A case for redundant arrays of inexpensive disks (RAID)
SIGMOD '88 Proceedings of the 1988 ACM SIGMOD international conference on Management of data
Streaming RAID: a disk array management system for video files
MULTIMEDIA '93 Proceedings of the first ACM international conference on Multimedia
A survey of approaches to fault tolerant design of VOD servers: techniques, analysis and comparison
Parallel Computing - Special issues on applications: parallel data servers and applications
Failure recovery algorithms for multimedia servers
Multimedia Systems
Performance Analysis of a Pull-Based Parallel Video Server
IEEE Transactions on Parallel and Distributed Systems
Data striping and reliability aspects in distributed video servers
Cluster Computing
Mitra: A Scalable Continuous Media Server
Multimedia Tools and Applications
Parallel Video Servers: A Tutorial
IEEE MultiMedia
An Implicitly Scalable, Fully Interactive Multimedia Storage Server
DIS-RT '98 Proceedings of the Second International Workshop on Distributed Interactive Simulation and Real-Time Applications
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
Performance and Reliability Study for Distributed Video Servers: Mirroring or Parity?
ICMCS '99 Proceedings of the 1999 IEEE International Conference on Multimedia Computing and Systems - Volume 02
Data consistent up- and downstreaming in a distributed storage system
SNAPI '03 Proceedings of the international workshop on Storage network architecture and parallel I/Os
Efficient allocation and composition of distributed storage
The Journal of Supercomputing
Hi-index | 0.00 |
We describe the design and the implementation of the CFS (Cluster File System) storage system which is dedicated to video streams. Our goal is to provide a system with the following features: 1) High number of supported streams at a low cost. 2) Transparent management with respect to the clients. 3) Reliability regarding data storage and service continuity. The CFS is implemented on a cluster of PCs connected with a high speed internal network. Its management is fully distributed among the cluster nodes so that there is no central component. Data are stored and retrieved among the cluster nodes using a Streaming RAID strategy, to enhance reliability. We evaluate the overhead introduced by the CFS management with respect to disk performance, by varying the number of nodes involved in file storage, the block size and the reliability level. The impact of the global server load on this overhead is moreover discussed to demonstrate scalability.