The Question is:
I have a relatively large image that acts as a server to clients using a remote
method invocation protocol. It crashes from time to time (~weekly?) with the
traceback triggered by %SYSTEM-F-STKOVF. It is threaded, using PTHREAD$RTL.
The only reference
to STKOVF (from help/message) refers to the linker facility. What I see is a
run time error. How do I increase the stack size such that it does not crash
in this fashion?
The Answer is :
The OpenVMS Wizard would assume this is either a bogus condition
value from the application (the application fell over with the
integer value 1364 left in register R0), or potentially an
internal error in a system component, or potentially insufficient
virtual address space. (The address shown is not part of the
OpenVMS user stack.)
First, please ensure that you have the available mandatory ECO
kits installed. For pointers and for an ECO severity search
engine, please see the FAQ.
Debugging discussions start in topic (1661) -- you will need to
perform at least some debugging of the error -- and you will
want to acquire the direct assistance of the support center.