Disk Scheduling

• Transferring data to/from secondary memory is the worst bottleneck of a general computing system
• As a computer deals with multiple processes over time, request to access the disk accumulate
• Operating systems use disk scheduling to determine which requests to complete first
• Scheduling algorithms primarily try to solve issues with seek time

First-Come, First-Served

• Similar to the CPU scheduling algorithm
• One of the easiest to implement but not the most efficient
• Process requests in order they arrive
  • Disregard the current position of the heads
• Head often moves over cylinders whose requests are pending

Shortest-Seek-Time-First

• Moves heads by the minimum amount necessary to satisfy a request
• Could result in the heads changing directions after each request is satisfied
• Does not guarantee smallest overall head movement
• Improvement over performance of FCFS
• It's possible an early request never gets processed
  • New requests keep getting added close to where the head is
  • New requests are closer --> higher priority
  • Called starvation

SCAN

• Read/write heads move in toward the spindle then out toward the platter edge
  • Back and forth between spindle and platter edge
• Satisfy requests along the way
• No chance of starvation