[leinad]

I had a feeling that was the error and i was going to check it tonight. If it is constantly printing out the compiling data to the terminal it should now be working. Just remember this part takes long

[mvbakker]

Yes it looks like it is compiling, seems like an infinite stream of stuff it is doing. I'll let you know if it succeeds!

[leinad]

That sounds about right I will validate it as well when I get home. Once it is finished compiling I suggest the first thing you do is to back up an image of the Rasbian.
That way you dont have to do it again should any thing go wrong and you can then copy that image to other nodes so you only have to do it once

[mvbakker]

Hey Daniel,

It's is working fine and I already ran a simple Poiseuille simulation on the four cores. Right now I am trying to set up a cluster like you did with multiple RasPi2's. Unfortunately that is not going so smooth so far. I sort of got the openMPI working (well they do get some sort of a connection it seems) however I do get all sorts of errors a do not understand. Have you maybe got some tips on how you did it?

update1:
I think I got my openMPI running, (or maybe not), but I now get the same error for either running the icoFoam application using 'mpirun -np 8 -hostfile machines icoFoam -parallel > log.icoFoam' or 'foamJob -p -s icoFoam'. Namely I get this,

Parallel processing using SYSTEMOPENMPI with 8 processors
Executing: /usr/bin/mpirun -np 8 -hostfile machines -x FOAM_SETTINGS /home/pi/OpenFOAM/OpenFOAM-2.4.0/bin/foamExec -prefix /home/pi/OpenFOAM icoFoam -parallel | tee log
/*---------------------------------------------------------------------------*\
| ========= | |
| \\ / F ield | OpenFOAM: The Open Source CFD Toolbox |
| \\ / O peration | Version: 2.4.0 |
| \\ / A nd | Web: www.OpenFOAM.org |
| \\/ M anipulation | |
\*---------------------------------------------------------------------------*/
Build : 2.4.0-dcea1e13ff76
Exec : icoFoam -parallel
Date : Nov 08 2015
Time : 20:20:33
Host : "master"
PID : 2585
Case : /home/pi/FoamCases/ppWallGradParallel
nProcs : 8
Slaves :
7
(
"master.2586"
"master.2587"
"master.2588"
"slave1.2694"
"slave1.2695"
"slave1.2696"
"slave1.2697"
)

Pstream initialized with:
floatTransfer : 0
nProcsSimpleSum : 0
commsType : nonBlocking
polling iterations : 0
sigFpe : Enabling floating point exception trapping (FOAM_SIGFPE).
fileModificationChecking : Monitoring run-time modified files using timeStampMaster
allowSystemOperations : Allowing user-supplied system call operations

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

Create mesh for time = 0

[4]
[4]
[4] --> FOAM FATAL ERROR:
[4] Cannot find file "points" in directory "polyMesh" in times 0 down to constant
[4]
[4] From function [5]
[5]
[5] --> FOAM FATAL ERROR:
[5] Cannot find file "points" in directory "polyMesh" in times 0 down to constant
[5]
[5] From function Time::findInstance(const fileName&, const word&, const IOobject::readOption, const word&)
[5] in file db/Time/findInstance.C at line 203.
[5]
FOAM parallel run exiting
[5]
[6]
[6]
[6] --> FOAM FATAL ERROR:
[6] Cannot find file "points" in directory "polyMesh" in times 0 down to constant
[6]
[6] From function Time::findInstance(const fileName&, const word&, const IOobject::readOption, const word&)
[6] in file db/Time/findInstance.C at line 203.
[6]
FOAM parallel run exiting
[6]
Time::findInstance(const fileName&, const word&, const IOobject::readOption, const word&)
[4] in file db/Time/findInstance.C at line 203.
[4]
FOAM parallel run exiting
[4]
--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 6 in communicator MPI_COMM_WORLD
with errorcode 1.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------
[7]
[7]
[7] --> FOAM FATAL ERROR:
[7] Cannot find file "points" in directory "polyMesh" in times 0 down to constant
[7]
[7] From function Time::findInstance(const fileName&, const word&, const IOobject::readOption, const word&)
[7] in file db/Time/findInstance.C at line 203.
[7]
FOAM parallel run exiting
[7]
--------------------------------------------------------------------------
mpirun has exited due to process rank 6 with PID 2696 on
node slave1 exiting without calling "finalize". This may
have caused other processes in the application to be
terminated by signals sent by mpirun (as reported here).
--------------------------------------------------------------------------
[master:02580] 3 more processes have sent help message help-mpi-api.txt / mpi-abort
[master:02580] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages

[leinad]

Hi

Glad to see you got openFoam working, there are a couple of things that need to be sorted before this works. I had to figure this out the hard way.

1. When you source your bashrc file make sure its at the top(the very first line) of your bashrc file not at the bottom.

2. You will need a network set up between all the nodes looks like you have already done this. I ended up using static ip addresses because I had some issues with a DHCP server.

3. You will also need to set up password less ssh between all the nodes with a ssh key for each node on every node.

4. A nfs file sharing system is also a requirement, I made a run directory on the master node so that all the nodes can read and write to the case directory on the master node.

There are lots of tutorials online on how to accomplish each of these steps, I would like to put together a tutorial but I have been so busy lately I just dont get the time. But if you have any issues just post and I will try to help you.
Note: There are lots of different ways to set up the points above and different ways to set up a cluster what I have described above is just how I got it to work.

[mvbakker]

Oke sounds like I got point one up until three already set up!

I'll look into point four you gave me since I haven't set that up correctly yet. I was already wondering how that would work. Thanks for the tips!

This morning I finished the tower so once I get two RasPi's working together I'll extent it to a total of five