NAME¶

ipcluster - start an IPython cluster

DESCRIPTION¶

Start an IPython cluster for parallel computing.

An IPython cluster consists of 1 controller and 1 or more engines. This command automates the startup of these processes using a wide range of startup methods (SSH, local processes, PBS, mpiexec, SGE, LSF, HTCondor, Slurm, Windows HPC Server 2008). To start a cluster with 4 engines on your local host simply do 'ipcluster start --n=4'. For more complex usage you will typically do 'ipython profile create mycluster --parallel', then edit configuration files, followed by 'ipcluster start --profile=mycluster --n=4'.

Subcommands -----------

Subcommands are launched as `ipcluster cmd [args]`. For information on using subcommand 'cmd', do: `ipcluster cmd -h`.

start

Start an IPython cluster for parallel computing

Start an ipython cluster by its profile name or cluster directory. Cluster directories contain configuration, log and security related files and are named using the convention 'profile_<name>' and should be creating using the 'start' subcommand of 'ipcluster'. If your cluster directory is in the cwd or the ipython directory, you can simply refer to it using its profile name, 'ipcluster start --n=4 --profile=<profile>`, otherwise use the 'profile-dir' option.

stop

Stop a running IPython cluster

Stop a running ipython cluster by its profile name or cluster directory. Cluster directories are named using the convention 'profile_<name>'. If your cluster directory is in the cwd or the ipython directory, you can simply refer to it using its profile name, 'ipcluster stop --profile=<profile>`, otherwise use the '--profile-dir' option.

engines

Start engines connected to an existing IPython cluster

Start one or more engines to connect to an existing Cluster by profile name or cluster directory. Cluster directories contain configuration, log and security related files and are named using the convention 'profile_<name>' and should be creating using the 'start' subcommand of 'ipcluster'. If your cluster directory is in the cwd or the ipython directory, you can simply refer to it using its profile name, 'ipcluster engines --n=4 --profile=<profile>`, otherwise use the 'profile-dir' option.

nbextension

Enable/disable IPython clusters tab in Jupyter notebook

for Jupyter Notebook >= 4.2, you can use the new nbextension API:

jupyter serverextension enable --py ipyparallel jupyter nbextension install --py ipyparallel jupyter nbextension enable --py ipyparallel

To see all available configurables, use --help-all.

EXAMPLES¶

ipcluster start --n=4 # start a 4 node cluster on localhost

ipcluster start -h # show the help string for the start subcmd

ipcluster stop -h # show the help string for the stop subcmd

ipcluster engines -h # show the help string for the engines subcmd