Deck 12: Disk Storage and Raid Technology

Question 1

Magnetic storage became popular because it was ________.

A) fast, inexpensive and persistent
B) rewritable, inexpensive and persistent
C) rewritable, fast and persistent
D) inexpensive, fast and reliable

Answer 1

rewritable, inexpensive and persistent

Question 2

Recently, disk capacities have ________, access times have ________ and cost per unit storage has ________.

A) increased, decreased exponentially, decreased exponentially
B) increased asymptotically, decreased, stayed the same
C) decreased, decreased, decreased
D) none of the above

Answer 2

none of the above

Question 3

________ is the amount of time required for the head to move from its current cylinder to its destination.

A) Rotational latency time
B) Transmission time
C) Seek time
D) none of the above

Answer 3

Seek time

Question 4

Moving-head disks store data in ________, which are part of larger units called _______, which form ________ when considered vertically.

A) tracks, sectors, cylinders
B) sectors, cylinders, tracks
C) cylinders, tracks, sectors
D) sectors, tracks, cylinders

Question 5

Disk scheduling algorithms are designed to minimize ________ and/or ________, depending on the system's requests.

A) seek time, rotational latency
B) seek latency, rotational time
C) request rate, turnaround time
D) none of the above

Question 6

Scheduling strategies are commonly evaluated by the system's ________.

A) throughput
B) mean response time
C) variance of response times
D) all of the above

Question 7

Maximizing ________and minimizing ________can lead to increased ________.

A) predictability, mean response time, rotational latency
B) throughput, mean response time, variance of response times
C) throughput, variance of response times, transmission time
D) none of the above

Question 8

FCFS scheduling tends to result in an acceptable amount of ________ at the expense of ________.

A) throughput, fairness
B) variance of response times, fairness
C) variance of response times, throughput
D) fairness, variance of response times

Question 9

FCFS scheduling does not exhibit ________.

A) high throughput
B) indefinite postponement of requests
C) low variance of response times
D) both a and b

Question 10

Because SSTF scheduling exhibits high ________, it is inappropriate for real-time and interactive systems.

A) variance of response times
B) throughput
C) mean response time
D) both a and b

Question 11

SCAN disk scheduling considers ________ in addition to seek times.

A) rotational latency
B) direction
C) variance of response times
D) transmission time

Question 12

Which if the following statements is false?

A) SCAN offers higher throughput than SSTF.
B) SCAN can suffer from indefinite postponement of requests.
C) SCAN is often called the elevator algorithm.
D) Both a and b.

Question 13

Under C-SCAN scheduling, which the disk arm finishes sweeping from the outermost track to the innermost track, it ________.

A) reverses direction
B) skips to the mid-range tracks
C) jumps back to the outermost track
D) services the next n requests using scan.

Question 14

The primary advantage of F-SCAN scheduling over other SCAN strategies is ________.

A) improved seek time
B) reduced variance in response times
C) avoidance of indefinite postponement
D) both b and c

Question 15

N-Step SCAN differs from F-SCAN in that ________.

A) N-Step SCAN prevents indefinite postponement, but F-SCAN does not.
B) N-Step SCAN always provides higher throughput than F-SCAN, but results in a larger variance in response times.
C) N-Step SCAN selects requests it will service during the current sweep according to the number of requests; F-SCAN selects such requests according to time.
D) None of the above-N-Step SCAN and F-SCAN are identical.

Question 16

The LOOK algorithm is based on the ________ strategy, but it eliminates unnecessary seeks when there are no more requests in the current direction.

A) SCAN
B) SSTF
C) FCFS
D) F-SCAN

Question 17

The C-LOOK algorithm reduces ________ compared to LOOK at the expense of ________.

A) variance of response times, fairness
B) variance of response times, throughput
C) mean response time, throughput
D) throughput, variance of response times

Question 18

Today's hard disks exhibit seek times that are ________ rotational latency times.

A) much greater than
B) greater than
C) the same order of magnitude as
D) much smaller than

Question 19

In SLTF scheduling, requests are ordered by ________, with the ________ one being serviced first.

A) track, nearest
B) size, largest
C) sector, nearest
D) size, smallest

Question 20

Both SPTF and SATF scheduling can lead to ________.

