Revisiting the double checkpointing algorithm

Fast check pointing algorithms require distributed access to stable storage. This paper revisits the approach base upon double check pointing, and compares the blocking algorithm of Zheng, Shi and Kalé, with the non-blocking algorithm of Ni, Meneses and Kalé, in terms of both performance and risk. We also extend their model proposed to assess the impact of the overhead associated to non-blocking communications. We then provide a new peer-to-peer check pointing algorithm, called the triple check pointing algorithm, that can work at constant memory, and achieves both higher efficiency and better risk handling than the double check pointing algorithm. We provide performance and risk models for all the evaluated protocols, and compare them through comprehensive simulations.