Really the question to ask yourself is why you want to cluster your machines. MPI (Message Passing Interface) is a programming tool in the form of libraries that can be used in C/C++ and Fortran (and I've heard there is a version for Java, but that may be PVM, Parallel Virtual Machine, something similar).
When you write your code with MPI, you explicitly tell it how to break up individual tasks between processors. If you don't have code that is written as parallel code, you won't see any advantage from using multiple processors. If you're interested in this, you should check out the MPICH home page,
http://www-unix.mcs.anl.gov/mpi/mpich/indexold.html
OpenMosix is very different. With this, it isn't necessary to write special parallel code. When a process is spawned, it is automatically migrated to the processor with the most available resources. Even so, unless you are attempting to solve a large problem, the communication time will most likely dominate any savings you may be gaining in computation time. If you want to play around with this, get the openMosix-kernel-2.4.18 package from the openmosix page. The -14 in your kernel is just saying that RH played around with custom kernels before deciding on that one as their stock kernel.
All in all, unless you're attempting to solve some kind of large problems, you won't see any gain from clustering your machines. It is fun to play around though.