| 摘要 |
PROBLEM TO BE SOLVED: To simplify the setting of response monitoring time and also to quickly perform failure recovery by preventing useless processing request to another server and cancellation processing from being performed even when many processing requests are accumulated due to the load increase of a self-server and, a call destination server and a network failure. SOLUTION: Response monitoring time-out time(TO) is added to processing request data to a call destination when a call source issues a PRC(remote procedure call). The response monitoring time-out time(TO) is inherited between respective servers. When a server after halfway furthers asks another server for processing, it sets difference between a call source time-out time(TO) and the present time as response monitoring time (steps 504 and 505). Also, when the call source time-out time(TO) passes (steps 501:N, 503:N and 508:N) and when response does not come from a call source even after the response monitoring time passes (step 506:N), processing is finished (step 510).
|
