A File Cannot Be Found or Opened When mpirun Is Running on Multiple Nodes

Symptom

When you run the mpirun command on multiple nodes, an error message is displayed indicating that an .so file cannot be found or opened.

$ mpirun --allow-run-as-root -np 4 -N 2 --hostfile /hf2  /AllReduce
mpirun: error while loading shared libraries: libopen-rte.so.40: cannot open shared object file: No such file or directory

When you run the mpirun command on multiple nodes, an error message is displayed indicating that a file cannot be found or opened.

$ mpirun --allow-run-as-root -np 4 -N 2 --hostfile /hf2  /AllReduce
bash: /Hyper-MPI_x.x.x_aarch64_CentOS7.6_GCC9.3_MLNX-OFED5.0/ompi/bin/orted: No such file or directory

Possible Causes

An .so file cannot be found or opened.
The LD_LIBRARY_PATH environment variable has not been configured at the end of the bashrc file.
A file cannot be found or opened.
The Hyper MPI installation paths on the two nodes are different.

Procedure

An .so file cannot be found or opened.
1. Use PuTTY to log in to a job execution node as a common Hyper MPI user, for example, hmpi_user.
2. Check whether environment variables are correctly configured. For details, see Configuring Environment Variables.
A file cannot be found or opened.
1. Use PuTTY to log in to a job execution node as a common Hyper MPI user, for example, hmpi_user.
2. You are advised to install Hyper MPI in a mounted shared directory.

Parent topic: FAQ