Determining the checkpoint frequency and interval to minimize the checkpoint overhead is main objectives in many research related to checkpoint algorithm in fault tolerant systems. Taking a checkpoint too infrequently may spend too much time to reprocess its execution after failures. On the other hand, taking checkpoint too frequently will increase the checkpoint overhead and will affect the total execution time of a process. In this paper, we propose memory profiling checkpoint interval algorithm which uses an adaptive time series analysis to adjust checkpoint interval dynamically. Experimental results from trace-driven simulation show that the total checkpoint overhead of the proposed algorithm is smaller than that of the static checkpoint interval algorithm and it is much closer to that of the optimal checkpoint interval.
展开▼
