A Database Approach for Accelerate Video Data Access

In this paper, we studied the efficiency and break-event point of storing video objects into DBMS and proved that storing "small" video objects into database is a suitable solution. To the video objects that stored in database as BLOB data type, we devised a database based time-oriented approach to speed up the video content access. Our experiments showed that, because of we extracted some system-aware metadata and stored into database transparently, the read performance was become practicable.

[1]  Alexandros Biliris.  The performance of three database storage structures for managing large objects , 1992, SIGMOD '92.

[2]  Michael Stonebraker,et al.  Large object support in POSTGRES , 1993, Proceedings of IEEE 9th International Conference on Data Engineering.

[3]  Margo I. Seltzer,et al.  The case for application-specific benchmarking , 1999, Proceedings of the Seventh Workshop on Hot Topics in Operating Systems.

[4]  Catharine van Ingen,et al.  Fragmentation in Large Object Repositories , 2006, CIDR.

[5]  Howard Gobioff,et al.  The Google file system , 2003, SOSP '03.

[6]  Wilson C. Hsieh,et al.  Bigtable: A Distributed Storage System for Structured Data , 2006, TOCS.

[7]  David J. DeWitt,et al.  Object and File Management in the EXODUS Extensible Database System , 1986, VLDB.

[8]  William H. Tetzlaff,et al.  Multimedia - An introduction , 1998, IBM J. Res. Dev..

[9]  C. Mohan,et al.  Coordinating backup/recovery and data consistency between database and file systems , 2002, SIGMOD '02.