We have a use case for RSH-Daemon and have been using it on many older versions
of OpenSuse without problems.  Recently we deployed two new OpenSuse 13.2
servers and both are failing in the same manner.  The rsh-server daemon fails
to respond to queries with great frequency.  Our networking guys have sniffed
the line and determined that the packets are getting to the server, but the
daemon simply is not always responding.

Xinetd has been tuned to accept a high number of connections, this is not an
issue of tuning.

The kernel has been patched to the latest release.

We have tried 10Gb copper, 10Gb fiber and 1Gb copper networking cards to see if
possibly it was just a buggy driver for one of the NICs....but all had the same
results.

Running xinetd in the foreground shows that the packets are not getting to the
daemon, when its failing nothing is displayed in the debug mode.  It just seems
to go nowhere.

the log files don't indicate failure.

On the side making the RSH request, it just hangs for a good while and then
indicates that the connection failed.  But then suddenly it will work again for
a while and then will stop for a period of time.

When RSH is failing, other services are working including SSH from the same
computer.

We're at a loss, it's a very nasty and low level issue that it seems like we
cannot fix.