… As an alternative, we present a method to increase resilience … This approach effectively
increases the time between faults … of some faults, it is now possible to increase the checkpoint …

… We use signals to alert rMPI of failed nodes and can thus simulate the failure of nodes for
testing purposes. However, since all nodes still are part of a complete MPI application and due …

… and scalable fault-tolerant environment requires the … on resilience in large-scale computing
with the focus of fault-… rMPI [9] is a fault-tolerant implementation of MPI through process …

… File and remote memory access operations are not supported by ULFM and are likely to
fail in a faulty environment (rather than raising an error, they throw a segmentation …

… problems, rMPI does not support certain complex MPI communicator operations, such as MPI
Comm split(). rMPI … Increasing fault resiliency in a messagepassing environment. Technical …

… the overhead of providing resiliency through redundancy. … ideas that could improve the
performance of rMPI and limit … Increasing fault resiliency in a message-passing environment. …

… must become tolerant to such faults. With large numbers of … contributes resiliency features
to the “Exascale Environment” … were also released following rMPI in 2011. After reiterating …

… languages for DS is Message Passing Interface (MPI). … rely on any specific MPI library
whereas rMPI is specific to MPICH. … , then explain the need for resilience in this library, and finally …

… We integrate our solution into an existing fault-resiliency … of fault management techniques in
the Message Passing Interface (… to fault presence consists of letting the MPI call raise an error…

… MPI/FT [9], P2P-MPI [19] and rMPI [16] are MPI libraries that provide fault-tolerance through
… failure or major performance degradations. Furthermore, we demonstrated that increasing …