Monitoring gird resources using Web Service is an important issue in grid computing.Although some concern has been given to the Index Service scalability with increasing grid resources, the overhead of processing thousands of resource information requests may cause performance problems in Index service for grid monitoring.This paper proposed a replica scheduling mechanism based on load balancing with adjusted capability to meet the performance requirements of Index Service. We have designed and implemented this mechanism on a Scalable Grid Monitoring System (SGMS) which is WSRF/WSRP-based.From the results of evaluations, we conclude that SGMS can enhance efficiency of Index Service and guarantee high-scalability of Index Service with increasing requests.