Installing MOOSE with OpenMPI3.1.3 and GCC 7 #16544
-
|
Hello, I followed the procedures of the Website: unset PETSC_DIR PETSC_ARCH cd petsc/ cd .. cd ~/projects/moose/test I was able to build PETSC and LibMesh, but when it comes to run the tests, I got the following errors:
|
Beta Was this translation helpful? Give feedback.
Replies: 5 comments 4 replies
-
|
There should be some more detailed error information farther up in your terminal window. I suspect there is a problem with the MPI. |
Beta Was this translation helpful? Give feedback.
-
|
@jinca I suggest you try the manual installation with gcc 9.2.0 and mpich 3.3 on the website Nicolò |
Beta Was this translation helpful? Give feedback.
-
|
The OS is Scientific Linux release 7.9 (Nitrogen) |
Beta Was this translation helpful? Give feedback.
-
|
50 Errors (the MAX before it quits) usually means an easy fix. Something fundamentally wrong. Not necessarily a compiler version or the like. As Andrew asked, seeing the actual error helps tremendously. The details of these errors can be found if you scroll up in your terminal history. |
Beta Was this translation helpful? Give feedback.
-
|
OK, When I set the following modules: I got this error while building petsc:
|
Beta Was this translation helpful? Give feedback.
-
|
Without PETSc, you wouldn't (shouldn't) be able to build libMesh. Without libMesh... you wouldn't be able to build MOOSE for an attempt at Its hard to tell if you copied exactly what you attempted, or if there is an error during your attempt: which mpicc
cd
mkdir projects # <-------- failed to enter projects before cloning moose
git clone https://github.com/idaholab/moose
cd moose
git checkout master
cd ~/projects/moose # <-------- not the moose you just cloned. A possible older version of moose? |
Beta Was this translation helpful? Give feedback.
-
|
Thank you so much for the quick replies guys. Many days ago, I was able to run MOOSE only with 7 FAILS. Now I am trying to reinstalling again, I can not install at least PETSC. I am very aware that I have to build PETSC, then LibMesh and at the end MOOSE. |
Beta Was this translation helpful? Give feedback.
-
|
Hi guys, this is the configuration I used: The PETSC was built as the libMesh, and therefore MOOSE. mesh/checkpoint.test_8 .................................................. [insufficient slots,min_cpus=8] SKIP
|
Beta Was this translation helpful? Give feedback.
-
|
Wrap three ticks around text when wanting to use a pre-rendered block (instead of one). Somehow, doing that entire thing around one broke GitHub! As I only got most of it, from email. tick tick tick tick tick tick Here is an excerpt of what was sent to me by email: The issue is with srun: Whatever this is, it prevented the following command from running: Which means, there would be no exodiff file generated, which would spawn the error: The same is happening with every error posted ( |
Beta Was this translation helpful? Give feedback.
Hi guys, this is the configuration I used:
module purge
module load slurm
module load dot
module load turbovnc/2.0.1
module load vgl/2.5.1/64
module load singularity/current
module load rhel7/global
module load cmake/latest
module load gcc/7
module load openmpi-3.1.3-gcc-7.2.0-b5ihosk
module load python/3.8
The PETSC was built as the libMesh, and therefore MOOSE.
I am attaching the output of the tests I've done recently.
Hope you can help me soon.
Best,
Julita Inca
mesh/checkpoint.test_8 .................................................. [insufficient slots,min_cpus=8] SKIP
mesh/checkpoint.test_8a ................................................. [insufficient slots,min_cpus=8] SKIP
parti…