Some of this behaviour should of changed in the first update set
for rhel3. (up2date-4.2.5)

Are you still seeing these problems with 4.2.5?

Yes, this behavior was observed in version 4.2.5, as indicated in the Version-Release 
information above ('rpm -q' returns "up2date-4.2.5-1").

I'm not sure how the current code would be failing to
detect the ip correctly. Heres a run down of what it
does to detect this:

1. Looks up hostname  (socket.gethostname)
2. resolves that name to an up (socket.gethostbyname())

If the above is not 127.0.0.1 and/or "localhost"
it tries:

opens a socket to serverURL (xmlrpc.rhn.redhat.com by default)
or to proxyUrl (if a proxy is configured)

figures out what interface that socket is on, and gets its
ip address. Then it resolves the hostname via a gethostbyaddr
on that address.

The first step can "fail" if the machine belives its hostname
is "localhost" and/or the hostname resolves to 127.0.0.1. 
Check the /etc/hosts file to see if this is whats happening.

That said, the second step should always get a useable IP
address, so I'm not sure whats going on. Anything unusual
about this machines hostname, dns, ip, etc?

It's possible that the behavior is a result of the /etc/hosts file.  The default (i.e., the 
one generated by the RHEL 3 ES install) seemed a little strange to me, and I've since 
fixed it (but I'm haven't re-registered the machine since).  The original /etc/hosts file 
looked something like this:

# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1               localhost.localdomain localhost my.example.com my

I subsequently changed it to the following:

# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1               localhost.localdomain localhost
XXX.XXX.XXX.XXX         my.example.com my

The interface was listening on the correct IP address, but it was not in the /etc/hosts 
file.

FYI:  I since performed an update on the hardware profile associated with the machine 
(by running up2date --hardware), and it now lists the correct hostname and IP 
address on my Systems profile page on RHN.  I imagine that the alterations to the /
etc/hosts file may be responsible for this, although I can't say for sure.

Perhaps the RedHat team may want to look into how the default /etc/hosts file is 
generated during the install of RHEL 3 ES.

Thanks!

I've just done my very first RHEL 3 WS U1 install (so I am using
up2date-4.2.5-1) and have had a similar problem. At the stage where
up2date was detecting my configuration it just sat there for a few
minutes and then listed ERROR in all the fields. I tried running
up2date at least 10 times with the same results.

After reading the above comments I checked my /etc/hosts file,
originally it read:

# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1               mymachine localhost.localdomain localhost

so I removed mymachine so the machine would rely on DNS to resolve the
its IP address, i.e. my /etc/hosts file now reads:

# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1               localhost.localdomain localhost

and this instantly fixed my up2date problem.

Other information: This was a totally fresh full install, I configured
up2date to work via a proxy, I was using an already existing RHN
account with 3 new and unused entitlements.

Like Brian, the default /etc/hosts file appears strange to me also and
I don't expect a "ping mymachine" command to return an IP address of
127.0.0.1