[molpro-user] MolPro Crashes in CCSD(t) routine

Andy May MayAJ1 at cardiff.ac.uk
Mon Aug 17 13:38:05 BST 2009


Neeraj.

The last message about Page Faults seems to indicate something is wrong
with the system, perhaps running low on memory:

http://en.wikipedia.org/wiki/Page_fault

Is this error reproducible?

The memory value passed to Molpro with -m option is for each process, so
for instance, with 8 processes and 15Gb this would be a maximum of
around 230 MW. With 4 processes it would be a maximum of 460 MW,
assuming you can ensure that no other jobs are running on the node using
up memory.

Best wishes,

Andy

Neeraj Rai wrote:
> Hello,
> 
>    I am running a CCSD(t) job but it crashes when it gets to the point
> of calculating triples. From the messages it seems like it is getting
> killed, probably trying to access more memory than requested by -m
> command.   I have cut and pasted last part of the output.. The node I am
> running it has 8 cores and 15Gb memory.  
> 
> uname -a
> Linux login2 2.6.16.60-0.39.3-smp #1 SMP Mon May 11 11:46:34 UTC 2009
> x86_64 x86_64 x86_64 GNU/Linux
> 
>  CCSD(T)     terms to be evaluated (factor= 1.000)
> 
> 
> 
>  Number of core orbitals:           5 (   5 )
>  Number of closed-shell orbitals:  16 (  16 )
>  Number of external orbitals:     369 ( 369 )
>  
>  Molecular orbitals read from record     2101.2  Type=RHF/CANONICAL
> (state 1.1)
> 
>  Number of N-1 electron functions:              16
>  Number of N-2 electron functions:             136
>  Number of singly external CSFs:              5904
>  Number of doubly external CSFs:          17431560
>  Total number of CSFs:                    17437465
> 
>  Length of J-op  integral file:               0.00 MB
>  Length of K-op  integral file:              10.47 MB
>  Length of 3-ext integral record:             0.00 MB
> 
>  Memory could be reduced to 972.8 Mword without degradation in triples
> tmp =
> /home/xe2/rain/pdir//soft/molpro/molpro2008/bin/molprop_2008_1_Linux_x86_6
> 4_i8.exe.p
>  Creating: host=cl1n208, user=rain,
>           
> file=/soft/molpro/molpro2008/bin/molprop_2008_1_Linux_x86_64_i8.exe,
> port=59536
>  Creating: host=cl1n208, user=rain,
>           
> file=/soft/molpro/molpro2008/bin/molprop_2008_1_Linux_x86_64_i8.exe,
> port=38807
>  Creating: host=cl1n208, user=rain,
>           
> file=/soft/molpro/molpro2008/bin/molprop_2008_1_Linux_x86_64_i8.exe,
> port=45928
>  Creating: host=cl1n208, user=rain,
>           
> file=/soft/molpro/molpro2008/bin/molprop_2008_1_Linux_x86_64_i8.exe,
> port=54610
>   4: interrupt(1)
> 3:SigIntHandler: interrupt signal was caught: 2
> 3:SigIntHandler: interrupt signal was caught: 2
> Last System Error Message from Task 3:: No such file or directory
>   3: ARMCI aborting 2 (0x2).
>   3: ARMCI aborting 2 (0x2).
> system error message: No such file or directory
> WaitAll: Child (20957) finished, status=0x9 (killed by signal 9).
> WaitAll: Child (20960) finished, status=0x100 (exited with code 1).
> WaitAll: Child (20959) finished, status=0x9 (killed by signal 9).
> WaitAll: No children or error in wait?
> 30190.06user 2843.76system 2:29:27elapsed 368%CPU (0avgtext+0avgdata
> 0maxresiden
> t)k
> 0inputs+0outputs (213major+3820962minor)pagefaults 0swaps
> 
> Could someone point me how to get around this problem or the only option
> is to find a m/c that has more memory to run these jobs?
> 
> Thanks.
> 
> 
> -- 
> Cheers,
> Neeraj.
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user



More information about the Molpro-user mailing list