| 摘要 |
PROBLEM TO BE SOLVED: To provide a cluster system and a device and a method for managing operation thereof with which trouble to be estimated in the case of power source fault occurrence can be automatically avoided. SOLUTION: When any abnormality occurs in a commercial power supply system for a UPS 30a, an operation managing device 10a sends a shutdown instruction S2 to a server 2a. The server 2a transfers the shutdown instruction S2 to other servers 2b-2d constituting the cluster system and performs shutdown processing by itself. The servers 2b-2d, which receive the shutdown instruction S2 from the server 2a, respectively perform shutdown processing as well. The servers 2a-2d respectively send a shutdown completion notice S3 to operation managing devices 10a-10c and the UPS 30a. The operation managing devices 10a-10d and UPS 20a, which receive this notice, respectively stop supplying power to the servers 2a-2d. Further, the UPS 20a stops supplying power to RAID 4a-4c. |
