Aloocatin Future Scope
Aloocatin Future Scope
Aloocatin Future Scope
Department of CSE
Guru Nanak Dev University, India
Department of CSE
Guru Nanak Dev University, India
ABSTRACT
Cloud computing has emerged as a popular computing model
to support on demand services and is rapidly becoming an
important platform for scintific applications. It provide users
with infrastructure, platform and software as amenity which is
effortlessly accessible via Internet. It has a huge user group
and has to deal with large number of task, so schedulling in
cloud plays a vital role for task execution. In this paper,
scheduling polices space-shared and time-shared
are
compared on the bases of some parametrs which are Task
Profit, Task Penalty, Throughput and Net Gain. In our
simulation results we shown that space-shared outperforms
than time-shared policy.
Index Terms
2. BACKGROUND
1. INTRODUCTION
29
3. SCHEDULING ARCHITECTURE
The main target of scheduling is to maximize the resource
utilization and minimize processing time of the tasks. The
scheduler should order the tasks so that balance between
improving the quality of services and at the same time
maintaining the efficiency and fairness among the tasks [4].
An efficient task scheduling strategy must goal to yield less
response time so that the execution of submitted tasks takes
place within a deadline as a consequences of this, tasks takes
place and more number of tasks can be submitted to the cloud
by the users which results in enhancement of the performance
of the cloud system [5].
4. SCHEDULING POLICIES
There are two polices which are defined under this section
these are Space-Shared scheduling policy and Time-Shared
scheduling policy.
Step 3:- It completes first task and then take the next task
from the queue.
Step 1:- All accepted task are arranged under the queue.
Step 2:- Then schedule the task simultaneously on the virtual
machine.
Step 3:- When queue is empty it checks for new task.
30
5. SIMULATION RESULTS
In this section, we examine the practical abilities of Spacesahred and Time shared approach. We firstly extant the
experimental situation and then compare the performances of
Space-shared and Time-shared scheduling algorithm of cloudsim.
A. Scenario
The performance measurement has been gathered by using the
CloudSim [2] framework. The goal of this study is to show
the ability of space-shared approach to meet deadlines in
conditions where the Time Shared could not able to meet
deadlines. To accomplish this, we recommend the following
situation. We label identical datacenter composed of hosts.
Each Processing element has a speed of 1000 MIPS. An
available bandwidth of 10 Gbit/s. A storage capacity of 1 TB.
A Random Access Memory (RAM) of 2048 MB.
We process 1000 cloudlets on the earlier demarcated
datacenter using two policies: Space-Shared and TimeShared approach. The lengths of cloudlets are equal and each
cloudlet is given a randomly generated deadline where
minimum value is half the number of cloudlets and maximum
value is equal to number of cloudlets. Cost of each cloudlets is
also depend upon the number of cloudlets and it is also
randomly generated.
B. Task Profit
The number of tasks which have completed successfully
before they meet the deadline. Space-shared approach is more
efficient to finish tasks before the deadline arrives than Timeshared scheduling algorithm. We compare number of tasks
completed by each approach for the given set of cloudlets. In
figure 3. We can see the results of experiment.
On the x-axis it shows the number of cloudlets and y-axis
indicates number of tasks completed successfully before
deadline reaches. By examining the figure we can say that
Time-shared algorithm inefficient to meet deadline before it
comes.
We can notice that, in this case both approaches seems to be
quiet indistinguishable but space-shared completes almost all
tasks before it meets the deadline. By increasing the number
of cloudlets and virtual machines we can observe the more
difference and conclude that Space-shared approach is better
than Time-shared approach.
1000
TaskProfit
800
600
400
200
0
No of Tasks
SpaceShared
TimeShared
31
C. Task Penalty
E. Net Gain
120
TaskPenality
100
80
60
40
600000
20
0
0
500
1000
1500
No of Tasks
SpaceShared
TimeShared
NetGain
500000
400000
300000
200000
100000
0
D. Throughput
SpaceShared
TimeShared
Throughput
No of Tasks
1000
900
800
700
600
500
400
300
200
100
0
7. REFERENCES
[1] The NIST definition of cloud computing, NIST special
publication 800-145.
No of Tasks
SpaceShared
TimeShared
32
IJCATM : www.ijcaonline.org
33