Posted by Hirotoshi Mori on November 20, 2008 at 08:49:36:
Dear MOLCAS developers and users,
This is Hirotoshi Mori at Ochanomizu University, Japan.
Today, I'm posting this message since I have a problem related with disk usage.
Now, I'm trying to run seward for medium size molecule with big basis sets (c.a.600). When I run my job, I got the next error. (The version is MOLCAS 7.2)
===(Error message)===
I/O statistics
--------------
Unit Name Xtent pages MBytes pages MBytes
(MBytes) in in out out
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
2 ONEINT 11.23 .16 2.90 .04 25.99
11 RUNFILE 6.88 .01 .05 .01 .05
22 RYSRW 1.26 .03 1.25 .00 .00
40 ORDINT 1023.78 .00 .00 131.50 2043.85
50 TEMP01 512.00 .00 .00 131.50 1022.02
77 NQGRID .00 .00 .00 .00 .00
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Total .20 4.20 263.05 3091.90
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
*******************************************************************************
*******************************************************************************
*** ***
*** ***
*** Location: AixWr ***
*** File: ORDINT ***
*** ***
*** ***
*** Premature abort while writing buffer to disk: Disk full? ***
*** ***
*** ***
*******************************************************************************
*******************************************************************************
From the above error, it looks like my disk was full.
But, actually, I have enough Working directory space (250GB) on 64 bit IBM-AIX system (Of cause, I used -64 option in compilation).
I tried with large $MOLCASDISK value, but I got same message.
When I use MOLCAS on my 64bit Linux PC-cluster, there was no problem.
What is the problem?
Best regards,
Hirotoshi Mori