I'm working on computing (MapReduce) security of the Big Data processing systems. The problem is that I do not know, in a system like Hadoop, it's possible or not to change the assignment of the tasks by the programmer.
For instance, Suppose there is a problem with one of the nodes and I want to not to be assigned task to it after identifying that node as a malicious node.
1. I know that Hadoop does the task assignment automatically. And I
want to get this out of the automatic mode.
2. Is it necessary to compile and modify Hadoop source?
3. Can be designed an additional module for Hadoop?
4. Is not it better to use cloudsim for this? (that's mean, I simulate an environment like Hadoop in cloudsim.)
On the other hand, Implementing MapReduce in cloudsim has its own problems.
Please Help me, Which method to continue?
Do you have a better idea?