Large Scale Multimedia Server Proc. Inet'94/jenc5 Design of a Large Scale Multimedia Storage Server I Introduction

Large scale multimedia storage servers will be an integral part of the emerging distributed multi-media computing infrastructure. However, given the modest rate of improvements in storage transfer rates, designing servers that meet the demands of multimedia applications is a challenging task that needs signiicant architectural innovation. Our research project, called Massively-parallel And Real-time Storage (mars) architecture, is aimed at the design and prototype implementation of a large scale multimedia storage server. It uses some of the well-known techniques in parallel I/O, such as data striping and Redundant Arrays of Inexpensive Disks (raid) and an innovative atm based interconnect inside the server to achieve a scalable architecture that transparently connects storage devices to an atm-based broad-band network. The atm interconnect within the server uses a custom asic called ATM Port Interconnect Controller (apic) currently being developed as a part of an arpa sponsored gigabit local atm testbed. Our architecture relies on innovative data striping and real-time scheduling to allow a large number of guaranteed concurrent accesses and uses separation of metadata from real data to achieve a direct ow of the media streams between the storage devices and the network. This paper presents our system architecture; one that is scalable in terms of the number of supported users and the throughput. Rapid advances in optical communication, high speed packet switching, data compression and processor and memory design will make a distributed multimedia computing infrastructure feasible in the near future. Multimedia mail, orchestrated presentations, high quality on-demand audio and video, collaborative multimedia document editing, distributed music rehearsals, browsing remote multimedia archives and virtual reality environments are a few examples of the exciting new applications that will be available to the users of such an infrastructure. An application in this environment will typically use one or more media streams, such as audio, graphics, video, images and text. Due to the data intensive nature of the component media streams, such applications will demand a large amount of compute, storage and network bandwidth. The user or the client equipment that runs these application may range from a simple display device such as a tv, to a sophisticated multime-dia workstation. A typical application may request data from a remote high bandwidth media storage/archival server on a periodic basis. The data may be rst passed on to a remote compute server for processing, such as content based real-time search on a video stream, and volume visu-alization. Subsequently, …