Failure of SAPT0 -job

1

Dear Psi4 developers,

I am trying to do SAPT0 for a complex of 62 atoms with jun-cc-pvtz. The dimer calculations are completed and then after it saves the occupied orbitals to File 180, it dies. I have checked the 190 File is generated in the scratch directory. I have also tested with increase in memory till 50 gb. I have performed SAPT0 on system slightly smaller than this with 2074 basis functions ( jun-cc-pvtz on another molecule) which gives me perfect results. Can you guys let me know whats the problem? and how can I make it work.

Input:

memory 8 Gb

molecule boron_complex_cl_anion {
0 1
C -3.775755 -1.997644 -0.057577
C -3.423139 -0.675398 -0.340032
C -4.303141 0.331916 0.054094
H -2.430740 -0.430271 -0.703398
C -3.885668 1.712927 0.141181
C -1.221050 4.187390 -0.006013
C -1.209947 5.559005 -0.296685
C -0.000005 3.521685 0.159377
C -0.000003 6.235331 -0.426975
H -2.152738 6.087030 -0.444065
C 1.221042 4.187387 -0.006016
H -0.000008 2.466604 0.435703
C 1.209941 5.559003 -0.296687
H -0.000002 7.300849 -0.660419
H 2.152732 6.087026 -0.444068
C 2.785054 -3.048579 -0.062635
C 3.775758 -1.997643 -0.057590
C 3.423145 -0.675396 -0.340047
C 4.303144 0.331918 0.054089
H 2.430749 -0.430261 -0.703418
C -2.785049 -3.048578 -0.062625
C 3.885668 1.712928 0.141173
N 1.443456 -2.772969 -0.166347
H 1.044423 -1.830021 -0.237935
C 0.726236 -3.923455 -0.082918
C 1.631881 -4.981434 0.060642
H 1.363159 -6.029566 0.150085
C 2.923583 -4.428051 0.079999
H 3.873216 -4.943833 0.181135
C -0.726232 -3.923455 -0.082924
C -1.631878 -4.981441 0.060563
H -1.363159 -6.029579 0.149939
C -2.923581 -4.428057 0.079933
H -3.873216 -4.943845 0.181016
C -4.647974 2.844120 0.421590
H -5.715646 2.830764 0.616394
C -3.777528 3.947042 0.400512
H -4.030123 4.984253 0.599642
C -2.494994 3.465860 0.116372
C 2.494989 3.465857 0.116361
C 3.777530 3.947048 0.400459
H 4.030126 4.984264 0.599560
C 4.647978 2.844128 0.421537
H 5.715655 2.830778 0.616314
N 2.583750 2.117921 -0.036363
H 1.792912 1.502704 -0.257876
N -2.583751 2.117926 -0.036359
H -1.792916 1.502718 -0.257909
N -1.443449 -2.772963 -0.166296
H -1.044413 -1.830008 -0.237804
O -5.517082 0.067487 0.434904
O -4.978475 -2.323537 0.305703
O 5.517080 0.067489 0.434913
O 4.978473 -2.323533 0.305709
B -6.073862 -1.299683 0.124881
B 6.073863 -1.299680 0.124893
F 6.473268 -1.319000 -1.197359
F 7.079793 -1.574495 1.013640
F -6.473261 -1.319002 -1.197373
F -7.079798 -1.574497 1.013623

-1 1
Cl -0.000004 0.104178 -0.592457
units angstrom
no_reorient
symmetry c1
no_com
}

set globals {
basis jun-cc-pvtz
df_basis_scf jun-cc-pvtz-jkfit
df_basis_sapt jun-cc-pvtz-ri
scf_type df
FREEZE_CORE true
guess SAD
}

energy(‘sapt0’)

output:

Final Occupation by Irrep:
A
DOCC [ 155 ]

Energy converged.

@RHF Final Energy: -2495.94772074361117

=> Energetics <=

Nuclear Repulsion Energy =           5125.3028602390031665
One-Electron Energy =              -13663.1890408569979627
Two-Electron Energy =                6041.9384598725991964
DFT Exchange-Correlation Energy =       0.0000000000000000
Empirical Dispersion Energy =           0.0000000000000000
Total Energy =                      -2495.9477207453965093

Properties will be evaluated at 0.000000, 0.000000, 0.000000 Bohr
==> Properties <==

Properties computed using the SCF density density matrix
Nuclear Dipole Moment: (a.u.)
X: 0.0000 Y: 21.4296 Z: 0.2729

Electronic Dipole Moment: (a.u.)
X: -0.0000 Y: -19.2875 Z: -0.0733

Dipole Moment: (a.u.)
X: 0.0000 Y: 2.1421 Z: 0.1996 Total: 2.1514

Dipole Moment: (Debye)
X: 0.0000 Y: 5.4447 Z: 0.5074 Total: 5.4683

Saving occupied orbitals to File 180.

Saving SAPT 2-DIMER file.

Thanks in advance!

Arka

2

So sapt calcs go: dimer scf, monoA scf, monoB scf, sapt0. Are you saying your calc dies just after dimer scf? I’m running your file and have made it into the sapt0 section.

The first time I ran it, it did fail fairly quickly but that was because I hadn’t set PSI_SCRATCH and the default scratch /tmp/ filled up. Any chance that could be happening to you?

If I don’t need the computer, I’ll let this run and send the result.

3

I wrote too soon. The job did fail but well into the sapt0 stage. It failed because this computer only has a small 230 GB scratch drive, and the job filled it. So I won’t be sending you an output file. Beware disk management on this job. :slight_smile:

>>> tail -f output.dat 
    NSO        =      1933
    NMO        =      1932
    NRI        =      4752
    NOCC A     =       146
    NOCC B     =         9
    FOCC A     =        42
    FOCC B     =         5
    NVIR A     =      1786
    NVIR B     =      1923

PSIO_ERROR: unit = 195, errval = 12
PSIO_ERROR: 12 (error writing to file)
An error has occurred python-side. Traceback (most recent call last):

  File "<string>", line 140, in <module>

  File "../../share//python/driver.py", line 444, in energy
    wfn = procedures['energy'][lowername](lowername, molecule=molecule, **kwargs)

  File "../../share//python/procedures/proc.py", line 2851, in run_sapt
    e_sapt = psi4.sapt(dimer_wfn, monomerA_wfn, monomerB_wfn)

RuntimeError: 
Fatal Error: PSIO Error
Error occurred in file: /shortened/psi4/src/lib/libpsio/error.cc on line: 128
The most recent 5 function calls were:

psi::PsiException::PsiException(std::string, char const*, int)
psi::psio_error(unsigned int, unsigned int)
psi::PSIO::rw(unsigned int, char*, psi::psio_address, unsigned long, int)
psi::sapt::SAPT::zero_disk(int, char const*, int, int)
psi::sapt::SAPT0::df_integrals()
4

Thanks for getting back!
the job dies with a message: Saving SAPT 2-DIMER file

I have set up scratch directory other than the default /tmp and mine one has more than 2TB of disk space.

5

I ran the job in a bigger cluster machine, and it worked! Memory must have been the problem in previous trials. Thanks!