An operation site has copy management information including state information that is used when executing a remote copy. The copy management information is stored in a storage device included in each of all storage device subsystems in this computer system. The copy management information is updated every time a remote copy is executed. When the operation site fails, by use of a copy management program in another operatable site, and with reference to the copy management information stored in the storage device of the storage device subsystem in each site, the copy is executed from a point to which the copy processing has progressed before the failure. As a result, when the production site in which an application is operated suffers from a disaster, it is possible to quickly recover data at the time of the disaster by executing a disaster recovery management program from another site in which no failure occurs, and then by restarting the remote copy.