We have had major problems with TCP/IP 3.2, trying to go to high levels of
TCP/IP maintenance. We run MVS 5.2, with our ADSM server at level 13.  The
problems, as I remember them were timeout problems - we would loose sessions
all through the night.  I would see ANR0480, ANR0481, and ANR0482 messages in
my log depending on how the communication was lost between client and server.
 We did not pin down what caused the problems.

There have been some postings here to the effect that you should apply PTF
UQ03848 to solve all of your problems with TCP/IP 3.2, and I think for some
environments, this may be ok.  However, we HAD the problem when UQ03848 was
applied, so I do not think it is a panacea.  As always, there is other
baggage here - we applied UQ03848 with 15 other PTFs to IP that addressed
problems we were having with another TCP/IP application.  The problem is
undoubtedly in there somewhere, but at the time, IBM had no idea where the
problem was.

To solve our problem, we have had to back off the IP maintenance.  We have
been talking to IBM, and they finally closed some PTFs that are supposed to
address the problems we had.  We are in the process of trying to reapply the
maintenance, but will not be able to for a little while yet, as we are in a
period of "heightened awareness" for year end processing.

Jerry Lawson
jlawson AT thehartford DOT com


______________________________ Forward Header __________________________________
Subject: TCP/IP Problems
Author:  INTERNET.OWNERAD at SNADGATE
Date:    1/20/98 9:22 PM


Hey, everyone,

     Our TCPIP folks here installed some maintenance last Thursday, 1/15.
On Saturday,
1/17 we had a scheduled shop shutdown, at which time ADSM was recycled.
Ever since
that time we have experienced major problems with scheduled backups
failing, as well
as batch production jobs that initiate a TSO client session with the
server.  What's wierd is
that not all backups fail, and not all production jobs fail.  Only those
jobs that initiate more
than one session fail (always when the second session is initiated); and I
have yet to figure
out a common thread for the client backups failing; they are not all of the
same platform,
however.

     Has anyone had a similar problem RECENTLY?  We are running TCPIP 3.2
at
9407+, MVS server 2.1 and most of the clients failing so far are NT 4.0 and
UNIX, some
NetWare.  I am not a TCPIP person, but I understand that the maintenance
applied was
very recently issued and was supposed to alleviate problems with CCB (?)
 buffer space.

     I know it's not much to go on - but it's all I got so far!

                              Ginny