Laboratory/GridComparisonPartII/Hadoop-0.20.1

Hadoop 0.20.1 (comparison of the grid/cloud computing frameworks - part II)

Test environment

During fail-over tests we used the same test environment as in the part I of our comparison, but with a small change in configuration. We had to increase dfs.replication from 1 to 3.

Moreover, we simulated node failures during our tests in the following order:

intel4 went down after 60 seconds from the beginning of computations
intel3 went down after 180 seconds from the beginning of computations
intel2 went down after 300 seconds from the beginning of computations

Code

During fail-over tests we used the same code as in the part I of our comparison. The only thing that has changed was new fail-over hook just before job.waitForCompletion(true) invocation. It allowed us to precisely synchronize the test environment.

You can find all the code in our repository: http://dacframe.org/lab

Results

You can see all the results with std deviation and average values on the following table:

Hadoop 0.20.1 <-->

	Computation time	Repeated tasks
Average	1 307 526.80
Std Deviation	181 570.01
	1 593 734.00	123
	1 194 784.00	45
	1 177 549.00	43
	1 164 319.00	44
	1 615 523.00	113
	1 178 317.00	40
	1 463 073.00	107
	1 145 385.00	44
	1 268 696.00	50
	1 273 888.00	63

As you can see on the above table, both average time and std deviation are much bigger in the fail-over test than in our first comparison. Average time increased from 384 331.60 by 240.21%, std deviation increased from 5 629.01 by 3125%. Moreover, a lot of tasks were computed several times. This means, that Hadoop is quite sensitive to unexpected node failures.