current right signature 0 != proper right signature


Gets Around
I've got an issue when run next job:
title "cubane CCSD/sto-3g hessian"

memory 3000 mb

scratch_dir /mnt/scratch

geometry
 symmetry C2v
 H                     2.05666470    -1.33213307     0.02407914
 C                     1.09615335    -0.79533448    -0.03911930
 H                     2.05666470     1.33213307     0.02407914
 C                     1.09615335     0.79533448    -0.03911930
 H                     0.00000000    -1.54000343    -2.07015649
 C                     0.00000000    -1.20256334    -1.02895254
 H                     0.00000000     1.54000343    -2.07015649
 C                     0.00000000     1.20256334    -1.02895254
 H                     0.00000000    -1.44960955     1.97883667
 C                     0.00000000    -0.78122663     1.10509614
 H                     0.00000000     1.44960955     1.97883667
 C                     0.00000000     0.78122663     1.10509614
 H                    -2.05666470    -1.33213307     0.02407914
 C                    -1.09615335    -0.79533448    -0.03911930
 H                    -2.05666470     1.33213307     0.02407914
 C                    -1.09615335     0.79533448    -0.03911930
end

basis spherical
 H library sto-3g
 C library sto-3g
end

scf
 direct
end

tce
 mkccsd
 2emet 1
 freeze atomic
end

mrccdata
 root 1
 nref 2
 22222222222222222222222222220
 22222222222222222222222222202
end

task tce freq

after first successful energy computation
Mk-MRCCSD (Hartrees) = -304.214919569704875

I've got the following:
MRCC tiling completed in             0.0            0.0
MA_verify_allocator_stuff: starting scan ...
stack block 'sorted MO coeffs', handle 126, address 0x721c7668:
	current right signature 0 != proper right signature 1431655765
stack block '', handle 125, address 0x721ceee8:
	current checksum 1914499648 != stored checksum 1914499739
stack block '', handle 125, address 0x721ceee8:
	current right signature 0 != proper right signature 1431655765

several lines skiped.............

stack block '', handle 125, address 0x721ceee8:
	current right signature 0 != proper right signature 1431655765
MA_verify_allocator_stuff: scan completed
 ------------------------------------------------------------------------
 gacratblock: not enough local mem                   0


Version of NWCHEM
Quote:

   source          = /opt/Nwchem-dev.revision26353-src.2014-10-22
nwchem branch = Development
nwchem revision = 26353
ga revision = 10508


Could anybody tell me why this input crashes ?
Thanks in advance.
Vladimir

Clicked A Few Times
We were able to perform Mk-MRCCSD(T) calculations for several geometries using 64 cores.
For comparisons the Mk-MRCCSD(T) energy for the first geometry is Mk-MRCCSD (Hartrees) = -304.214916805420671
In frequency calculations you have to be extra cautious since the active orbitals may change for various geometries (simply you have to analyze the output). In your runs you can also use simple "cas 2 2" definition of the model space.

Please use the following input

title "cubane CCSD/sto-3g hessian"

echo

start cubane


  1. memory 3000 mb
memory stack 200 mb heap 100 mb global 64 mb verify


  1. scratch_dir /mnt/scratch

geometry
symmetry C2v
H 2.05666470 -1.33213307 0.02407914
C 1.09615335 -0.79533448 -0.03911930
H 2.05666470 1.33213307 0.02407914
C 1.09615335 0.79533448 -0.03911930
H 0.00000000 -1.54000343 -2.07015649
C 0.00000000 -1.20256334 -1.02895254
H 0.00000000 1.54000343 -2.07015649
C 0.00000000 1.20256334 -1.02895254
H 0.00000000 -1.44960955 1.97883667
C 0.00000000 -0.78122663 1.10509614
H 0.00000000 1.44960955 1.97883667
C 0.00000000 0.78122663 1.10509614
H -2.05666470 -1.33213307 0.02407914
C -1.09615335 -0.79533448 -0.03911930
H -2.05666470 1.33213307 0.02407914
C -1.09615335 0.79533448 -0.03911930
end

basis spherical
H library sto-3g
C library sto-3g
end

scf
direct
thresh 1.0e-9
tol2e 1.0e-10
end

tce
mkccsd
2emet 1
freeze atomic
thresh 1.0d-4
maxiter 500
end

mrccdata
root 1
nref 2
22222222222222222222222222220
22222222222222222222222222202
end

task tce freq


Best regards,
Kiran & Karol

Gets Around
Hello Karol.

btw cas 2 2 works only if TARGETSYM is set, but according to the documentation
Quote:http://nwchemgit.github.io/index.php/Release66:TCE
On the other hand, in the frequency calculations, TARGETSYM must be none, since the finite displacements given in the course of frequency calculations will lift the spatial symmetry of the equilibrium geometry.
. So I can't use cas 2 2.

Gets Around
I try another hardware, still no luck:
Quote:username
MRCC tiling completed in 0.0 0.0

Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:
  1. 0 0x7F1A794427D7
  2. 1 0x7F1A79442DDE
  3. 2 0x7F1A78D93C2F
  4. 3 0x2BDBD77 in list_verify at ma.c:?
  5. 4 0x2BDC8F8 in MA_verify_allocator_stuff
  6. 5 0x2BDE5B5 in MA_sizeof
  7. 6 0xBCEAE0 in rtdb_broadcast at rtdb.c:?
  8. 7 0xBCF1E1 in rtdb_get
  9. 8 0xBCE371 in rtdb_get_
  10. 9 0xA8108B in scf_get_fock_param_ at scf_get_info.F:366
  11. 10 0x177B98B in tce_ao1e_fock2e_
  12. 11 0x16FB59A in tce_mrcc_energy_
  13. 12 0x16E012C in tce_energy_fragment_ at tce_energy_fragment.F:92
  14. 13 0x63F028 in task_energy_doit_ at task_energy.F:277
  15. 14 0x640511 in task_energy_ at task_energy.F:111
  16. 15 0x64A331 in num_grad_ at task_num_grad.F:282
  17. 16 0x64C8EE in task_num_grad_ at task_num_grad.F:85
  18. 17 0x640FD8 in task_gradient_doit_ at task_gradient.F:383
  19. 18 0x642219 in task_gradient_ at task_gradient.F:110
  20. 19 0x7207EA in stpr_gen_hess_at_ at stpr_gh_at.F:520
  21. 20 0x6443BF in task_hessian_doit_ at task_hessian.F:254 (discriminator 1)
  22. 21 0x6446CE in task_hessian_ at task_hessian.F:80
  23. 22 0x643937 in task_freq_ at task_freq.F:83
  24. 23 0x6331B6 in task_ at task.F:390
  25. 24 0x62BB29 in nwchem at nwchem.F:283
Segmentation fault


Forum >> NWChem's corner >> Running NWChem