VASP hangs after NELM or LABORT
Posted: Tue Aug 09, 2016 6:51 am
Why would VASP hang after completing NELM electronic loops or receiving a hard stop (LABORT)?
I'm running a 128 atom cell with hybrid functionals (HSE06) and trying to calculate band structure. However, the cluster (SDSC Comet) has a short queue timeout of 48 hours. The job completes 2-3 electronic steps (to 2E-6) with ICHARG=11 within the first 24 hours, and it will acknowledge either LABORT=.TRUE. in STOPCAR or just NELM=2 in INCAR. But nothing happens after that. It does not write the PROCAR file. I read somewhere that hybrid functionals require self-consistent runs (ICHARG=1), but I've run many band structures with ICHARG=11 so far. Is it just the inclusion of too many k-points?
I'm using ISTART=1; INIWAV=1; ICHARG=11; LHFCALC=.TRUE.; HFSCREEN=0.2; PRECFOCK=Fast; LREAL=.TRUE.
ISMEAR=0; SIGMA=0.05; ALGO=Damped; TIME=0.5; LDIAG=.TRUE.; PREC=Accurate; LPLANE=.TRUE.
NSW=0; NELM=3; IBRION=-1; ENCUT=520; EDIFF=1E-8; LWAVE=.TRUE.; LCHARG=.TRUE.
CHGCAR and WAVECAR were converged to 1E-8 in previous HSE runs. I don't know why dE is so large in this run after being well converged previously.
KPOINTS contains:
3x3x3 plus 79: duplicate Gamma removed
83
Reciprocal lattice
0.00000000000000 0.00000000000000 0.00000000000000 1
0.33333333333333 0.00000000000000 0.00000000000000 8
0.33333333333333 0.33333333333333 0.00000000000000 6
-0.33333333333333 0.33333333333333 0.00000000000000 12
2.00000000 2.00000000 2.00000000 0.0
(...plus 78 more points with weight 0.)
STDERR/STDOUT ends with (last updated 19 hours ago, but job still "running"):
reading WAVECAR
the WAVECAR file was read successfully
charge-density read from file: unknown
entering main loop
N E dE d eps ncg rms ort
gam= 0.000 g(H,U,f)= 0.200E-05 0.392E-06 0.307E-13 ort(H,U,f) = 0.000E+00 0.000E+00 0.000E+00
SDA: 1 -0.709590074730E+03 -0.70959E+03 -0.11975E-05 31872 0.239E-05 0.000E+00
gam= 0.333 g(H,U,f)= 0.658E-06 0.267E-06 0.328E-13 ort(H,U,f) = 0.980E-06 0.324E-06 0.317E-13
DMP: 2 -0.709590075661E+03 -0.93139E-06 -0.68007E-06 31872 0.926E-06 0.130E-05
hard stop encountered! aborting job ...
OUTCAR ends with:
127 1.122 1.245 0.062 2.429
128 0.857 1.380 0.000 2.237
------------------------------------------------
tot 142.961 161.502 8.280 312.743
CHARGE: cpu time 5.0712: real time 5.0830
FORLOC: cpu time 0.0410: real time 0.0415
I'm running a 128 atom cell with hybrid functionals (HSE06) and trying to calculate band structure. However, the cluster (SDSC Comet) has a short queue timeout of 48 hours. The job completes 2-3 electronic steps (to 2E-6) with ICHARG=11 within the first 24 hours, and it will acknowledge either LABORT=.TRUE. in STOPCAR or just NELM=2 in INCAR. But nothing happens after that. It does not write the PROCAR file. I read somewhere that hybrid functionals require self-consistent runs (ICHARG=1), but I've run many band structures with ICHARG=11 so far. Is it just the inclusion of too many k-points?
I'm using ISTART=1; INIWAV=1; ICHARG=11; LHFCALC=.TRUE.; HFSCREEN=0.2; PRECFOCK=Fast; LREAL=.TRUE.
ISMEAR=0; SIGMA=0.05; ALGO=Damped; TIME=0.5; LDIAG=.TRUE.; PREC=Accurate; LPLANE=.TRUE.
NSW=0; NELM=3; IBRION=-1; ENCUT=520; EDIFF=1E-8; LWAVE=.TRUE.; LCHARG=.TRUE.
CHGCAR and WAVECAR were converged to 1E-8 in previous HSE runs. I don't know why dE is so large in this run after being well converged previously.
KPOINTS contains:
3x3x3 plus 79: duplicate Gamma removed
83
Reciprocal lattice
0.00000000000000 0.00000000000000 0.00000000000000 1
0.33333333333333 0.00000000000000 0.00000000000000 8
0.33333333333333 0.33333333333333 0.00000000000000 6
-0.33333333333333 0.33333333333333 0.00000000000000 12
2.00000000 2.00000000 2.00000000 0.0
(...plus 78 more points with weight 0.)
STDERR/STDOUT ends with (last updated 19 hours ago, but job still "running"):
reading WAVECAR
the WAVECAR file was read successfully
charge-density read from file: unknown
entering main loop
N E dE d eps ncg rms ort
gam= 0.000 g(H,U,f)= 0.200E-05 0.392E-06 0.307E-13 ort(H,U,f) = 0.000E+00 0.000E+00 0.000E+00
SDA: 1 -0.709590074730E+03 -0.70959E+03 -0.11975E-05 31872 0.239E-05 0.000E+00
gam= 0.333 g(H,U,f)= 0.658E-06 0.267E-06 0.328E-13 ort(H,U,f) = 0.980E-06 0.324E-06 0.317E-13
DMP: 2 -0.709590075661E+03 -0.93139E-06 -0.68007E-06 31872 0.926E-06 0.130E-05
hard stop encountered! aborting job ...
OUTCAR ends with:
127 1.122 1.245 0.062 2.429
128 0.857 1.380 0.000 2.237
------------------------------------------------
tot 142.961 161.502 8.280 312.743
CHARGE: cpu time 5.0712: real time 5.0830
FORLOC: cpu time 0.0410: real time 0.0415