> I've looked at the code fairly carefully and I don't see how this
could happen. Can you provide a self-contained test case that makes it
happen?
I would try to build a test program. Currently this is happening with my
production app on longevity run.
But still would like to know the reason to get such error. Any Postgres
document or other links would be helpful.
Is there any help from the stat stuff provided in the earliar mail?
Any fine tuning or changes in postgresql.conf file required?
My app makes 18 conncurrent connection with the Postgres server?
I was using periodic reindexing, but that too did'nt helped much, so I
have disbled it again. Jes doing VACCUM ANALYSE on every table after
every 10 mins.
BTW the source code was compiled on SunOS5.8 using Sun Forte's cc
compiler:
/configure --prefix=/usr/local/pgsql --enable-thread-safety
--without-readline CFLAGS='-D_REENTRANT' CPPFLAGS='-mt'
cc=/export/home/uxapps/workshop/SUNWspro/bin/cc
Does this helps?
Thanks,
Rajan Bhide