Fine-Grained Device Management in an Interactive Media Server Raju Rangaswami, Zoran Dimitrijevic, Edward Chang, and Shueng-Han Gary Chan IEEE Trans. on.

Slides:



Advertisements
Similar presentations
Continuous Media 1 Differs significantly from textual and numeric data because of two fundamental characteristics: –Real-time storage and retrieval –High.
Advertisements

Ying Wai Wong, Jack Y. B. Lee, Victor O. K. Li, and Gary S. H. Chan CSVT 2007 FEB Supporting Interactive Video-on-Demand With Adaptive Multicast Streaming.
Chapter 20: Multimedia Systems Silberschatz, Galvin and Gagne ©2005 Operating System Concepts Chapter 20: Multimedia Systems What is Multimedia.
Silberschatz, Galvin and Gagne ©2009 Operating System Concepts – 8 th Edition Chapter 20: Multimedia Systems.
Allocating Memory.
Allocation Methods - Contiguous
1 Multimedia Chapter Introduction to multimedia 7.2 Multimedia files 7.3 Video compression 7.4 Multimedia process scheduling 7.5 Multimedia file.
1 Adaptive Live Broadcasting for Highly-Demanded Videos Hung-Chang Yang, Hsiang-Fu Yu and Li-Ming Tseng IEEE International Conference on Parallel and Distributed.
Lecture 17 I/O Optimization. Disk Organization Tracks: concentric rings around disk surface Sectors: arc of track, minimum unit of transfer Cylinder:
VCR-oriented Video Broadcasting for Near Video-On- Demand Services Jin B. Kwon and Heon Y. Yeon Appears in IEEE Transactions on Consumer Electronics, vol.
Prefix Caching assisted Periodic Broadcast for Streaming Popular Videos Yang Guo, Subhabrata Sen, and Don Towsley.
Wavelet-Based VBR Video Traffic Smoothing Dejian Ye, J. Cam Barker, Zixiang Xiong, and Wenwu Zhu IEEE Trans. Multimedia, Aug
A Novel Video Layout Strategy for Near-Video-on- Demand Servers Shenze Chen & Manu Thapar Hewlett-Packard Labs 1501 Page Mill Rd. Palo Alto, CA
Cost-Based Plan Selection Choosing an Order for Joins Chapter 16.5 and16.6 by:- Vikas Vittal Rao ID: 124/227 Chiu Luk ID: 210.
Efficient Support for Interactive Browsing Operations in Clustered CBR Video Servers IEEE Transactions on Multimedia, Vol. 4, No.1, March 2002 Min-You.
Bandwidth Allocation in a Self-Managing Multimedia File Server Vijay Sundaram and Prashant Shenoy Department of Computer Science University of Massachusetts.
Adaptive Streaming and Rendering of Large Terrains: a Generic Solution WSCG 2009 Raphaël Lerbour Jean-Eudes Marvie Pascal Gautron THOMSON R&D, Rennes,
Design of an Interactive Video- on-Demand System Yiu-Wing Leung, Senior Member, IEEE, and Tony K. C. Chan IEEE Transactions on multimedia March 2003.
1 Exploring Data Reliability Tradeoffs in Replicated Storage Systems NetSysLab The University of British Columbia Abdullah Gharaibeh Matei Ripeanu.
File Management Chapter 12.
CS Spring 2012 CS 414 – Multimedia Systems Design Lecture 34 – Media Server (Part 3) Klara Nahrstedt Spring 2012.
Secondary Storage Management Hank Levy. 8/7/20152 Secondary Storage • Secondary Storage is usually: –anything outside of “primary memory” –storage that.
MM File Management Karrie Karahlaios and Brian P. Bailey Spring 2007.
Memory Management ◦ Operating Systems ◦ CS550. Paging and Segmentation  Non-contiguous memory allocation  Fragmentation is a serious problem with contiguous.
File Management Chapter 12. File Management File management system is considered part of the operating system Input to applications is by means of a file.
IT The Relational DBMS Section 06. Relational Database Theory Physical Database Design.
File Implementation. File System Abstraction How to Organize Files on Disk Goals: –Maximize sequential performance –Easy random access to file –Easy.
1 Multimedia Storage Issues. NUS.SOC.CS5248 OOI WEI TSANG 2 Media vs. Documents large file size write once, read many deadlines!
Storage Management in Virtualized Cloud Environments Sankaran Sivathanu, Ling Liu, Mei Yiduo and Xing Pu Student Workshop on Frontiers of Cloud Computing,
XE33OSA Chapter 20: Multimedia Systems. 20.2XE33OSA Silberschatz, Galvin and Gagne ©2005 Chapter 20: Multimedia Systems What is Multimedia Compression.
Budget-based Control for Interactive Services with Partial Execution 1 Yuxiong He, Zihao Ye, Qiang Fu, Sameh Elnikety Microsoft Research.
Silberschatz, Galvin and Gagne ©2009 Operating System Concepts – 8 th Edition, Multimedia Systems.
Subject: Operating System.
CSCI-375 Operating Systems Lecture Note: Many slides and/or pictures in the following are adapted from: slides ©2005 Silberschatz, Galvin, and Gagne Some.
File Management Chapter 12. File Management File management system is considered part of the operating system Input to applications is by means of a file.
Chapter 4 Memory Management Virtual Memory.
Lecture 3 Page 1 CS 111 Online Disk Drives An especially important and complex form of I/O device Still the primary method of providing stable storage.
CS Spring 2009 CS 414 – Multimedia Systems Design Lecture 30 – Media Server (Part 5) Klara Nahrstedt Spring 2009.
Disk & File System Management Disk Allocation Free Space Management Directory Structure Naming Disk Scheduling Protection CSE 331 Operating Systems Design.
Improving Disk Throughput in Data-Intensive Servers Enrique V. Carrera and Ricardo Bianchini Department of Computer Science Rutgers University.
Virtual Memory The memory space of a process is normally divided into blocks that are either pages or segments. Virtual memory management takes.
Storing and Serving Multimedia. What is a Media Server? A scalable storage manager Allocates multimedia data optimally among disk resources Performs memory.
Data dissemination in wireless computing environments
Content caching and scheduling in wireless networks with elastic and inelastic traffic Group-VI 09CS CS CS30020 Performance Modelling in Computer.
NUS.SOC.CS5248 Ooi Wei Tsang 1 Course Matters. NUS.SOC.CS5248 Ooi Wei Tsang 2 Make-Up Lecture This Saturday, 23 October TR7, 1-3pm Topic: “CPU scheduling”
Disk storage systems Question#1 (True/False) A track is divided into multiple units called sectors.
Operating Systems: Summary INF1060: Introduction to Operating Systems and Data Communication.
Saving Bitrate vs. Users: Where is the Break-Even Point in Mobile Video Quality? ACM MM’11 Presenter: Piggy Date:
Minimum Cost Scheduling of Stored Video in Dynamic Bandwidth Allocation Networks Reporter : M 張益瑞 IEEE Transactions on Consumer Electronics, Vol.
CS Spring 2009 CS 414 – Multimedia Systems Design Lecture 27 – Media Server (Part 2) Klara Nahrstedt Spring 2009.
Chapter 20: Multimedia Systems
Practical Database Design and Tuning
Memory Management.
CS 414 – Multimedia Systems Design Lecture 31 – Media Server (Part 5)
Chapter 20: Multimedia Systems
Chapter 11: File System Implementation
The Impact of Replacement Granularity on Video Caching
Operating Systems Disk Scheduling A. Frank - P. Weisberg.
Chapter 9 – Real Memory Organization and Management
Chapter 20: Multimedia Systems
HashKV: Enabling Efficient Updates in KV Storage via Hashing
Practical Database Design and Tuning
Data Orgnization Frequently accessed data on the same storage device?
Secondary Storage Management Brian Bershad
Secondary Storage Management Hank Levy
Chapter 20: Multimedia Systems
Performance-Robust Parallel I/O
Chapter 20: Multimedia Systems
Chapter 20: Multimedia Systems
Presentation transcript:

