[slurm-users] siesta jobs with slurm, an issue

[Renfro, Michael]

You’re getting the same fundamental error in both the interactive and batch version, though.

The ‘reinit: Reading from standard input’ line seemed off, since you were providing an argument for the input file. But all the references I find to running Siesta in their manual (section 3 and section 16) show something more like:

  mpirun -np 4 /share/apps/chem/siesta-4.0.2/spar/siesta < dimer1prime.fdf > dimer1prime.out

and those examples line up with the idea that Siesta reads its commands from standard input, not by literally opening an input file specified as a command-line argument.

If the version of the command using < and > works correctly, then it’s definitely not a Slurm issue, it’s an issue with how you invoked Siesta.

> On Jul 22, 2018, at 12:11 PM, Mahmood Naderan <mahmood.nt at gmail.com> wrote:
> 
> I am able to directly run the command on the node. Please note in the following output that I have pressed ^C after some minutes. So, the errors are related to ^C.
> 
> 
> [mahmood at compute-0-3 ~]$ mpirun -np 4 /share/apps/chem/siesta-4.0.2/spar/siesta dimer1prime.fdf dimer1prime.out
> Siesta Version  : v4.0.2
> Architecture    : x86_64-unknown-linux-gnu--unknown
> Compiler version: GNU Fortran (GCC) 4.8.5 20150623 (Red Hat 4.8.5-16)
> Compiler flags  : mpifort -g -O2
> PP flags        : -DMPI -DFC_HAVE_FLUSH -DFC_HAVE_ABORT
> PARALLEL version
> 
> * Running on    4 nodes in parallel
> >> Start of run:  22-JUL-2018  21:39:41
> 
>                            ***********************
>                            *  WELCOME TO SIESTA  *
>                            ***********************
> 
> reinit: Reading from standard input
> ************************** Dump of input data file ****************************
> ^C--------------------------------------------------------------------------
> MPI_ABORT was invoked on rank 0 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.
> --------------------------------------------------------------------------
> No species found!!!
> Stopping Program from Node:    0
> ************************** End of input data file *****************************
> 
> reinit: -----------------------------------------------------------------------
> reinit: System Name:
> reinit: -----------------------------------------------------------------------
> reinit: System Label: siesta
> reinit: -----------------------------------------------------------------------
> 
> initatom: Reading input for the pseudopotentials and atomic orbitals ----------
> No species found!!!
> Stopping Program from Node:    0
> 
> 
> Regards,
> Mahmood
> 
> 
> 
> On Sun, Jul 22, 2018 at 8:44 PM, Bill Barth <bbarth at tacc.utexas.edu> wrote:
> That doesn't look like a slurm problem to me necessarily. Looks like SIESTA quit of its own volition (thus the call to MPI_ABORT()). I suggest you ask your local site support to take a look or go to the SIESTA developers. I doubt you'll find any SIESTA experts here to help you. 
> 
> All I can suggest is to check that all the paths you have provided SIESTA are correct (the path to the executable is clearly fine b/c SIESTA starts, but can it fine prime.fdf?). Otherwise start with your local support team.
> 
> Best,
> Bill.
> 
> -- 
> Bill Barth, Ph.D., Director, HPC
> bbarth at tacc.utexas.edu        |   Phone: (512) 232-7069
> Office: ROC 1.435            |   Fax:   (512) 475-9445
>