>
> If you are sure, that the problem occurs in the same time interval, not
> approximately, but exactly, then it may not even be apache causing the
> problem, rather something else which triggers apache to have the problem.
>
> I'm thinking of some sort of DHCP server enabled ...
>
> How apache counts exactly, is partly explain in the configuration files, but
> more indepth at apache.org, the documentation.
>
> If you have a load that you can't quite figure out when it has it's maximum,
> the documentations tell you which settings to alter one by one in order to
> have apache handle all requests in due time.
>
> Last, you have an option about the max_lifetime of a child process, if you
> locate the exact directive and eventually reduce the lifetime, apache might
> be able to kill these hanging childprocesses faster and thus stay stable.
>
> /Andreas


Is there need to tune the TCP parameters for Solaris if I am running
Apache under high load?