Calculation keep using swap file
 

Hello all,

Short description of my problem. I have a case running in parallel on a Linux cluster - 8 nodes w/ 2 Gb memory on each node. The case has approx. 3.3 million tets.

When running the case on the nodes only about 1.0-1.4 Gb is used on each node and a swap memory of 180-250Mb is created - one node only uses 30Mb though. Does anybody have an idea as to why the nodes are not using the full potential of the available memory? There is no other processes running on the nodes.

I suspect the swap to be the reason for my long calculation times.

I have started the simulation from the command line using the -s option for increased memory estimation.

Regards,
Per