Efficient scheduling algorithms are essential to irregular data redistribution in cluster grid. Cluster grid is an environment with heterogeneous computing nodes and complex network. It is important for schedulers to keep an eye on load balance and low communication cost while distributing different size of data segment on various processors. High Performance Fortran Version 2 (HPF2) provides GEN_BLOCK distribution format which facilitates generalized block distributions. In this paper, we present a message clustering technique to derive low communication cost when performing such operation in cluster grids. The main idea of the proposed technique is to cluster three kinds of messages and normalize the cost. The performance evaluation is given and show the proposed method successfully adapts to heterogeneous environment.