INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 2, ISSUE 1, JANUARY 2013

ISSN 2277-8616

A New Heuristic Disk Scheduling Algorithm

*Sandipon Saha, **Md. Nasim Akhter, ***Mohammod Abul Kashem
Abstract:- Since the invention of the movable head disk, people have improved I/O performance by intelligent scheduling of dis k accesses. Processor
speed and memory capacity are increasing several times faster than disk speed. This disparity suggests that disk I/O performance w ill become an
important bottleneck .Methods are needed for using disks more efficiently . Past analysis of disk scheduling algorithms has largely been experimental
and little attempt has been made to develop algorithms w ith provable performance guarantees. Disk performance management is an increasingly
important aspect of operating system research and development. In this paper a new dis k scheduling algorithm has been proposed to reduce the
number of movement of head. It is observed that in existing scheduling algorithms the number of head movement is high. But we proposed a new realtime disk scheduling algorithm that reduces the head movement therefore it maximizes throughput for modern storage devic es.
Key words: Disk scheduling, SSTF, C-LOOK, SCAN, FCFS, C-SCAN, LOOK, HEAD MOVEMENT

1. Introduction
Scheduling is a fundamental operating system function,
since almost all computer resources are scheduled before
use. The disk is of course, one of the computer resources.
For the disk drivers, meeting this responsibility entails
having fast access time and large disk bandwidth.
Processor speed and disk and memory capacity are
increasing by over 40% 40%per year. In contrast, disk
speed is increasing more gradually, growing by only 7%per
year [13]. Since this rate is unlikely to change substantially
in the near future, I/O performance may become the system
bottleneck. However, despite the difficulty of improving
mechanical components, we can still aim to use the disks
more efficiently. The access time has two major
components. For example, disks generally operate at a
small fraction of their maximum bandwidth. Researchers
have demonstrated experimentally that sophisticated disk
head scheduling algorithms can deliver higher throughput
[20, 12, 23]. This past research has focused almost
exclusively on two types of work loads : synthetic work
loads , where disk requests are randomly and uniformly
distributed across the disk, and more recently, traces,
where the requests to an actual disk are recorded and used
as a testing ground for algorithms. However, for these or for
general work loads, researchers have made little attempt to
develop algorithms with provable performance guarantees.
In addition, no one has determined the computational
complexity of the disk scheduling problem. There is a risk
that synthetic workloads and traces from a few
environments may not represent all possible situations. And
the seek time is the time for the disk arm to move the
heads to the cylinder containing the desired sector. The
disk bandwidth is the total number of bytes transferred ,
divided by the total time between the first request for
service and the completion of the last transfer.

We can improve both the access time and bandwidth by

scheduling the servicing of disk I/O requests in a good
order. Whenever a process needs I/O to or from the disk, it
issues a system call to the operating system. The request
specifies several pieces of information:
Whether this operation is input or output.
Whether the disk address for the transfer is
What the memory address for the transfer is
What the number of sectors to be transferred is
If the disk driver and controller are available, the request
can be serviced immediately. If the driver or controller is
busy, any new, request for service will be placed in the
queue of pending requests for the drive. For a
multiprogramming system with many processes, the disk
queue may often have several pending requests. Thus,
when one request is complete, the operating system
chooses which pending request to service next.

2. Proble m State ment and motivation:

Several algorithms exist to schedule the servicing of disk
I/O requests. We illustrate them with a requests queue(10199) :36,180,120,10,15,40,188,150 , 120 , 168.Head starts
at 130.
1. FCFS:The simplest form of disk scheduling is, of course,
the first-come,first-served algorithm.But it generally does
not provide the faster service.Concider , for example a disk
queue with requests for I/O to blocks on cylinders:
Queue(10-199) : 36,180,120, 10 , 15,40,188,150 , 168 .And
head current position is :130. This schedule is diagrammed
in figur 1.1

______________________________
Sandipon Saha, Md. Nasim Akhter, Mohammod Abul
Kashem
Department of CSE, Dhaka University of Engineering
& Technology.
Department of CSE, Dhaka University of Engineering
& Technology.
Department of CSE, Dhaka University of Engineering
& Technology.
49
IJSTR2013
www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 2, ISSUE 1, JANUARY 2013

200

ISSN 2277-8616

3. SCAN:The disk arm starts at one end of the disk and

moves toward the other end, servicing requests until will get
to the other end of the disk, where the head movement is
reversed and the servicing continues.Some time called the
elevator algorithm.Concider , for example a disk queue with
requests for I/O to blocks on cylinders:Queue(10-199)
:36,180,120,10,15,40,188,150 , 168. And head current
position is :130. This schedule is diagrammed in figur 1.3

