A. Mourad. I/o scheduling in a storage server for video-on-demand applications. In Proceeding of the IASTED/ISMM International Conference. Distributed Multimedia Systems and Applications, page 31, Honolulu, HI, Aug 1994. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
The Design, Implementation, and Analysis of the Stony Brook Video .. - Vernick (1996) (Correct)
....Schemes for striping data using course grain striping and fine grain striping are examined in [ORS96] For each scheme, the optimal amount of data to be retrieved for a video during each disk access is computed so that the number of concurrent streams supported by the server is maximized. Mou94] analyzes the performance of using a RAID 3 array for delivery of video services. In [PBC95] the authors compare two media block placement schemes where the size of a block is a group of frames (CTL) In the balanced placement approach, each media block is striped across all disks in the array ....
....CHAPTER 7. VIDEO STORAGE SYSTEM ADMISSION CONTROL 88 Compared to the Deterministic Average algorithms, the Statistical Worst Case is about 10 to 20 better, across different I O cycle times and disk array configurations. This result is less optimistic than are reported in earlier papers [Mou94] VGG94b, VGG94a] This is because the I O service time model presented in the previous section is more accurate and realistic. In addition, the measurements from real implementations include various overheads, such as SCSI bus arbitration overheads, that are simply overlooked in earlier studies. ....
[Article contains additional citation context not shown here]
A. Mourad. I/o scheduling in a storage server for video-on-demand applications. In Proceeding of the IASTED/ISMM International Conference. Distributed Multimedia Systems and Applications, page 31, Honolulu, HI, Aug 1994.
An Empirical Study of Admission Control Strategies in Video.. - Vernick, Chiueh (1995) (1 citation) (Correct)
....deterministic approach: Worst case seek delay and rotational latency are used to decide if new video streams should be admitted [AOG92, RVR92, RV91, VR93] In each of these papers, the admission control algorithms for single disk I O systems are presented without concrete performance results. In [Mou94], a deterministic algorithm is compared to a statistical algorithm. Simulations show that the statistical algorithm can service up to 70 more streams than the deterministic algorithm while reducing the overflow probability to the order of 10 Gamma7 . This work, however, did not describe the ....
....overloaded I O cycles, we will only focus on the former. Compared to the Deterministic Average algorithms, the Statistical Worst Case is about 10 to 20 better, across different I O cycle times and disk array configurations. This result is less optimistic than are reported in earlier papers [Mou94] [VGG94b, VGG94a] We believe this is because the I O service time model presented in Section 2 is more accurate and realistic. In addition, the measurements from real implementations include various overheads such as SCSI bus arbitration overheads that are simply overlooked in earlier studies. ....
A. Mourad. I/o scheduling in a storage server for video-on-demand applications. In Proceeding of the IASTED/ISMM International Conference. Distributed Multimedia Systems and Applications, page 31, Honolulu, HI, Aug 1994.
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