MPICH-V2
Aurélien Bouteiller, Franck Cappello, Thomas Herault A et al. · Proceedings - Supercomputing · 2003
Execution of MPI applications on clusters and Grid deployments suffering from node and network failures motivates the use of fault tolerant MPI implementations. We present MPICH-V2 (the second protocol of MPICH-V project), an automatic fault tolerant MPI implementation using an innovative protocol that removes the most limiting factor of the pessimistic message logging approach: reliable logging of in transit messages. MPICH-V2 relies on uncoordinated checkpointing, sender based message logging and remote reliable logging of message logical clocks. This paper presents the architecture of MPICH-V2, its theoretical foundation and the performance of the implementation. We compare MPICH-V2 to MPICH-V1 and MPICH-P4 evaluating a) its point-to-point performance, b) the performance for the NAS benchmarks, c) the application performance when many faults occur during the execution. Experimental results demonstrate that MPICH-V2 provides performance close to MPICH-P4 for applications using large messages while reducing dramatically the number of reliable nodes compared to MPICH-V1.
Cited by 174 works. doi:10.1145/1048935.1050176
Authors
References
- Impossibility of distributed consensus with one faulty process
- The NAS Parallel Benchmarks
- Distributed snapshots
- A high-performance, portable implementation of the MPI message passing interface standard
- MPI: The Complete Reference
- A survey of rollback-recovery protocols in message-passing systems
- Optimistic recovery in distributed systems
- Introduction to Distributed Algorithms
- The Network Architecture of the Connection Machine CM-5
- CoCheck: checkpointing and process migration for MPI