180
160
140
120
100
80
60

200
180
160
140
120
100
80
60
40
20
0

40
20
0

Figure:1.1
Head movement: (130-36)+ (80-36)+(120-80) +(120-10)
+(15-10) +(40-15) +(188-40)+ (188-150) +(150-120) +(168120)=582
In FCFS total head movement: 582
2. SSTF: Shortest Seek Time First-Selects the request with
the minimum seek time from the current head osition.Since
seek time increases with the number of cylinders traversed
by the head, SSTF chooses the pending request closest to
the current head position.SSTF scheduling is a form of SJF
scheduling:may cause starvation of some requests.
Concider , for example a disk queue with requests for I/O to
blocks on cylinders:
Queue(10-199): 36,180,120,10,15,40,188,150 , 168. And
head current position is :130. This schedule is diagrammed
in figur 1.2

200
180
160
140
120
100
80
60
40
20
0

Figure:1.3
Head movement:(130-120) +(120-80) +(80-40) +(40-36)
+(36-15) +(15-10) +(10-0) +(150-0) +(168-150) +(188168)=318
In SCAN total head movement: 318.
4. C-SCAN: Provides a more uniform wait time than SCAN.
The head moves from one end of the disk to the other.
Servicing requests as it goes. However, when it reaches of
the other end, it immediately will return to the beginning of
the disk, without servicing any requests on the return trip.
Treats the cylinders as a wraparound circular list from the
first cylinder to the last one. Concider , for example a disk
queue
with
requests
for
I/O
to
blocks
on
cylinders:Queue(10-199):36,180,120,10,15,40,188,150,
168. And head current position is :130. This schedule is
diagrammed in figur 1.4

Figure :1.2
Head Movement:(130-120) +(150-120) +(168-150) +(188168) +(188-80) +(80-40) +(40-36) +(36-15) +(15-10)=256
In SSTF total head movement: 256.
50
IJSTR2013
www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 2, ISSUE 1, JANUARY 2013

ISSN 2277-8616

innermost and outermost cylinders each sweep, LOOK will

change directions when it has reached the last request in
the current direction. Concider , for example a disk queue
with requests for I/O to blocks on cylinders : Queue(10199): 36,180,120,10,15, 40,188,150 ,168. And head current
position is :130. This schedule is diagrammed in figur 1.6

250
200
150

200
180
160
140
120
100
80
60
40
20
0

100
50

Figure:1.4
Head movement: (150-130) +(168-150) +(188-168)+(199188) +(199-0) +(10-0) +(15-10) +(36-15) +(40-36) +(80-40)
+(120-80)=388.
In C-SCAN total head movement: 388

Figure:1.6
5. C-LOOK: A version of C-SCAN. Arm goes only as far as
the last request in each direction, the reverses direction
immediately, without first going all the way to the end of the
disk. Concider , for example a disk queue with requests for
I/O
to
blocks
on
cylinders:
Queue(10-199):
36,180,120,10,15, 40,188,150 ,168. And head current
position is :130. This schedule is diagrammed in figur 1.5

Head movement: (130-120) +(120-80) +(80-40)+(40-36)

+(36-15) +(15-10) +(150-10) +(168-150) +(188-168)=298
In LOOK total head movement: 298

3. Proposed Disk Scheduling Algorithm:

At first sorting in ascending order of all cylinders input
blocks by using any sorting method. Find the distance
between the smallest block number and current head
position. Let it is P and again find the distance between the
largest block number and current head position. Let it is Q.
Sequentially move and reached head from these block to
the highest block number. Else head moves sequentially
from its current position to the highest block number in
forward and again in backward which block is not visited.
Then sequentially move and reached head from these block
to the largest block number.

200
180
160
140
120
100
80
60
40
20
0

Figure 1.5
Head movement: (150-130)+(168-150) +(188-168) +(18810) +(15-10) +(36-15) +(40-36) +(80-40) +(120-80)=346
In C-LOOK total head movement: 346
6. Look: LOOK is similar to SCAN in that the heads sweep
across the disk surface in both directions performing reads
and writes. However, unlike SCAN, which visits the

New Heuristic Disk Scheduling Algorithm (a, n, count,

h)
1. // a [] is an array that contains cylinders number. N
is the number of cylinder. Count is use
2. // for counting head movements. h denote the
present head position.
3. Sorting input blocks of cylinder number in
ascending order by any sorting method.
4. Input present head position h.
5. Temp:=h;
6. For i :=1 to n do
If(a[i]>=h) { Position=i; break;}
7. Left_distance:=head-a [1];
Right_distance:=a[n]-head;
8. Count:=0;
9. If (Left_distance< Right_distance)
{
51

IJSTR2013
www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 2, ISSUE 1, JANUARY 2013

For i:=position-1 to 1 step -1 do {

count:=count+Temp-a [i]; Temp:=a[i];
}
count :=count+[position]-a[1];
For I: =position+1 to n do
count: =count + a [i]-a [i-1];

ISSN 2277-8616

Rd=a[n]-Hp

Rd a[n] a[1]; Ld
Ld a[n] a[1]; Ld

Hm =

Rd
Rd

Calculation: Let ,
Queue(10-199) :36,180,120,10,15,40,188,150 , 168.
Head starts at: 130.
Ld=130-10=120
Rd=188-130=58
Here, Ld>Rd so,
Hm=58+188-10=236
Total head movement=236(For new algorithm)

}
Else
{
For i:=position to n do
{count:=count+a[i]-head;
head:=a[i];}
Count:=count+a[n]-a[position-1];
For i:=position-1 to 2 do
Count:=count+a[i]-a[i-1];

Comparisons table among proposed and existing

algorithms:

}
10. Return count; // total head movement
Graphical representation of proposed algorithm:
Concider , for example a disk queue with requests for I/O to
blocks on cylinders : Queue(10-199) :36,180,120,10,15,
40,188,150 ,168. And head current position is :130. This
schedule is diagrammed in figur 2

200
180
160
140
120
100
80
60
40
20
0

SL.NO

Name of
Algorithm

Number of head
movement

1.

