In recent years, an efficiency of task scheduling is evolved as a major challenge in cloud platforms. Especially, identifying the optimal resources for input tasks is the major challenges faced by the task scheduler. So, in this research, a Modified Whale Optimization Algorithm (MWOA) is proposed to improve the behaviour in task scheduling by applying the parameters such as resource allocation and load balancing. Capacity criteria based MWOA algorithm determines the effective Virtual Machine (VM) for execution of tasks in queue. An effectiveness of proposed WOA algorithm is analysed by the utilization of performance measures such as memory storage, execution time, cost and makespan. The attained results from the proposed WOA algorithm are outperformed in various predictable optimization algorithms such as Storm, Spark, Flink and Kafka. The results of proposed WOA algorithms demonstrates the better performances in minimum execution time of 612ms, and memory storage of 309Kb on Kafka platform.