Computational grid is made up of virtual resources and differs from HPC, as it's used in scientific and technological computation-intensive problem solving. In computational grid, resources are identified as underutilized, less loaded and over loaded. Load balancing becomes even more challenging in grid because of its heterogeneous nature and computation-data separation among resources. Utilization of these resources through co-ordination of various loads is always considered an optimization problem. In order to achieve this, approach of adaptive resource ranking is applied in this research. This adaptive methodology of resource co-ordination in the computational grid is based on historical or average load of each resource along with their current load for a defined period. Except resource co-ordination, the proposed algorithm also introduces periodical runtime backup to another available resource for retaining quality-of-service as approved in service level agreement. The experiments carried out in this book establish the proposition for optimized model and fault tolerant model, in case of unexpected resource failure. Hence, quality of service is achieved for both of these models.