Resource selection using fuzzy logic for Dynamic Service Placement and Replication

One of the core components of the DSPR framework is the resource evaluation function. The Dynamic Service Placement and Replication (DSPR) framework is designed to manage services on distributed environment by repetitively searching for the best combination of distributed machines from a resource pool until the best working group is attained to satisfy the requirements of the service. In this research, a fuzzy logic based resource evaluation technique is proposed for the DSPR framework. The proposed technique is then compared with existing resource evaluation methods such as rule-based, utility-based and constraint-based. Besides that, the ability of DSPR to increase service availability in the event of multiple continuous failures is measured. The performance difference between using DSPR and conventional system with failover capability in the event of disaster was carried out. These experiments justify the suitability of using the proposed resource evaluation techniques compare to the existing work.