Hello everyone. Over the past 2 weeks we had 3 server crashes.
The dsmserv.err file contained the following entry for each crash:

07/28/1999 01:14:38  ANR7838S Server operation terminated.
07/28/1999 01:14:38  ANR7837S Internal error BUF006 detected.
07/28/1999 01:14:38  ANR7833S ADSM thread 1 terminated in response to
program aborted.
07/28/1999 01:14:38  ANR7833S ADSM thread 2 terminated in response to
program aborted.
.
.
According to ADSM, this is due to low buffer pool. Okay so we increased
the buffer pool from 32MB to 48MB per their recommendation. Again
another
crash. Finally I decided to increase the buffer pool to 128MB. It has
been
3 days know and the server is still up.

Yesterday after the administrative processes are completed, I changed
the
buffer pool to 32MB and ran a complete test backup using all drives.
Server did
not crash.  I lowered yet the buffer pool to 16MB and ran another test
backup
using all drives. Server did not crash.  So whats up?

ADSM development is currently working on a PTF to fix the problem.
However,
it will not be available until next year, as I was told. Oh well.  I
just hope the
128MB buffer pool is enough. What makes me wonder is why should the
server
crash if its low in buffer pool(cache) where it could use disk? Any idea
anyone?

I feel so aloooone.

Angel