[molpro-user] gparam,lenbuf= ?
Benj FitzPatrick
benjfitz at uchicago.edu
Tue Jan 23 16:46:46 GMT 2007
Karel,
Did you compile global arrays with the large file option? I ran into this when
GA did not have large files enabled.
Cheers,
Benj FitzPatrick
University of Chicago
---- Original message ----
>Date: Tue, 23 Jan 2007 17:13:06 +0100
>From: Krapnik <krapnik at gmail.com>
>Subject: [molpro-user] gparam,lenbuf= ?
>To: molpro-user at molpro.net
>
> I have forgotten to attach the end of output file (Memory was set for
> 800 MW, disk was in maximum only 30% full):
>
> AO integral compression algorithm 1 Integral accuracy 1.0D-13
>
> 109655.622 MB (compressed) written to integral file ( 40.5%)
>
> Node minimum: 109655.622 MB, node maximum: 109655.622 MB
>
> NUMBER OF SORTED TWO-ELECTRON INTEGRALS:33128280906. BUFFER
> LENGTH: 32768
> NUMBER OF SEGMENTS: 189 SEGMENT LENGTH: 175999431 RECORD
> LENGTH: 524288
>
> Memory used in sort: 176.56 MW
>
> ERROR WRITING 32768 WORDS AT OFFSET34091302912. TO FILE24
> IMPLEMENTATION=eaf FILE HANDLE= 2 IERR= -2000
> ? Error
> ? I/O error
> ? The problem occurs in writew
>
> GA ERROR fehler on processor
> 0
> 0:0:fehler on processor 0:: 4925500
> 0: ARMCI aborting 4925500 (0x4b283c).
> tmp =
>
/home/berka/pdir//usr/local/chem/molpro/2006.1_amd64-tcgmsg-64/lib/molprop_2006_1_i8_x86_64_tcgmsg.exe.p
> Creating: host=cerny2, user=berka,
>
>
file=/usr/local/chem/molpro/2006.1_amd64-tcgmsg-64/lib/molprop_2006_1_i8_x86_64_tcgmsg.exe,
> port=38453
>
> --
> Karel "CaCO3" Berka
More information about the Molpro-user
mailing list