We are having much the same problem as Jim Laur at AT&T Bell Labs was having
with the -208 errors, but on a different platform: AT&T 3B2/600G running
AT&T Unix System V 3.2.2. Our code is not written in C, but in 4GL instead.
We are having the problem with prepares randomly failing, etc., just as Jim
Laur did, and we already have hunted down each and every prepare to ensure we
have a close and free statement to go with it. It hasn't helped.
We also clearly have a resource allocation problem, and the idea of setting
SHMBASE to a different value was an exciting prospect to us.
One problem: there doesn't seem to be any way to set SHMBASE on this
platform; if there is a way, it isn't documented any where we can find it.
Can someone (maybe Jim Laur or Alan Denney) tell us how to modify SHMBASE on
our platform so we, too, can fix this horrible and insistent problem?
Bill Williams