"Non-zero return code" after DMRG-SCF

[dmrg-learner]

Hi,

I tried to run a DMRG calculation and got this ending part

...
 Convergence after ... iterations
...
--- Stop Module: dmrgscf at ...
*** files: ...
    saved to directory ...
--- Module dmrgscf spent ---

.########################.
.# Non-zero return code #.
.########################.

I cannot release the molecule information. The active space is beyond the standard CASSCF limit and it works for small active space. Could it be memory too small? (I did not find notice for memory problem in the output.) If yes, where to set memory in the input?

Thank you very much

[dmrg-learner]

in a slurm file, it is said slurmstepd: error: Detected 1 oom_kill event in StepId=....batch. Some of the step tasks have been OOM Killed. seems out of memory. Is there any place in the input file to specify memory? besides #SBATCH --mem-per-cpu

[weymutht]

As far as I know, QCMaquis takes as much memory as it needs; there is no specific setting limiting it except for the option in your queueing system. I suggest you resubmit your job, requiring more memory from the queueing system.