| 摘要 |
When a fault occurs, a hot-standby cluster takes over the job of the cluster which is at fault, and the load distribution for each job of load-sharing clusters is changed, simplifying the processing for dealing with the fault. Thus the clusters are efficiently used and arranged freely. The system is provided with a plurality of clusters which are operated by one or more of an in-use hot-standby system, a hot-standby waiting system, and a load sharing system for each job according to a table, and of the cluster is provided with a means which instructs a hot-standby waiting cluster to take over the job of a faulty cluster when the job is of the in-use hot-standby system, or instructs another load-sharing cluster to take over the job in a case the job is of the load-sharing system, referring to the table for dealing with a fault at any one of the clusters. |
