FLUENT: What causes "Connection Refused" message when running in parallel?


When attempting to start FLUENT in parallel, a user may get a message similar to:

Host spawning Node 0 on machine "node1" (unix).
node1: Connection refused
Host unable to spawn process on machine "node1".

Where node1 is the name of the machine you are trying to spawn to.


This usually means that ssh or rsh is not working between all nodes in the cluster.

The MPICH (-pnmpi) communicator uses ssh or rsh depending on whether or not the -ssh flag is used on the command line. The Socket (-pnet) communicator uses rsh only.

You need to set up your cluster so that you can ssh or rsh between all nodes without requiring a password. The command:
rsh node1
or
ssh node1

should log you in to machine node1
If you are asked for a password, and you are using rsh, see:
man rhosts
or if you are using ssh, see:
man ssh-keygen

Contact your system administrator if you have problems.

Additional installation documentation can be found here:

<a target=_blank href="http://www.fluentusers.com/support/installation/unixfaq/index.htm">http://www.fluentusers.com/support/installation/unixfaq/index.htm</a>http://www.fluentusers.com/support/installation/unixfaq/index.htm





Show Form
No comments yet. Be the first to add a comment!