[bolbol]

Currently I'm using openfoam6 on Ubuntu 16 / virtualbox
The dedicated ram for the system is 12GB and when I tried to run a study case using pimpleFoam, only 1 GB is consumed along with 1 core from the CPU.
Is there a way to increase the resources used to speed up the process?
>> attached the resources consumption

[agustinvo]

Hi


I'm not a Virtualbox user, but you should modify your VM settings to use more RAM or CPU.



Regards

[bolbol]

Originally Posted by agustinvo

Hi


I'm not a Virtualbox user, but you should modify your VM settings to use more RAM or CPU.



Regards

I did that already, you can check the attachment above.
There are 5 cores and 11GB ram not used

[agustinvo]

Hi, I was using my smartphone and I didn't check properly.


The RAM usage is linked to the needed memory to allocate mesh, fields... It is what it is. About the CPU... are you using:

pimpleFoam // one processor in terminal

or

foamJob pimpleFoam // one processor in background

or

 decomposePar // decompose your mesh and fields
              // to be used by the different processors
 
 foamJob -p pimpleFoam // parallel simulation

to run your simulation?

[bolbol]

Originally Posted by agustinvo

Hi, I was using my smartphone and I didn't check properly.


The RAM usage is linked to the needed memory to allocate mesh, fields... It is what it is. About the CPU... are you using:

pimpleFoam // one processor in terminal

or

foamJob pimpleFoam // one processor in background

or

 decomposePar // decompose your mesh and fields
              // to be used by the different processors
 
 foamJob -p pimpleFoam // parallel simulation

to run your simulation?

I compiled my own solver actually but it's based on pimpleFoam
And I just write pimpleParticleFoam.

Is all I have to do is writing "foamJob -p pimpleParticleFoam" or there are some setup required before?

=======


I tried that command and it worked already.
"mpirun -np 4 simpleFoam -parallel"


Thank you for you hints

[agustinvo]

Just remember to decompose your simulation before running. In the corresponding decomposeParDict you can select the number of processors and the decomposition method.

After that you can use mpi or foamJob.