The performance of coordinated and independent checkpointing

Checkpointing is a very effective technique to tolerate the occurrence of failures in distributed and parallel applications. The existing algorithms in the literature are basically divided into two main classes: coordinated and independent checkpointing. This paper presents an experimental study that compares the performance of these two classes of algorithms. The main conclusion of our study is that coordinated checkpointing is more efficient than independent checkpointing and all the arguments against the performance of coordinated algorithms were not verified in practice