[aeroiitkgp]

Hello,

I have a grid size of 3 million cell volumes. It took just few seconds to partition using 24 cores, but took almost 15 mins to partition with 240 cores. My machine is a Linux cluster running redhat with 24 cores in each node. I have attached the configuration file

I also tried a37 million mesh with 480 processors which took 3 hours to partition. And it does partitioning for each case

Is there any way to save the partition file so that it can be used for other cases? Also can distance file for SA model be saved to be used in other cases? Kindly help me resolve this issuses.

[talbring]

Actually it shouldn't take that long for the partitioning ... are you sure your cluster is properly set up and there is a high-speed connection between the nodes (like e.g. infiniband)?

[aeroiitkgp]

Originally Posted by talbring

Actually it shouldn't take that long for the partitioning ... are you sure your cluster is properly set up and there is a high-speed connection between the nodes (like e.g. infiniband)?

Yes our cluster has an infiniband link. We regularly use parallel programs and they run very efficiently.

But I am not sure what is slowing down SU2 Partioning. If you can guide on what should I check in order to ensure that everything is ok with cluster or SU2, it would be useful.

Also is there any way to save partitioning file?