Supporting continuous media (CM) data-such as video and audio-imposes strin
gent demands on the retrieval performance of a multimedia server. In this p
aper, we propose and evaluate a set of data-placement and retrieval algorit
hms to exploit the full capacity of the disks in a multimedia server. The d
ata-placement algorithm declusters every object over all of the disks in th
e server-using a time-based declustering unit-with the aim of balancing the
disk load. As for runtime retrieval, the quintessence of the algorithm is
to give each disk advance notification of the blocks that have to be fetche
d in the impending time periods, so that the disk can optimize its service
schedule accordingly. Moreover, in processing a block request for a replica
ted object, the server will dynamically channel the retrieval operation to
the most lightly loaded disk that holds a copy of the required block. We ha
ve implemented a multimedia server based on these algorithms. Performance t
ests reveal that the server achieves very high disk efficiency. Specificall
y, each disk is able to support up to 25 MPEG-1 streams. Moreover, experime
nts suggest that the aggregate retrieval capacity of the server scales almo
st linearly with the number of disks.