Chapter 11 (I/O Management and Disk Scheduling)

Question 1

What is programmed I/O?

Answer 1

CPU issues I/O and busy waits for it to complete

Question 2

What is interrupt-driven I/O?

Answer 2

CPU issues I/O and continues, is interrupted when I/O completes

Question 3

What is DMA?

Answer 3

CPU issues I/O and DMA module takes over, freeing the CPU

Question 4

What information is contained in a call to the DMA module?

Answer 4

Whether to read/write

Address of the I/O device

Starting memory address to read or write to

Number of words to read or write

Question 5

What are the two OS design objectives for I/O?

Answer 5

Efficiency

Generality

Question 6

What is generality?

Answer 6

Have all devices use the same interface

Question 7

What are two types of I/O devices?

Answer 7

Block-oriented

Stream-oriented

Question 8

What are block-oriented I/O devices?

Answer 8

Information stored in fixed size blocks

Transfers made a block at a time

Question 9

What are stream-oriented I/O devices?

Answer 9

Transfers information as a stream of bytes (no block structure)

Question 10

What are the advantages of a single buffer?

Answer 10

Allows next unit to be read while previous is processed

Process can be swapped since I/O is using OS buffer

Question 11

When might a double buffer be inadequate?

Answer 11

Process performs rapid bursts of I/O

Question 12

What can be used if a process performs rapid bursts of I/O?

Answer 12

Circular buffering

Question 13

Define: Access time

Answer 13

Sum of seek time and rotational delay

Question 14

Define: Seek time

Answer 14

Time it takes to position the head at the desired track

Question 15

Define: Rotational delay

Answer 15

Time it takes for the beginning of the sector to reach the head

Question 16

Define: Data transfer time

Answer 16

Occurs as the sector moves under the head

Question 17

What is the primary reason for differences in performance in disk scheduling policies?

Answer 17

Seek time

Question 18

What are the three characteristics of RAID?

Answer 18

Multiple physical drives viewed by the OS as a single drive

Data distributed across the drives

Redundancy to ensure recoverability if a drive fails

Question 19

What RAID levels are commercially unavailable?

Answer 19

2 and 4

Question 20

What are the characteristics of RAID level 1?

Answer 20

Duplicates all data

Every disk has a mirror disk

Simple recovery but higher cost

Question 21

Why is RAID level 2 not used?

Answer 21

Only useful if lots of disk errors occur, which is not common today

Question 22

How does RAID level 2 handle recovery?

Answer 22

Parity disk

Question 23

What are the characteristics of RAID level 3?

Answer 23

Single additional parity disk

Uses a parity bit instead of an error correcting code

Question 24

What are strips in RAID level 4?

Answer 24

Blocks

Question 25

What are the characteristics of RAID level 5?

Answer 25

Distributes parity disks across all disks

Keeps parity disk from being a bottleneck

Question 26

What are the characteristics of RAID level 6?

Answer 26

Uses two parity strips on different disks instead of just 1

Requires N+2 disks

Question 27

How many disks must fail in RAID 6 to be unavailable?

Answer 27

3

Question 28

What are the replacement algorithms for the disk cache?

Answer 28

LRU - Least recently used

LFU - Least frequently used

Question 29

What’s the problem with LFU disk cache replacement algorithms?

Answer 29

Blocks may be used repeatedly for a short amount of time, building up a high reference count, then not used again for a long time. This can cause LFU to make a poor choice.

Question 30

How can you fix the problem with LFU?

Answer 30

Divide the queue into new, middle, and old sections