Re: [R] Rmpi task-pull

])) mpi.remote.exec(sessionInfo()) mpi.close.Rslaves() mpi.quit() For some more debugging you can start your cluster with log output: mpi.spawn.Rslaves(nslaves=3, needlog=TRUE) Then there should be logfiles for every node in your working directory. Best Markus Daniel Ferrara wrote: Hi, I'm testing

[R] Rmpi task-pull

Hi, I'm testing the efficiency of the Rmpi package regarding parallelization using a cluster. I've found and tried the task pull programming method, but even if it is described as the best method, it seems to cause deadlock, anyone could help me in using this method? here is the code I've found