In this paper, we investigate a mechanism for handling resource sharing among tasks under a server-based scheduling technique in multiprocessor platforms, which combines partitioned and global scheduling to benefit a better scheduling method compared to conventional techniques.