Hybrid metaheuristic model based performance-aware optimization for map reduce scheduling

Abstract

Because of the rapid rise in the count of corporations using cloud-dependent infrastructure as the foundation for big data storing and analysis. The fundamental difficulty in scheduling big data services in cloud-dependent systems is ensuring the shortest possible makespan while simultaneously reducing the number of resources being used. We have created a new, secure map reduce scheduling method that functions as follows. Initially, the cloud architecture is designed and the tasks are generated. In the pre-processing phase, the huge set of tasks was processed by the map-reduce scheduling framework. Afterward, the optimal task scheduling task is conducted which utilizes a hybrid algorithm named Tunicate Combined Moth Flame Algorithm (TCMFA), which provides better task scheduling via providing optimal makespan, execution time, and security. This proposed TCMFA is the hybridization of both Moth Flame Optimization (MFO) and Tunicate Swarm Algorithm (TSA). The error rate of the TCMFA gets reduced to 320 approximately over other conventional methods such as RSA, ACO, GHO, BTS, OWPSO, BES, PRO, SOA, COOT, TSA & MFO which proves the accuracy of our TCMFA and makes it more efficient and secure for optimal map-reduce scheduling.