On Oct 22, 10:31am, Bob Booth - CSO wrote:
> Subject: And yet another ANR9999D..
> Any know how to track this one down:
>
> ANR9999D tcpcomm.c(352): Unknown TCPIP Failure.
> ANR9999D tcpcomm.c(352): Unknown TCPIP Failure.
> ANR9999D tcpcomm.c(352): Unknown TCPIP Failure.
> ANR9999D tcpcomm.c(352): Unknown TCPIP Failure.
> ANR9999D tcpcomm.c(352): Unknown TCPIP Failure.
> ANR9999D tcpcomm.c(352): Unknown TCPIP Failure.
> ANR9999D tcpcomm.c(352): Unknown TCPIP Failure.
> ANR9999D tcpcomm.c(352): Unknown TCPIP Failure.
> ANR9999D tcpcomm.c(352): Unknown TCPIP Failure.
> ANR9999D tcpcomm.c(352): Unknown TCPIP Failure.
>
>
> Is there a trace level that can be performed, or a 'command' that will shed
> some light on this?  They have been appearing in groups like this all night.
>
> Thanks again!
>
> Bob Booth
>-- End of excerpt from Bob Booth - CSO

We are observing that kind of error for several months, first under 2.1.0.3,
now under 2.1.0.9. I have opened a pmr and we have produced a lot of trace
data. Analyzing the traces it comes out that we are receiving an error other
than ETIMEDOUT, ECONNREFUSED or ENETUNREACH.  Since IBM do not trace the errno
received we do not know what errno we did get. Network sniffing would be
necessary for further investigations.
It seems that the error is triggered by SunOS clients under heavy load. But
that's not for sure.

Werner

I have opened a pmr and we have produced a lot of trace data.





--
--
--
 ===========================================================================
 ===========================================================================
Werner Baur
Leibniz-Rechenzentrum     X.400:  S=Baur;OU=lrz;P=lrz-muenchen;A=d400;C=de
Barer Str. 21            RFC822:  Werner.Baur AT lrz DOT de
D-80333 Muenchen           Tel.:  ++49-89-28928781
Germany                     Fax:  ++49-89-2809460
 ===========================================================================