RAID0.5: Active Data Replication for Low Cost Disk Array Data Protection

RAID has long been established as an effective way to provide highly reliable as well as high-performance disk subsystems. However, reliability in RAID systems comes at the cost of extra disks. In this paper, we describe a mechanism that we have termed RAID0.5 that enables striped disks with very high data reliability but low disk cost. We take advantage of the fact that most disk systems use backup systems for disaster recovery. With the use of these backup systems, the disk system needs to only replicate data since the last backup, thus drastically reducing the storage space requirement. Though RAID0.5 has the same data loss characteristics of traditional mirroring, the lower storage space comes at the cost of lower availability. Thus, RAID0.5 is a tradeoff between lower disk cost and lower availability while still preserving very high data reliability. We present analytical reliability models and experimental results that demonstrate the enhanced reliability and performance of the proposed RAID0.5 system.

[1]  Eran Gabber,et al.  Data logging: a method for efficient data updates in constantly active RAIDs , 1998, Proceedings 14th International Conference on Data Engineering.

[2]  Tzi-cker Chiueh,et al.  Track-based disk logging , 2002, Proceedings International Conference on Dependable Systems and Networks.

[3]  Masaru Kitsuregawa,et al.  Hot mirroring: a method of hiding parity update penalty and degradation during rebuilds for RAID5 , 1996, SIGMOD '96.

[4]  David A. Patterson,et al.  Virtual log based file systems for a programmable disk , 1999, OSDI '99.

[5]  Michelle Y. Kim,et al.  Synchronized Disk Interleaving , 1986, IEEE Transactions on Computers.

[6]  John Wilkes,et al.  A trace-driven analysis of disk working set sizes , 1993 .

[7]  Alexander A. Stepanov,et al.  Mime: a high performance parallel storage device with strong recovery guarantees , 1997 .

[8]  Tzi-cker Chiueh Trail: a track-based logging disk architecture for zero-overhead writes , 1993, Proceedings of 1993 IEEE International Conference on Computer Design ICCD'93.

[9]  Hai Jin,et al.  Parity Logging Overcoming the Small Write Problem in Redundant Disk Arrays , 2002 .

[10]  Yiming Hu,et al.  DCD—disk caching disk: a new approach for boosting I/O performance , 1996, ISCA '96.

[11]  Jai Menon,et al.  Floating Parity and Data Disk Arrays , 1993, J. Parallel Distributed Comput..

[12]  Hector Garcia-Molina,et al.  Disk striping , 1986, 1986 IEEE Second International Conference on Data Engineering.

[13]  Ming Zhang,et al.  Implementation and Performance Evaluation of RAPID-Chache under Linux , 2002, PDPTA.

[14]  Randy H. Katz,et al.  A case for redundant arrays of inexpensive disks (RAID) , 1988, SIGMOD '88.