Section: New Results

Optimal checkpointing period with replicated execution on heterogeneous platforms

Participants : Anne Benoit, Aurélien Cavelan [University of Basel, Switzerland] , Valentin Le Fèvre, Yves Robert.

In this work, we design and analyze strategies to replicate the execution of an application on two different platforms subject to failures, using checkpointing on a shared stable storage. We derive the optimal pattern size $W$ for a periodic checkpointing strategy where both platforms concurrently try and execute $W$ units of work before checkpointing. The first platform that completes its pattern takes a checkpoint, and the other platform interrupts its execution to synchronize from that checkpoint. We compare this strategy to a simpler on-failure checkpointing strategy, where a checkpoint is taken by one platform only whenever the other platform encounters a failure. We use first or second-order approximations to compute overheads and optimal pattern sizes, and show through extensive simulations that these models are very accurate. The simulations show the usefulness of a secondary platform to reduce execution time, even when the platforms have relatively different speeds: in average, over a wide range of scenarios, the overhead is reduced by $30\%$. The simulations also demonstrate that the periodic checkpointing strategy is globally more efficient, unless platform speeds are quite close.

This work is available as a research report. A preliminary version appears in the proceedings of the FTXS'17 workshop.