runs on a network, the network includes a communication access means, and said monitoring means monitors the communication access means for relative times at which messages related to the at least one transaction pass therethrough to and from the application.

3. The system for obtaining performance data according to claim 2, wherein the performance data include host and network times between messages passing back and forth between the application and a resource employing the application for each of the at least one transaction.

4. The system for obtaining performance data according to claim 1, wherein the performance data can be sought relative to any one of the application and one or more resources employing the application.

5. The system for obtaining performance data according to claim 1, wherein said communicating means sets information in memory accessible by the application which defines situations to switch to definite response mode.

6. The system for obtaining performance data according to claim 1, wherein the messages include an inbound path information unit bound for the application from a resource employing the application, an outbound path information unit bound for the resource from the application and an inbound definite response path information unit bound for the application from the resource.

7. The system for obtaining performance data according to claim 1 wherein said means for communicating information includes means for deactivating said definite response mode for said selected resource to revert dynamically to exception response mode for said selected resource following said at least one transaction.

8. A method for performance monitoring of transactions in a network associated with applications operating in exception response mode, comprising the steps of:

a) modifying the applications to check memory for instructions for changing modes;

b) placing the instructions in memory regarding applications and resources for which mode changes are desired;

c) monitoring for messages for transactions for which any application has changed modes to definite response mode; and

d) processing data relative to the monitored messages to obtain performance data relative to the applications and resources employing the application.

9. The method according to claim 8, wherein said step a further comprises modifying the applications to check the memory relative to each transaction processed therein.