map_reduce_vertices/edges and transform_vertices/edges are not parallelized on Mac OS X
These operations currently rely on OpenMP for parallelism.
On OS X 10.6 and earlier, gcc 4.2 has several OpenMP bugs and is not stable enough to use reliably.
On OS X 10.7, the clang++ compiler does not yet support OpenMP.
We are now working to eliminate the OpenMP dependency.
map_reduce_vertices/edges and transform_vertices/edges use a lot more processors than what was specified in –ncpus
This is related to the question above. While there is a simple temporary solution (omp_set_num_threads), we intend to properly resolve the issue by not using openMP at all.
Unable to launch distributed GraphLab when each machine has multiple network interfaces
The communication initialization currently takes the first non-localhost IP address as the machine’s IP. A more reliable solution will be to use the hostname used by MPI. A solution is in the works.