Staggered consistent checkpointing

A consistent checkpointing algorithm saves a consistent view of a distribut ed application's state on stable storage. The traditional consistent checkp ointing algorithms require different processes to save their state at about the same time. This causes contention for the stable storage, potentially resulting in large overheads. Staggering the checkpoints taken by various p rocesses can reduce checkpoint overhead. This paper presents a simple appro ach to arbitrarily stagger the checkpoints. Our approach requires that the processes take consistent logical checkpoints, as compared to consistent ph ysical checkpoints enforced by existing algorithms. Experimental results on nCube-2 are presented.