FCFS

582

2.

C-SCAN

388

3.

SSTF

256

4.

C-LOOK

346

5.

LOOK

298

6.

SCAN

318

7.

NEW
Heuristic

236

Comparisons Graph among proposed and existing

algorithms. That show performances:

600
500
Figure: 2

400

Head movement: (150-130)+(168-150) +(188-168) +(188120) +(120-80) +(80-40) +(40-36) +(36-15) +(15-10)=236

300

In new heuristic algorithm total head movement: 236

100

200

General Equeation :
Let ,
Left distance=Ld
Right distance =Rd
Head position=Hp
Request queue=a[]
Max queue position=n
Total Head Movement=Hm
Ld=Hp-a[1]

Figure:3
52
IJSTR2013
www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 2, ISSUE 1, JANUARY 2013

ISSN 2277-8616

Limitations:
1. Sometime number of head movement is equal to
SSTF or LOOK scheduling.
2. When input blocks are stay in ascending order
without sorting then FCFS is best. But in dynamic
allocation of cylinder it is almost impossible.

[12] Ruemmler, C., And wiklkes,J. An introduction to

disk drive modeling. Computer 27, 3(1994),17-28.

Conclusion:
In conclusion, we have presented a new real-time disk
scheduler that imposes almost no performance penalty over
non-real-time optimal schedulers when given sufficient
slack time. We showed how to characterize a devices
performance. From the above experiment and comparison
of proposed algorithm with existing algorithm it is clear to us
that the existing algorithm reduces head movement. Who
wants to work with disk scheduling this algorithm will open
new era for them. This would help the new generation to be
go ahead.

[14] Seaman, P.H., Lind, R. A., And Wilson, T. L. An

analysis of auxiliary-storage activity. IBM System
Journal 5, 3(1996), 158-170.

References
[1]

Operating System Principles (6th edition) Abraham

Silberschatz, Peter Bare Galvin, Greg Gagne

[2]

Mordern operating system (2nd edition) Andrew S.

Tanenbaum .

[3]

Operating
Systems:
A
Approach(2E) D.M. Dhamdhere

Concept-based

[4]

Intel, And Seagate. Serial Ata native command

queuing july 2003.

[5]

Kumar, R. Fairness in disk scheduling. Master

thesis, Indian Institute of science, Bangladesh,
India, Jan. 1993.

[6]

SCSI architecture model-3(SAM3).Tech. Rep. T10

project 1561-D ,revision 14,International for
information
Committee Technology standards
(INCITS),T10 Technical Committee. Sept. 2004.

[7]

Chang, R., Shih, W., And Chnag, R. Real-time disk

scheduling for multimedia application with
deadline-modification-scan scheme. International
Journal of Time-critical Computing System
19(2000),149-168.

[8]

Dees, B. Native Command queuing-advanced

performance in desktop storage. Potentials, IEEE
24, 4(2005),4-7.

[9]

Frank, H .Analysis and optimization of disk storage

devices for Time-sharing system. J ACM 16,
4(1969),602-620.

[13] C. Ruemmler and J. Wilkes. An introduction to disk

drive modeling. IEEE Computer, 27(3):1729,
March 1994.

[15] Seltzer, M., Chen, P., And Ousterhout, J. Disk

scheduling revisited, Jan . 1990.
[16] Yu, Y . J.,Shin, D. I., Eom, H., And Yeom , H. y.
NCQ vs.I/O scheduler: Preventing unexpected
misbehaviors. Trans. Storage 6, (2010), 1-37.
[17] Zhu, Y. Evolution of scheduling algorithms for realtime disk I/O. Tech.rep., Department of computer
science and Engineering, University of Nebraska,
May 2002.

[18] Reddy, A.L.N., Wyllie, J., and Wijyaratne, K.B.R.

disk scheduling in a multimedia I/O system. ACM
Trans. Multimedia Comput. Commun. Appl. 1,
1(2005), 37-59.

[19] Mesnier, M. P.,Wachs, M., Sambasivan, R.R. ,

Zheng, A.X., And Ganger, G .R . Modeling the
relative fitness of storage. Sigmetrics Perform.
Eval. Rev. 35, 1(2007),37

[10] Huang, Y., And Huang, J. Disk scheduling on

multimedia storage servers. Computers, IEEE
Transactions on 53, 1(2004), 77-82.
[11] Reddy,A.L.N.,And Wyllie,J. Disk Scheduling in a
multimedia I/O system. ACM, pp.225-223.

53
IJSTR2013
www.ijstr.org