Well...
I Guess AOL Server and Linux themselves are not accused in this case, top shows a normal server behavior during the hang..
Heres the server threads settings any how:
ns_param maxconnections 20
ns_param maxdropped 0
ns_param maxthreads 10
ns_param minthreads 0
ns_param threadtimeout 120
This type of problem happens very rarely, and more clearly, when there is a big count of users (400+) using the web-end for Writing.
Scheduled Procedures run normally when the above condition disappears... Also, when users are purely reading from the web, Ive got no problem also.
My other OpenACS is accessing the other oracle application schemas. And during the hang, mostly, the locked tables are those of the other schemas.
A question out of this scope... Does anybody test Oracle 10G with OpenACS??
I am playing around to test the compatibility, 10G sounds great, but I am having some weird things also.
I dont exactly know why the new 10G enterprise manager shows NO locks list anymore...
Anyhow, I am a having a High CPU consumption by the nsd process after a while running AOL Server 4.01 Over 10G, Still cant figure why...
But Ill sure have more clues in while