Ff84602 - CHP - ISM Book Exc Solutions
Ff84602 - CHP - ISM Book Exc Solutions
Ff84602 - CHP - ISM Book Exc Solutions
Chapter 2
1. What are the benefits of using multiple HBAs on a host? Solution/Hint: - High availability An application specifies a requirement of 200GB to host a database and other files. It also specifies that the storage environment should support 5,000 IOPS during its peak processing cycle. The disks available for configuration provide 66GB of usable capacity, and the manufacturer specifies that they can support a maximum of 140 IOPS. The application is response time sensitive and disk utilization beyond 60 percent will not meet the response time requirements of the application. Compute and explain the theoretical basis for the minimum number of disks that should be configured to meet the requirements of the application. Solution: Number of disk required = max (size requirement, IOPS requirements) To meet the size requirement = 200 GB/66 GB= 4 disks To meet the IOPS requirement= 5000 IOPS/(1400.6 IOPS)= 60 disks = max (4, 60) = 60 disks 3. Which components constitute the disk service time? Which component contributes the largest percentage of the disk service time in a random I/O operation? Solution: - seek time, rotational latency and transfer rate - seek time Why do formatted disks have less capacity than unformatted disks? Solution: In order to make storage device functional, it need to be formatted. Common types of drive formats are FAT32, NTFS and ext2. In each of the formatting schemes, a portion of the storage space is allocated to configured file system to enable cataloging data on the disk drive.
2.
4.
Page
1 of 4
5.
The average I/O size of an application is 64 KB. The following specifications are available from the disk manufacturer: average seek time = 5 ms, 7,200 rpm, transfer rate = 40 MB/s. Determine the maximum IOPS that could be performed with the disk for this application. Taking this case as an example, explain the relationship between disk utilization and IOPS. Solution: The disk service time (Rs) is a key measure of disk performance; and Rs along with disk utilization rate (U) determines the I/O response time for applications. The total disk service time (Rs) is the sum of seek time (E), rotational latency (L), and the internal transfer time (X): Rs = E+L+X E is determined based on the randomness of the I/O request. L and X are measures provided by disk vendors as technical specifications of the disk. Average seek time of 5ms in a random I/O environment, or E=5ms Disk rotation speed of 7,200 rpm from which rotational latency (L) can be determined, which is one half of the time taken for a full rotation or L= (0.5/7,200 rpm expressed in ms) 40 MB/s internal data transfer rate, from which the internal transfer time (X) is derived based on the block size of the I/O. With a block size of 64 KB, X= 64 KB/40 MB Consequently Rs = 5 ms + (0.5/7,200) + 64 KB/40 MB = 5+4.167+1.6 = 10.767 ms The maximum no. of I/Os serviced per second or IOPS = 1/ Rs In other words, for an I/O with a block size of 64 KB and Rs = 10.767 ms, the maximum IOPS will be 1/(10.767*10-3) = 92.876 IOPS
Page
2 of 4
6.
Consider a disk I/O system in which an I/O request arrives at the rate of 80 IOPS. The disk service time is 6 ms. a. Compute the following: Utilization of I/O controller, Total response time, Average queue size, and Total time spent by a request in a queue. b. Compute the preceding parameter if the service time is halved. Solution: Arrival rate (a) = 80 IOPS, consequently, the arrival time Ra = 1/a = 1/80 = 12.5 ms Rs = 6 ms (given) 1. Utilization (U) = Rs/Ra = 6/12.5 = 0.48 or 48% 2. Response time (R) = Rs/(1-U) = 6/(1-0.48) = 11.5 ms 3. Average queue size = U2/(1-U) = (0.48)2/(1-0.48) = 0.44 4. Time spent by a request in a queue = UxR, or the total response time-service time = 0.48x11.5 = 5.52 ms Now, if controller power is doubled, or the service time is halved; consequently Rs = 3 ms in this scenario. 1. Utilization (U) = Rs/Ra = 3/12.5 = 0.24 or 24 % 2. Response time (R) = Rs/(1-U) = 3/(1-0.24) = 3.9 ms 3. Average queue size = U2/(1-U) = (0.24)2/(1-0.24) = 0.08 4. Time spent by a request in a queue = 0.24x3.9 = 0.936 ms
7.
Refer to Question 6 and plot a graph showing the response time and utilization, considering 20 percent, 40 percent, 60 percent, 80 percent, and 100 percent utilization of the I/O controller. Describe the conclusion that could be derived from the graph. Solution: 1. Utilization = 20 % = 0.2 Response time = 7.5 ms 2. Utilization = 40 % = 0.4
Page
3 of 4
Response time = 10 ms 3. Utilization = 60 % = 0.6 Response time = 15 ms 4. Utilization = 80 % = 0.8 Response time = 30 ms 5. Utilization = 100 % = 1 Response time = infinity
Page
4 of 4