[reSIProcate] Crashes in Resiprocate stack - help.
Hi,
I am not sure if this is the right forum to ask this question - but I
will
ask anyways as Resiprocate stack is involved here. I have a B2BUA server on
reciprocate running on two machines - 1st a Pentium 4 based PC running a
Linux Enterprise server and 2nd a Celeron based PC running a Linux
Enterprise server. Both the Linux are same versions. The B2BUA server is a
multi threaded application running on linux.
Well the two machines are not clean in the sense that other
installations
have been done time and again on these two PCs and hence libraries may
differ.
Now here is the problem part, the B2BUA runs pretty well on the P4,
but
it often and frequently crashes on the Celeron machine during our
sanity/system testing. The crashes are as simple as a call being made and
then being disconnected. We typically get a segmentation fault or an assert
abort and backtracing through gdb shows the stack crashing at a simple
InviteSessionHandler member function is being used.
I want to know does the speed of the machine, or a difference in
libraries
or something can cause this behavior indifference between the two nodes
running the same code ? Is it that because of Celeron being slower, we are
seeing some "potential" race conditions when traffice is more which are not
arising in the P4 ?
Has anyone been through this experience before and suggest some
pointers as
to how to go about debugging or finding the cause of such problems.
-Regards
Asheesh.