Abstract: In distributed computing environment, the system-wide large-scale data storage is necessary, The conventional approach to distribute throughout the system is "Network file system" (NFS), which limits the size of data on the file system by the size of single NFS server volume, eventhough applying RAID disks. In addition, the central server raises the single point of failure and the I/O bottleneck problem which reducing overall system performance.To provide the system a single large storage space, Scalable Distributed file system (SDFS) was concocted. The SDFS illustrates users a logical huge volume of a disk drive connecting to user's machines. This volume is accumulated from several partitions of disk drives in the distributed environment. Unknown to users, an I/O access to an area on this volume is redirected to the corresponding on certain machines in which data are actually stored. Thanks to its native distributing operation, the SDFS help to alleviate the bottleneck problem tremendously.