insufficient virtual memory or low CPU usage

Message

cgzhou · #1 Post by **cgzhou** » Wed Aug 15, 2007 1:12 am

Dear All,

I am running VASP MD for calculating a 120 atoms system containing just C and H. The input parameters:

! INCAR

! electronic degrees of freedom
PREC = med
ISYM = 0 ! no symmetry
ISMEAR = 1 ; SIGMA = .2
IALGO = 48 ! algorithm to minimiz e
ISIF = 0 ! no stress calculations
!LREAL = O ! use real-space procectors (automatic)
LREAL=.TRUE.
NSIM = 8 ! use matrix-matrix operations

! ionic degree of freedom
NSW = 4000 ! number of steps
IBRION = 0 ! MD
POTIM = 1.0 ! 1 fs timestep
SMASS = 0.05 ! mass of heatbath
TEBEG = 300

MAXMIX = 40 ! rank of dielectric function is <40
IWAVPR = 10 ! TMPCAR is kept in RAM

LCHARG = F
NWRITE = 1

! end of INCAR

It is very sad that the program keeps eating my memory and finally lead to a crash:

-------------------------------------------
[blade@node19 c60h50]$ mpirun -np 2 ~/bin/vasp &
[1] 19622
[blade@node19 c60h50]$ running on 2 nodes
distr: one band on 1 nodes, 2 groups
vasp.4.6.28 25Jul05 complex
POSCAR found : 2 types and 110 ions
LDA part: xc-table for Ceperly-Alder, standard interpolation

-----------------------------------------------------------------------------
| |
| W W AA RRRRR N N II N N GGGG !!! |
| W W A A R R NN N II NN N G G !!! |
| W W A A R R N N N II N N N G !!! |
| W WW W AAAAAA RRRRR N N N II N N N G GGG ! |
| WW WW A A R R N NN II N NN G G |
| W W A A R R N N II N N GGGG !!! |
| |
| VASP found 327 degrees of freedom |
| the temperature will equal 2*E(kin)/ (degrees of freedom) |
| this differs from previous releases, where T was 2*E(kin)/(3 NIONS). |
| The new definition is more consistent |
| |
-----------------------------------------------------------------------------

POSCAR, INCAR and KPOINTS ok, starting setup
WARNING: wrap around errors must be expected
FFT: planning ... 7
reading WAVECAR
/usr/local/mpich/ch_p4/bin/mpirun: line 1: 19706 Killed
/home/blade/bin/vasp -p4pg /home/blade/mycal/c60h50/PI19622 -p4wd /home/blade/mycal/c60h50
p1_19715: p4_error: net_recv read: probable EOF on socket: 1
bm_list_19707: (148.332031) wakeup_slave: unable to interrupt slave 0 pid 19706

---------------------------------------------

I traced the memory usage and found that the memory usage keeps increasing. My workstation has
2 GB memory, and another PC has 4 GB memory. On either machine, the program always told me
"insufficient virtual memory" and then crashed. If I adjust the precision to "LOW", the program
does not crash but has a very very low CPU usage (maximum 0.5%). It is quite a headache.

I compiled VASP using ifort9+cmkl8 under RedHat AS 3. Could anyone help me solve the problem?

Thanks,

Chenggang Zhou

#2 Post by **admin** » Fri Aug 17, 2007 9:31 am

this rather looks like a bug in the installation of your MPI software.

cgzhou · #3 Post by **cgzhou** » Fri Aug 17, 2007 10:03 am

Hi Admin,

Do you have any suggestions? Thanks in advance!

#4 Post by **admin** » Tue Aug 21, 2007 9:31 am

please ask your system administrator or the person who installed the compiler and the parallelization software on your machines. We cannot give any hints on problems which are related to your local computing environment