Hybrid reliable load balancing with MOSIX as middleware and its formal verification using process algebra

For many years, load balancing in distributed computing environment has been researched to enhance and optimize the scalability of the entire system. Although numerous works have been proposed on the issues of process migration and load balancing, a comprehensive approach still misses out on various fronts. An effort has been made in the present work to cite a reliable and comprehensive load balancing approach for a trusted small scale distributed computing environment based on the priority of the processes. The proposed approach introduces a Process Migration Server (PMS) that also acts as future cluster management server ensuring that the latency time in migrated process execution is reduced along with no starvation policy for any process. We also propose an effective resource sharing mechanism where our system is able to distribute the load evenly among various nodes without much overhead. This in turn increases the reliability of the overall system. To verify our approach, we have implemented hybrid reliable load balancing algorithm where MOSIX is deployed as middleware and significant performance improvement has been observed. We have compared the performance of the proposed algorithm to the default MOSIX. Our results indicate that, hybrid reliable load balancing has lowered the network traffic by 80%–90%, increased CPU utilization by 40%–50%, with 4%–9% lesser memory and computational requirement. A relationship between the formal aspect of hybrid reliable load balancing algorithm and its architectural model JMM is established through process algebra approach.

Research highlights
► A hybrid load balancing (HLB) approach in trusted clusters is proposed for HPC.
► HLB reduces network traffic by 80%–90% and increases CPU utilization by 40%–50%.
► The AWT and MRT of remote processes are reduced by 13%–26% using ReJAM.
► The stability analysis of JMM using PA ensures the finite sequences of transitions.
► On the basis of these properties, JM model has been proved safe and reliable.