How can I use the MATLAB Parallel Server products with the LSF scheduler in a nonshared file system?

4 views (last 30 days)
I would like to know if it is possible to configure MATLAB Parallel Server and Parallel Computing Toolbox to work with the Platform LSF scheduler when the data is stored in a fire space that is not accessible to the cluster. This setup may also involve running PCT from a different type of operating system than the cluster.

Accepted Answer

MathWorks Support Team
MathWorks Support Team on 25 Mar 2019
Edited: MathWorks Support Team on 25 Mar 2019
The LSF configurations with Parallel Computing Toolbox are only supported for use when:
- The MATLAB client operating system is the same type as the cluster (Windows to Windows, or Unix to Unix)
- The DataLocation set in the configuration is accessible by all computers from the same path (shared file system)
If the configuration does not meet these requirements, the LSF configurations in PCT will not work.
This limitation can be resolved by using the generic scheduler interface. The instructions and some sample scripts for integration with 3rd party schedulers can be found under the $MATLABROOT/toolbox/distcomp/examples/integration/lsf/nonshared directory.
Depending on your network and cluster configuration, you might need to modify these files before they will work in your setting. These configurations require that you have passwordless SSH/RSH and SCP/FTP setup on your client to the head node in the cluster. If you need assistance to setup passwordless remote access, ask your system administrator for help.
  • NOTE: Starting in R2019a the following name changes occurred:
  • MATLAB Distributed Computing Server was renamed to MATLAB Parallel Server
  • mdce_def was renamed to mjs_def
  • mdce binary was renamed to mjs

More Answers (0)

Categories

Find more on Startup and Shutdown in Help Center and File Exchange

Tags

No tags entered yet.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!