A Novel Chicken Pecking Order Algorithm for Efficient Map-Reduce

摘要

The ‘Map Reduce System’ works as a distributed system comprises of concurrent servers that runs several parallel tasks, provides data communication within the system, provides redundancy and fault tolerance. But, the most common difficulty seen in MapReduce is the Un-even load balancing problem. (i.e.,) Whenever a node in a distributed environment takes up a task of which is highly difficult for it to compute or very easy to perform (two extremities), un-even load balancing can occur. This paper suggests a new bio-inspired algorithm called ‘Chicken Pecking Order (CPO) Algorithm’ to effectively allocate and monitor dominant and recessive nodes for faster computation. The algorithm is explained theoretically and a real-time transactional example is addressed and solved using the developed CPO Algorithm.