Hello group, I have a big problem. I can't ssh to nodes after they boot up brom PXE i get: /bin/bash: Permission denied Connection to node01 closed. Any idea why is this happening?? I did the mrsh ins

Hello group! I was wondering, is there any way to change the provision port of warewulf with its httpd? I know about defining virtualhosts, but how do I create them and how my nodes would find the new

Hello!! I have a cluster runnig just like the one in the install guide https://github.com/openhpc/ohpc/releases/download/v1.3.5.GA/Install_guide-CentOS7-Warewulf-SLURM-1.3.5-x86_64.pdf With 1 master n

Hello group! I was wondering, for a cluster in a university, that is being use to run OpenFOAM and WRF models (WRF running as an operational service, two times a day) WRF jobs HAVE TO preemt every oth

Hello group! If I turn off my cluster, using openhpc 1.3.3 .... the jobs that are in the queue get erased? Or they will still be there on power on? Thanks to all for the help!

And how could I block the execution of jobs in the master node that are OpenMP programs for example, that don't make use of MPI and ssh for computing... Is there a way? 2018-04-23 14:51 GMT-03:00 Alex

Hello group! I was wondering, is there anyway to block the execution of commands like "mpirun -np 4 -hosts node1,node2" in my cluster runnning OpenHPC 1.3.3 I want to make sure that every job is sent

Adrian!! I'm trying to build your guide: https://opensource.com/article/18/1/how-build-hpc-system-raspberry-pi-and-openhpc#comment-152896 How did you do this???

Has anyone find out a way to run CentOS 7 on RP3? I want to build a small cluster based on them and can't find a way to load Centos on aarch64

Yup! You were right! Thank you very much for all your help Karl 2018-03-23 14:01 GMT-03:00 Karl W. Schulz <karl@...>:

You know I realized that, so I added the folder on the master in /etc/exports with the others: /home *(rw,no_subtree_check,fsid=10,no_root_squash) /opt/ohpc/pub *(ro,no_subtree_check,fsid=11) /opt/int

I had a problem when I tried to run a test domain in WRF on my cluster. [prun] Error: Expected Job launcher mpiexec.hydra not found for impi So I shh to my node, and try to run the command myself (aft

Hello! I have a problem. My master node (also a compute node) is the only one with internet connection. And I have 1 compute node (c1). I want to run a job in using both nodes, but the first part (a m

Thanks Karl! It is working now. You were right

Hello group! I have a problem with OpenFOAM running over sbatch.... Whenever I try to run the cavity example over 16 cores with 2 nodes from console, so that would be doing: module load openfoam srun

The problem is when I check the IB traffic with "watch ifconfig ib0" I get: Every 2,0s: ifconfig ib0 Tue Jan 9 10:35:09 2018 Infiniband hardware address can be incorrect! Please read BUGS section in i

Thanks, when I run the Benchmark I obtained: prun IMB-MPI1 PingPong [prun] Master compute host = c1 [prun] Resource manager = slurm [prun] Launch cmd = mpiexec.hydra -bootstrap slurm IMB-MPI1 PingPong

Hello!! Thanks so much for the responses. What I want to do is use MPI over Infiniband (sorry for the misunderstanding) But I use Intel MPI, not OpenMPI or MVAPICH. Intel MPI is not supported over IB?

Hello group! I have just installed a new Infiniband connection. After some troubleshooting I can now ping between them and everything. ############################################################ ibst

Hello!! I was wondering... How can I make my env modules to force swap between them? For example: If I have loaded gnu 7.2 and I want to load intel 18 compiler... I want to do # module load intel ....