CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Running in parallel gives error (https://www.cfd-online.com/Forums/openfoam-solving/235329-running-parallel-gives-error.html)

lumpor April 11, 2021 08:00

Running in parallel gives error
 
I made code changes in the radiation library of OF2012, which give no error when running my case serially. However, when trying to run it in parallel, I get the following error when the simulation is supposed to start:


Code:

Rank 0 [Sun Apr 11 12:41:09 2021] [c2-0c2s15n3] Fatal error in MPI_Recv: Message truncated, error stack:
MPI_Recv(212)...........................: MPI_Recv(buf=0x7ffffffefabc, count=4, MPI_BYTE, src=2, tag=1, MPI_COMM_WORLD, status=0x7ffffffefa20) failed
MPIDI_CH3_PktHandler_EagerShortSend(376): Message from rank 2 and tag 1 truncated; 8 bytes received but buffer size is 4
srun: error: nid00575: task 0: Aborted
srun: Terminating job step 6781862.0
slurmstepd: error: *** STEP 6781862.0 ON nid00575 CANCELLED AT 2021-04-11T12:41:09 ***
srun: error: nid00575: tasks 1-31: Terminated
srun: Force Terminated job step 6781862.0


This dissapears when I comment out a line in the code that calls the function distributedTriSurface.findline(start, end, hitInfo). Calling the function from an empty triSurface gives no error as well.


My case has cyclic patches, for which I added preservePatches lines in the decomposeParDict file.


Does anyone know what the error means? Does it have anything to do with preparePar?


Thanks

lumpor April 14, 2021 09:18

I solved it. Turned out I had to use the returnReduce function for a variable I was using in my while loops.


All times are GMT -4. The time now is 21:21.