Data Orgnization

1
Data Orgnization

• Frequently accessed data on the same storage
  device?
• Conversely, all infrequently accessed data on the
  same storage device?
• Retrieval efficiency through proper data
  placement
• Load balancing and availability issues

2
Storage Manager

• Allocation and storage of multimedia files on
  various storage devices (maintain metadata that
  maps logical blocks of files to physical blocks)
• Enforce bandwidth reservations on storage devices
  to ensure that devices are not overloaded
• Recover from hardware component failures
• Coordinate data storage on several hierarchies

3
Frames to Storage Blocks

• A one-to-one mapping between frames and blocks is
  inefficient
• Partitioning a multimedia file (viewed as a
  stream of bytes) into fixed-size storage blocks
• Extracting the frames from the storage blocks is
  left to the decoding software
• Convert frame rate to average block delivery rate

4
Data Block Placement

• Retrieval Blocks
• Contiguous Placement
• a single seek is enough (but fragmentation is a
  problem)
• Scattered Placement
• many intrafile seeks (can be alleviated using
  large block sizes and to read one block in each
  round)
• Constrained Placement
• intrafile seeks may be unavoidable if multiple
  blocks are needed in each round
• reduce the seeks to a reasonable bound by
  bounding the separation of successive blocks

5
Data Striping

• RAID Scheme -- data is striped across each disk
• achieves intrafile and interfile parallelism
• Disk arrays are a good solution to the high
  bandwidth requirements of multimedia services

6
RAID

• Data striping for improved peformance
• Redundancy for improved reliability
• Data striping distributes data over multiple
  disks to make them appear as a single, fast,
  large disk
• Multiple I/Os in parallel
• multiple requests on separate disks (decreases
  queueing time seen by I/O requests)
• single multiple block request on multiple disks
  (increases the effective transfer rate seen by a
  single request)