parallel (cluster vs local) and UDFs
Hi,
Is running parallel on a cluster of different machines the same as running parallel on a local machine? I am asking this based from the point of view of compiling UDFs. I have not been able to load my library after compilation; getting the following messages: Opening library "libudf"... Primitive Error at Node 0: open_udf_library: The operation completed successfully. Primitive Error at Node 1: open_udf_library: The system cannot find the file specified. Primitive Error at Node 2: open_udf_library: The system cannot find the file specified. Primitive Error at Node 3: open_udf_library: The system cannot find the file specified. Error: open_udf_library: The operation completed successfully. Error Object: #f All helps are deeply appreciated. Thank you. Sincerely, EH |
Re:parallel (cluster vs local) and UDFs
You need to copy the compiled library folders to all the nodes on your cluster.
|
Thank you for your reply. What do you mean by nodes on your cluster? Because at the moment, I only selected parallel in the FLUENT Launch window but did not do any partitioning to actually solve the problem in parallel.
|
In the start-up window, you have a Tab - Parallel Setting, the run type defaults to shared memory, but in your case this will not work of course, you have to work with the distributed memory. However, I do not have experience with it, so good luck :)
|
All times are GMT -4. The time now is 12:49. |