Fine-Grained Device Management in an Interactive Media Server Raju Rangaswami, Zoran Dimitrijevic, Edward Chang, and Shueng-Han Gary Chan IEEE Trans. on Multimedia, Dec 2003

Outline Introduction Interactive media proxy Device profiling Device management High-level data organization Low-level data organization IO scheduling System evaluation

Introduction Interactive media Fast-forward Interactive media proxy (IMP) Transform non-interactive broadcast or multicast streams into interactive ones for servicing a large number of end users.

Interactive Media Proxy (IMP) Device profiling Collect detailed disk parameters to manage a device more effectively. Device management Perform fine-grained device management to improve the overall disk access efficiency.

Disk profiling The authors present a SCSI disk profiling tool that extracts detailed disk parameter. Why disk profiling is necessary? Inaccurate information (worst case assumption) Dynamic information (ex: file fragmentation) Manufacturing variance

Device management High-level data organization Low-level disk placement IO scheduling

High-level data organization For fast-scan Skip B frames Display a P frame only if the corresponding I frame is also included. Adaptive tree scheme Use a truncated binary tree to store videos. Each level of the tree forms a substream and is stored as a sequential file.

Truncated binary tree original I + P sampled I

Adaptive tree scheme Height (h) The number of levels The number of supported fast-scan streams Density (η) Range from 0 to 1 The smaller η eliminates some tree level and decreases the tree density.

Low-level disk placement (1/2) Zoning placement Zone – multiple cylinders Combine similar bit-rate streams in the same logical zone. Outer zones have higher data-transfer rate. High bit-rate streams should be stored in fast zones. (to maximize throughput) Cylinder placement Exploit the deterministic nature of write streams and use a best-effort approach for reads.

Low-level disk placement (2/2) When any write stream uses up its allocated cylinders, a new set of free cylinders within the same zone and adjacent to the previous cylinder set is allocated. Cylinder placement maintains the same relative cylinder distance between the stream pairs. Minimize IO variability. The seek overhead for switching from one write stream to the next write stream requires the disk to seek typically less than 50 cylinders. (almost equal to the minimum seek time for a single cylinder) S1S1 S2S2 S3S3 S1S1 S2S2 write …

IO scheduling Goals Maximize throughput Minimize response time Step-sweep IO scheduling Using Cylinder Placement, the seek overheads for write streams can be minimized. Thus, step-sweep schedules write streams optimally.

Step-sweep IO scheduling

System evaluation Truncated Binary Tree (TBT) η = 1 Partial TBT (PTBT) η = 0.5 Sequential (SEQ) η = 1/h (original video stream) Reduce seek overhead for writes. Suffer from fast-scan.

SEQ v.s. PTBT v.s. TBT SEQ reduces seek operations

Zoning Placement Zoning placement improves throughput for read-intensive loads.

Cylinder Placement

Step-Sweep

Cumulative Effect