In this paper, the design issues on a scaleable popularity independent vide
o on demand (VOD) server which has been designed and implemented in Ajou Un
iversity are discussed. The server with parallel architecture aims to resol
ve a basic performance bottleneck, which has been encountered in the design
steps of many other VOD servers. The performance degradation caused by the
skew in video popularity becomes minimized through the realization of serv
er-level striping. The server-level striping increases the scalability of s
erver: the increase in performance follows closely the increase in number o
f video pumps. The communication and processing power overhead introduced b
y the parallel server architecture is relieved by a specially designed mesh
ed pipeline type interconnection network. The design issues on the intercon
nection network and data placement policy are also addressed with empirical
data. In the last part of this paper, the real-time operating system and t
he schedulers for video pump are described.