Biomedical Engineering Reference
In-Depth Information
Hadoop cluster
in
f
ormation
1
Aggregation
Interval
0.95
0.9
-60
40
140
Number of client requests/s
Fig. 5 Accuracy changes with number of nodes and aggregation level
Hadoop cluster
consumption
0.015
Consumption
vs Nodes
0.01
0.005
0
-60
40
140
Number of nodes/client requests
Fig. 6 Consumption level is decreasing as client requests are increasing
5.2 Experiment to Test Accuracy
In this experiment, we considered only the important metrics: CPU, memory, and
disk usage and network workload. Figure
5
shows the metric accuracy, which
varies with aggregation interval and number of nodes in the cluster. With the
increase in aggregation interval, the accuracy of monitoring information will
decrease, but at a certain point, direction is changed toward increase in accuracy.
The accuracy is also increasing as the number of nodes in the cluster is increasing.
5.3 Consumption Experiments
Figure
6
shows the consumption with different number of nodes or client requests.
As the number of nodes is very less, consumption is high; it gradually reached
stable state and then reduced gradually as the node number is increasing.
6 Conclusion and Future Work
In this paper, we presented a new methodology of providing integrity and
authenticity for the data outsourced by the user over cloud using FNV, a non-
cryptographic hashing technique. This paper implemented an ef
cient and secure
Search WWH ::
Custom Search