[molpro-user] failure of qcisd optg
Andy May
MayAJ1 at cardiff.ac.uk
Wed Feb 27 17:54:59 GMT 2013
Shachar,
The input you sent takes a long time to run so I cut down the basis set
and the job failed in the same location (although with a different error
reported).
I found that removing 'gdirect' allowed the program to run correctly,
perhaps you can try to see if this works for you.
I've opened a bug report about the crash:
https://www.molpro.net/bugzilla/show_bug.cgi?id=4007
Best wishes,
Andy
On 25/02/13 15:44, Shachar Klaiman wrote:
> Dear Andy,
>
> Please find attached the input and output files as well as the pbs log
> with the error I get.
>
> Thanks for you help.
>
> Best regards,
> Shachar
>
>
>
>
> On Sat, Feb 23, 2013 at 12:09 AM, Andy May <MayAJ1 at cardiff.ac.uk
> <mailto:MayAJ1 at cardiff.ac.uk>> wrote:
>
> Shachar,
>
> If you can send the input and output files then I'll see if I can
> reproduce the error.
>
> Best wishes,
>
> Andy
>
>
> On 15/02/13 08:13, Shachar Klaiman wrote:
>
> Dear Molpro users,
>
> I am trying to perform a qcisd optimization but after the
> initial QCISD
> calculation the program crashes with a segmentation fault. I am
> using
> Molpro 2012 parallel installation with intel compiler 12.0.5
> impi 4.0.2
> and MKL 10.3.5 . All the tests pass.
>
> The error only appears in the PBS log:
>
> forrtl: severe (174): SIGSEGV, segmentation fault occurred
> Image PC Routine Line
> Source
> libmpi.so.4 00002AAAAACC63A0 Unknown
> Unknown Unknown
> molpro.exe 00000000018FA41D Unknown
> Unknown Unknown
> molpro.exe 00000000018EBEA8 Unknown
> Unknown Unknown
> molpro.exe 000000000046DEB5 Unknown
> Unknown Unknown
> molpro.exe 0000000000421226 Unknown
> Unknown Unknown
> molpro.exe 0000000001D0D99B Unknown
> Unknown Unknown
> molpro.exe 0000000001D0B0C3 Unknown
> Unknown Unknown
> molpro.exe 0000000001E07524 Unknown
> Unknown Unknown
> molpro.exe 0000000001E5DF2D Unknown
> Unknown Unknown
> molpro.exe 0000000001E52D44 Unknown
> Unknown Unknown
> molpro.exe 0000000001E509F6 Unknown
> Unknown Unknown
> molpro.exe 00000000005B5DD9 Unknown
> Unknown Unknown
> molpro.exe 0000000000413506 Unknown
> Unknown Unknown
> molpro.exe 0000000000411BA4 Unknown
> Unknown Unknown
> molpro.exe 00000000004A5B4F Unknown
> Unknown Unknown
> molpro.exe 000000000040EAC1 Unknown
> Unknown Unknown
> molpro.exe 000000000040EA5C Unknown
> Unknown Unknown
> libc.so.6 0000003A3041D994 Unknown
> Unknown Unknown
> molpro.exe 000000000040E969 Unknown
> Unknown Unknown
> forrtl: severe (174): SIGSEGV, segmentation fault occurred
>
> ------------------------------__--------
> the last part of the molpro output:
>
>
> RESULTS
> =======
>
> Reference energy -339.599577245823
> QCISD singlet pair energy -0.935891040246
> QCISD triplet pair energy -0.464332775793
> QCISD correlation energy -1.400223774967
>
> !QCISD total energy -340.999801020791
>
> Timing summary (sec):
>
> STEP CPU(USER) SYS CPU(TOT) WALL
> Transformation 104.43 11.04 115.47 189.47
> CCSD iterations 4424.35 228.90 4653.25 4847.98
>
> Program statistics:
>
> Available memory in ccsd: 199998300
> Min. memory needed in ccsd: 106382167
> Max. memory used in ccsd: 149342907
> Max. memory used in cckext: 110566686 (17
> integral passes)
>
> Starting CP-QCISD iterations. Author: G. Rauhut (1999)
>
> Loop Var(Pairs) Var(Singles) Time/Loop Tot-Time DIIS
> 1 0.37D+00 0.44D-10 4.2 677.6 0 0
> 2 0.67D-02 0.14D-01 1700.1 2377.6 1 1
> 3 0.11D-02 0.36D-03 1701.4 4079.0 2 2
> 4 0.16D-03 0.67D-03 1712.8 5791.8 3 3
> 5 0.32D-04 0.12D-03 1702.1 7493.9 4 4
> 6 0.24D-05 0.51D-04 1707.6 9201.5 5 5
> 7 0.14D-05 0.15D-04 1704.7 10906.3 6 6
> 8 0.29D-06 0.29D-05 1705.8 12612.0 6 2
> 9 0.88D-07 0.42D-06 1707.0 14319.1 6 1
> 10 0.10D-07 0.64D-07 1706.5 16025.6 6 4
> 11 0.23D-08 0.11D-07 1707.9 17733.5 6 3
> 12 0.34D-09 0.21D-08 1705.5 19438.9 6 6
> 13 0.95D-10 0.33D-09 1704.1 21143.0 6 5
> 14 0.15D-10 0.70D-10 1708.4 22851.4 6 2
>
> Time needed for computing effective density matrices:
> 1713.8 sec.
>
> ------
>
> and then it stops.
>
> Any ideas?
>
> Thanks for your help.
>
> Best,
> Shachar
>
>
>
> _________________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net <mailto:Molpro-user at molpro.net>
> http://www.molpro.net/mailman/__listinfo/molpro-user
> <http://www.molpro.net/mailman/listinfo/molpro-user>
>
>
More information about the Molpro-user
mailing list