A) low throughput
B) high mean response times
C) high variance of response times
D) indefinite postponement

Question 21

Access time is defined as ________ plus ________.

A) seek time, positioning time
B) positioning time, transmission time
C) rotational latency, seek time
D) seek time, transmission time

Question 22

Today's hard disks counteract efforts to improve performance using disk scheduling ________.

A) by using alternate sectors dispersed throughout the disk to improve reliability
B) because little effort has been made to increase disk rotational speeds
C) reordering disk requests
D) all of the above

Question 23

Complex disk scheduling strategies are most appropriate ________.

A) under all circumstances
B) for complex file organization structures
C) when accessing localized sequential files
D) under heavy system load

Question 24

________ scheduling is appropriate for online transaction-processing systems.

A) FCFS
B) SSTF
C) C-LOOK
D) none of the above

Question 25

________ caching flushes data to disk immediately, ensuring it cannot be lost in the event of a system failure.

A) Write-back
B) Buffer
C) Write-through
D) Coherent

Question 26

Disk cache buffers are stored ________.

A) in main memory
B) in a memory module attached to the disk drive
C) on a disk interface controller
D) all of the above

Question 27

Compression can improve disk performance at the expense of ________.

A) processor time
B) capacity
C) fault tolerance
D) data integrity

Question 28

Partitioning a drive can potentially ________ fragmentation.

A) increase
B) decrease
C) eliminate
D) None of the above. Partitioning has no effect on fragmentation

Question 29

Disk arm anticipation can be implemented by ________.

A) moving the disk arm to the middle tracks
B) partitioning a disk
C) moving the disk arm to a "hot spot" of disk activity
D) both a and c

Question 30

The letter "I" in RAID originally stood for ________, but has been revised to stand for ________.

A) inexpensive, independent
B) interleaved, independent
C) independent, inexpensive
D) independent, interleaved

Question 31

The original motivation for RAID came from the observation that ________.

A) main memory was faster than secondary storage
B) disk capacities were increasing, but disk I/O transfer rates were increasing at a much slower pace
C) disks were becoming increasingly susceptible to failures
D) all of the above

Question 32

RAID relies on units of data called ________, which when combined into larger chunks are known as ________.

A) stripes, strips
B) fields, blocks
C) sectors, tracks
D) strips, stripes

Question 33

A primary drawback to RAID schemes is greatly decreased ________.

A) performance
B) mean-time-to-failure
C) disk space
D) redundancy

Question 34

RAID level 0 can increase performance ________.

A) at the expense of capacity
B) but does not implement striping
C) but does not provide fault-tolerance
D) all of the above

Question 35

RAID level 1 ________ of all RAID levels.

A) offers the best performance
B) is the most fault-tolerant
C) is the most popular
D) offers the highest total capacity

Question 36

RAID level 1 ________.

A) uses mirroring
B) can yield higher throughput for reads compared to writes
C) is the lowest "true" RAID level
D) all of the above

Question 37

________ is determined according to evenness or oddness of a sum of bits.

A) ECC
B) XOR
C) Parity
D) None of the above

Question 38

RAID level 2 schemes ________.

A) offer higher throughput than RAID level 0
B) are more space-efficient than RAID level 1
C) are rarely implemented
D) both b and c

Question 39

RAID level 3 performs ________ and incurs ________ storage overhead than RAID level 2

A) better, more
B) worse, less
C) better, less
D) worse, more

Question 40

Disks in RAID level 4 are ________.

A) striped at the bit level
B) striped at the block level
C) less fault tolerant than RAID level 3
D) both b and c

Question 41

The most significant bottleneck in RAID level 4 is its ________.

A) read performance
B) write performance
C) parity generation
D) both b and c

Question 42

One technique that can improve RAID level 5 performance, called ________, involves performing a read-modify-write cycle only when the system load is light.

A) AFRAID
B) write combining
C) update imaging
D) parity logging

Question 43

RAID level 5 provides performance that ________.

A) increases fault tolerance at the expense of capacity
B) improve capacity at the expense of fault tolerance
C) balances capacity and fault tolerance
D) depends on the presence a hardware controller