Ansys Fluent 14.5 Distribution using Platform MPI on Windows workstations

Test Workstations Details

Node Hostnames: t1.test.com, t2.test.com, t3.test.com
Cores: 8 cores/node

Prerequisites

• Ansys Fluent 14.5 should be installed on all the workstation in the default installation location.
• Platform MPI Installation & service should be running on workstations (Installation Instructions can be found at “%AWP_ROOT145%\commonfiles\MPI\Platform\8.2.1\Windows\INSTALL_Platform-MPI_README.mht”)
• Connect all three workstations to AD or have identical username & password across all the workstations

Hostname File Preparation

Crete a text file called hostnames.txt on the user’s desktop
In my case, Each cores will be defined per line. You can also follow the Platform MPI hostfile definitions.
Cores Distribution Example : Fore cores from t1.test.com, fore cores from t2.test.com & all eight cores from t3.test.com


hostnames.txt

t1.test.com
t1.test.com
t1.test.com
t1.test.com
t2.test.com
t2.test.com
t2.test.com
t2.test.com
t3.test.com
t3.test.com
t3.test.com
t3.test.com
t3.test.com
t3.test.com
t3.test.com
t3.test.com

or 

t1.test.com:4
t2.test.com:4
t3.test.com:8

Launch Ansys Fluent 14.5 & select necessary options

ansys14.5-1

Click on Parallel Settings and do the following things

1. Click on Validate Platform MPI Password (for first time or alter password change)
2. Click on Distributed Memory on a Cluster -> File Containing Machine Names -> Browse the hostnames.txt file

ansys14.5-2
3. Click on OK button to start the job in distributed mode

4. Just skip the warning message & click on OK

ansys14.5-4

5. Now Ansys Fluent Main application will be launched. This will get all the required licenses & make mpi communication between selected workstations

ansys14.5-3

6. Now user can select an existing job / create a new job & submit for solving!

Advertisements

One thought on “Ansys Fluent 14.5 Distribution using Platform MPI on Windows workstations

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s