| 摘要 |
With respect to a storage system in which quick formatting and sequential formatting can be run concurrently, the time it takes to process an access request from a host is prevented from becoming prolonged even when a normal sequential formatting process is executed with respect to a storage volume which frequently incurs I/O penalties. The storage device measures the load from the host per configurational unit (storage medium) of LUs, and divides the LUs into a group of LUs whose load per storage medium is low, and a group of LUs whose load per storage medium is high. Further, the density per unit of LU capacity of I/O penalties incurred in a storage volume for which quick formatting is being executed is calculated. Sequential formatting is then executed, with priority, with respect to the LUs belonging to the group with low loads and in order of descending density of incurred I/O penalties.
|
