| R. Muntz, J. Santos, and S. Berson. RIO: A Realtime Multimedia Object Server. In ACM Sigmetrics Performance Evaluation Review, volume 25, September 1997. |
....are large cluster to cluster throughput, high I O performance delivered to the application, scalable disk bandwidth, good matching of disk and network throughput. The concept of cluster based server has already been proposed in connection with specific domains of applications such as video servers [13,14] and Internet data caches for large dataset acquisition [15] In this paper we focus on clusters employed as general purpose data servers in the context of high performance dataintensive computing. In our study, a (possibly parallel) application is running on a client cluster and we want to ....
R. Muntz, J.R Santos, and S. Berson. RIO: A real-time multimedia object server, ACM Performance Evaluation Review, 25(2), pp. 29-35, September 1997.
....to disks in a round robin manner starting with an arbitrarily chosen disk [11, 16, 1] For example, Figure 7 depicts a system with four disk drives where the assignment of X starts with disk do. Another way to distribute the system load across the disks is to employ a random placement of data [14, 20], see Figure 8. Figure 9.a shows the average startup latency of these two alternatives as a function of system load (utilization) Results from various configurations showed similar trends (see details in [8] In all cases, the average startup latency with round robin is higher than with random. ....
R. Muntz, J. Santos, and S. Berson. RIO: A Real-time Multimedia Object Server. ACM Sigmetrics Performance Evaluation Review, 25(2), Sep. 1997.
....modeled using our computational timing model. An example system is a set of avionics tasks that runs on an avionics application specific system [1] 45] The task set consists of 24 independent periodic tasks. Another example is a multimedia server that provides multimedia data to multiple users [37]. Without loss of generality, we assume that all tasks have identical periods. When this is not the case, a simple preprocessing step and application of the least common multiple (LCM) theorem [33] transforms an arbitrary set of periods to this design scenario in polynomial time. As a ....
R. Muntz, J. R. Santos, and S. Berson, "RIO: A real-time multimedia object server," Performance Eval. Rev., vol. 25, no. 2, pp. 29--35, 1997.
....the future direction of this work. 2. Related work Several studies have focused on data placement and retrieval scheduling of continuous media objects, for collections see [9, 13, 6, 10, 4, 5] Among these, only a few studied the random placement of data blocks on continuous media servers [3, 12, 11, 14]. Traditional constrained placement techniques such as round robin data placement allow for deterministic service guarantees while random placement techniques are modeled statistically. Overall, random placement increases the flexibility to support various applications while it maintains a ....
R. Muntz, J. Santos, and S. Berson. RIO: A Real-time Multimedia Object Server. In ACM Sigmetrics Performance Evaluation Review, volume 25, September 1997.
....cycle length V. Related works A lot of techniques have been investigated in order to achieve scalability for video servers. Those techniques have been implemented within video server prototypes. For example, studies have been conducted by Biersack et al. Server Array [BB96] Muntz et al. RIO [MSB97] Ghandeharizadeh et al. Mitra [GZS 97] Lee and Wong [LW00] A comprehensive study of architectural alternatives and the approaches employed by existing systems can be found in [GVK 95] Lee98] The main di erence between the mentioned studies and the architecture studied in this ....
R. Muntz, J.R Santos, and S. Berson. RIO: A realtime multimedia object server. ACM Performance Evaluation Review, 25(2):29-35, September 1997.
....natural for a video server, given the intrinsic parallelism provided by independent clients on one hand, and the possibility of easily fetching in parallel multiple blocks for a given stream on the other hand. There are several prototypes design for distributed architecture : RIO done at UCLA[MSB97] MITRA developed at USC[GZS 97] Server Array done at Eurecom [BB96a] and Tiger the prototype of Microsoft [BB 96b] etc. Within the set of video server prototypes, a lot of research have been done concerning disk optimizations (scheduling, data placement) however, almost none of them ....
R. Muntz, J.R Santos, and S. Berson. RIO : A real-time multimedia object server. ACM Performance Evaluation Review, 25(2):29-35, September 1997.
.... To store a large media object X on such a platform it is commonly striped into n blocks: X 0 , X 1 , X n 1 [Pol91] There are two basic techniques to assign these blocks to the magnetic disk drives that form the storage system: a) in a round robin sequence [SGM86] or (b) in a random manner [MSB97]. Traditionally, the round robin placement utilizes a cycle based approach [GZS 97] to scheduling of resources to guarantee a continuous display, while the random placement utilizes a deadline driven approach [RW94] We have previously implemented paradigm (a) and now plan to extend our prototype ....
Muntz, R.R.; Santos, J.; and Berson, S. "RIO: A Real-time Multimedia Object Server." ACM Sigmetrics Performance Evaluation Review, Volume 25, Number 2, September, 1997.
.... media object X on such a platform it is commonly striped into n blocks: X 0 ; X 1 ; Xn 1 [14, 21, 7, 3, 12] There are two basic techniques to assign these blocks to the magnetic disk drives that form the storage system: a) in a round robin sequence [15, 3] or (b) in a random manner [11, 4, 16]. Traditionally, the round robin placement utilizes a cyclebased approach to scheduling of resources to guarantee a continuous display, while the random placement utilizes a deadline driven approach. We plan to implement the random block placement in conjunction with a deadline driven 4 1000 ....
R. Muntz, J. Santos, and S. Berson. RIO: A Real-time Multimedia Object Server. In ACM Sigmetrics Performance Evaluation Review, volume 25, September 1997.
....which focus on di erent design issues: Fellini: is developped by AT T[MNz 96] and is designed to handle a verywide range of requirements, from non realtime data to video streams. It is target for a SMP box and not for a cluster as the other works mentioned below. RIO: done at UCLA[MSB97] is based on random allocation of stripping nodes, which, as Fellini, makes it possible to support a wide range of multimedia needs, but which is now targeted for a cluster of PC (the exibility of random allocation allowed this move from a SMP solution) MITRA: done at USC[GZS 97] is ....
R. Muntz, J.R Santos, and S. Berson. RIO : A real-time multimedia object server. ACM Performance Evaluation Review, 25(2):29-35, September 1997.
....and implement a virtual walk through system such that a user can query and retrieve information about the virtual world. One major limitation of the work is that it only allows a single user to explore the virtual world 1 and therefore, there is no communication and interaction between users. In [9, 10], the authors described how to build a storage system that can support applications like the video on demand and 3D walk through of a virtual world. The result is particularly interesting in the sense that the storage server can guarantees the timely delivery of data to the different multimedia ....
R. Muntz, J.R. Santos, and S. Berson. RIO: A Real-time Multimedia Object Server, ACM Performance Evaluation Review, ACM Press, vol. 25, no. 2, p.29-35, September, 1997.
....and implement a virtual walk through system such that a user can query and retrieve information about the virtual world. One major limitation of the work is that it only allows a single user to explore the virtual world 2 and therefore, there is no communication and interaction between users. In [9, 10], the authors described how to build a storage system that can support applications like the video ondemand and 3D walk through of a virtual world. The result is particularly interesting in the sense that 2 Of course, the system allows many users to explore the same virtual world but under ....
R. Muntz, J.R. Santos, and S. Berson. RIO: A Real-time Multimedia Object Server, ACM Performance Evaluation Review, ACM Press, vol. 25, no. 2, p.29-35, September, 1997.
No context found.
R. Muntz, J. Santos, and S. Berson. RIO: A Realtime Multimedia Object Server. In ACM Sigmetrics Performance Evaluation Review, volume 25, September 1997.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC