A case for redundant arrays of inexpensive disks (RAID)
SIGMOD '88 Proceedings of the 1988 ACM SIGMOD international conference on Management of data
Computer architecture: a quantitative approach
Computer architecture: a quantitative approach
Streaming RAID: a disk array management system for video files
MULTIMEDIA '93 Proceedings of the first ACM international conference on Multimedia
On-line extraction of SCSI disk drive parameters
Proceedings of the 1995 ACM SIGMETRICS joint international conference on Measurement and modeling of computer systems
Fault tolerant design of multimedia servers
SIGMOD '95 Proceedings of the 1995 ACM SIGMOD international conference on Management of data
Fault-tolerant architectures for continuous media servers
SIGMOD '96 Proceedings of the 1996 ACM SIGMOD international conference on Management of data
Distributed schedule management in the Tiger video fileserver
Proceedings of the sixteenth ACM symposium on Operating systems principles
Data striping and reliability aspects in distributed video servers
Cluster Computing
Chained Declustering: A New Availability Strategy for Multiprocessor Database Machines
Proceedings of the Sixth International Conference on Data Engineering
Disk striping in video server environments
ICMCS '96 Proceedings of the 1996 International Conference on Multimedia Computing and Systems
Design and Performance Tradeoffs in Clustered Video Servers
ICMCS '96 Proceedings of the 1996 International Conference on Multimedia Computing and Systems
| Hi-index | 0.00 |
Mirroring-based reliability as compared to parity-based reliability significantly simplifies the design and the implementation of video servers, since in case of failure mirroring does not require any synchronization of reads or decoding to reconstruct the lost video data. While mirroring doubles the amount of storage volume required, the steep decrease of the cost of magnetic disk storage makes it more and more attractive as a reliability mechanism. We present in this paper a novel data layout strategy for replicated data on a video server. In contrast to classical replica placement schemes that store original and replicated data separately, our approach stores replicated data adjacent to original data and thus does not require additional seek overhead when operating with disk failure. We show that our approach considerably improves the server performance compared to classical replica placement schemes such as the interleaved declustering scheme and the scheme used by the Microsoft Tiger video server. Our performance metric is the maximum number of users that a video server can simultaneously support (server throughput).