STFS: a novel file system for efficient small writes

This paper presents a novel high performance file system: Same Track File System (STFS) based on the researches of the character of modern disk drive. This paper firstly analyses the non-linear seek latency character of modern disk drive, and then indicates the drawback of FFS due to small writes. Then, this paper describes a novel disk layout of STFS: putting the meta-data and data into the same track. This disk layout of STFS can effectively depress the seek overhead of small writes and greatly improve file system throughput. The experiments of this paper show that STFS is a high performance file system.

[1]  Wei Hu,et al.  Scalability in the XFS File System , 1996, USENIX Annual Technical Conference.

[2]  John Wilkes,et al.  An introduction to disk drive modeling , 1994, Computer.

[3]  M. Frans Kaashoek,et al.  Embedded Inodes and Explicit Grouping: Exploiting Disk Bandwidth for Small Files , 1997, USENIX Annual Technical Conference.

[4]  J. Kent Peacock,et al.  The Counterpoint Fast File System , 1988, USENIX Winter.

[5]  Yale N. Patt,et al.  On-line extraction of SCSI disk drive parameters , 1995, SIGMETRICS '95/PERFORMANCE '95.

[6]  Steve R. Kleiman,et al.  Extent-like Performance from a UNIX File System , 1991, USENIX Winter.

[7]  Mendel Rosenblum,et al.  The design and implementation of a log-structured file system , 1991, SOSP '91.

[8]  Robert S. Fabry,et al.  A fast file system for UNIX , 1984, TOCS.

[9]  Mary Baker,et al.  Measurements of a distributed file system , 1991, SOSP '91.