Abstract: In a business system constructed by a web server, an AP server, and a DB server serially connected in this order, the times required for the transmission and reception of messages with respect to the servers or business applications located in downstream processes of the servers, are respectively measured. Then, performance information associated with, at the very least, these times is created for the identifier that identifies the business application, and this is transmitted to and collected in the administration server. The processing time spent in each server and business application, if under normal circumstances, is very small. Therefore if this deviates from a predetermined range, it can be easily identified that the business application thereof is being a bottleneck.

Abstract: A program recording medium recording a communication control program to be implemented in a computer including a plurality of control means for controlling data communication in a hierarchical configuration and data transfer control means for controlling data transfer among the plurality of control means. The program causes the data transfer control means to add anomaly information into data to be transferred to the plurality of control means upon detection of the anomaly indication, to send discard indication, that is the data containing the added anomaly information, at substantially the same time to the plurality of control means, and to discard the data containing the anomaly indication and data relating thereto.

Abstract: A program recording medium recording a communication control program to be implemented in a computer including a plurality of control means for controlling data communication in a hierarchical configuration and data transfer control means for controlling data transfer among the plurality of control means. The program causes the data transfer control means to add anomaly information into data to be transferred to the plurality of control means upon detection of the anomaly indication, to send discard indication, that is the data containing the added anomaly information, at substantially the same time to the plurality of control means, and to discard the data containing the anomaly indication and data relating thereto.

Abstract: A load sharing system composed of a plurality of servers that are interconnected over a network for providing optimum and efficient operations at high reliability. The system employs a plurality of servers and a communication controller. A state manager in the communication controller first establishes connection paths on the network to reach the state management agents. The state manager then performs a centralized management of activity status of the servers based on status information collected from the state management agents via the established connection paths. A distribution processor disposed in the communication controller receives service requests from user terminals via a network and determines their respective destinations according to management data provided by the state manager. The distribution processor then forwards the service requests to the servers over the communication medium to direct their execution.