Symmetrical declustering: a load balancing and fault tolerant strategy for clustered video servers

Authors:
Song Wu;Hai Jin;Guang Tan
Affiliations:
Huazhong University of Sci. & Tech., Wuhan, China;Huazhong University of Sci. & Tech., Wuhan, China;Huazhong University of Sci. & Tech., Wuhan, China
Venue:
ICCSA'03 Proceedings of the 2003 international conference on Computational science and its applications: PartI
Year:
2003

Citing 10
Cited 0

Long-term movie popularity models in video-on-demand systems: or the life of an on-demand movie

MULTIMEDIA '97 Proceedings of the fifth ACM international conference on Multimedia
Comparing random data allocation and data striping in multimedia servers

Proceedings of the 2000 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
Efficient striping techniques for variable bit rate continuous media file servers

Performance Evaluation
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
Scalable Multimedia Servers

IEEE Concurrency
Segmented Information Dispersal (SID) Data Layouts for Digital Video Servers

IEEE Transactions on Knowledge and Data Engineering
Symmetrical Pair Scheme: A Load Balancing Strategy to Solve Intra-Movie Skewness for Parallel Video Servers

IPDPS '02 Proceedings of the 16th International Parallel and Distributed Processing Symposium
Load management in distributed video servers

ICDCS '97 Proceedings of the 17th International Conference on Distributed Computing Systems (ICDCS '97)
A Popularity-Based Data Allocation Scheme for a VOD Server

COMPSAC '96 Proceedings of the 20th Conference on Computer Software and Applications

Quantified Score

Hi-index	0.00

Visualization

Abstract

Data placement strategy is important for modeling and designing large-scale video servers. To optimize the overall system performance, numerous schemes of data layout are proposed to obtain availability and load balancing. Clustered video servers stripe video objects across multiple nodes in order to achieve server-level fault tolerance and solve the load imbalance caused by video popularity. However, due to clients' skew access pattern, the clustered video servers with traditional fault-tolerance replication schemes probably suffer performance decline in both normal and fault mode. In this paper, we propose a novel algorithm called Symmetrical Declustering to improve system performance and availability. Our analysis shows the effectiveness of the scheme.