Using MPI with IPython#

Often, a parallel algorithm will require moving data between the engines. One way of accomplishing this is by doing a pull and then a push using the direct view. However, this will be slow as all the data has to go through the controller to the client and then back through the controller, to its final destination.

A much better way of moving data between engines is to use a message passing library, such as the Message Passing Interface (MPI). IPython’s parallel computing architecture has been designed from the ground up to integrate with MPI. This document describes how to use MPI with IPython.

Additional installation requirements#

If you want to use MPI with IPython, you will need to install:

A standard MPI implementation such as OpenMPI or MPICH.
The mpi4py package.

Note

The mpi4py package is not a strict requirement. However, you need to have some way of calling MPI from Python. You also need some way of making sure that MPI_Init() is called when the IPython engines start up. There are a number of ways of doing this and a good number of associated subtleties. We highly recommend using mpi4py as it takes care of most of these problems. If you want to do something different, let us know and we can help you get started.

Starting the engines with MPI enabled#

To use code that calls MPI, there are typically two things that MPI requires.

The process that wants to call MPI must be started using mpiexec or a batch system (like PBS) that has MPI support.
Once the process starts, it must call MPI_Init().

There are a couple of ways that you can start the IPython engines and get these things to happen.

Automatic starting using mpiexec and ipcluster#

The easiest approach is to use the MPI Launcher, which will first start a controller and then a set of engines using mpiexec:

cluster = ipp.Cluster(engines="mpi")
cluster.start_cluster_sync()

or on the command-line

$ ipcluster start -n 4 --engines=mpi

Automatic starting using batch systems such as PBS or Slurm#

IPython Parallel also has launchers for several batch systems, including PBS, Slurm, SGE, LSF, HTCondor. Just like mpi, you can specify these as the controller

cluster = ipp.Cluster(engines="slurm", controller="slurm")

New in version 8.0: The controller and engines arguments are new in IPython Parallel 8.0. In 7.x, these arguments had to be called controller_launcher_class and engine_launcher_class, respectively.

Actually using MPI#

Once the engines are running with MPI enabled, you are ready to go. You can now call any code that uses MPI in the IPython engines. And, all of this can be done interactively. Here we show a simple example that uses mpi4py version 1.1.0 or later.

First, lets define a function that uses MPI to calculate the sum of a distributed array. Save the following text in a file called psum.py:

from mpi4py import MPI
import numpy as np

def psum(a):
    locsum = np.sum(a)
    rcvBuf = np.array(0.0, 'd')
    MPI.COMM_WORLD.Allreduce([locsum, MPI.DOUBLE],
        [rcvBuf, MPI.DOUBLE],
        op=MPI.SUM)
    return rcvBuf

Now, we can start an IPython cluster and use this function interactively. In this case, we create a distributed array and sum up all its elements in a distributed manner using our psum() function:

In [1]: import ipyparallel as ipp

In [2]: cluster = ipp.Cluster(engines="mpi", n=4)

In [3]: rc = cluster.start_and_connect_sync()

In [4]: view = rc[:]

In [5]: view.activate() # enable magics

# run the contents of the file on each engine:
In [6]: view.run('psum.py')

In [6]: view.scatter('a', np.arange(16,dtype='float'))

In [7]: view['a']
Out[7]: [array([ 0.,  1.,  2.,  3.]),
         array([ 4.,  5.,  6.,  7.]),
         array([  8.,   9.,  10.,  11.]),
         array([ 12.,  13.,  14.,  15.])]

In [8]: %px totalsum = psum(a)
Parallel execution on engines: [0,1,2,3]

In [9]: view['totalsum']
Out[9]: [120.0, 120.0, 120.0, 120.0]

Any Python code that makes calls to MPI can be used in this manner, including compiled C, C++ and Fortran libraries that have been exposed to Python.