Bruce
No doubt Stuart Willis will answer for himself but, in case he's become
confused, I need to sort out a few points here. Also you have become a bit
confused over Enterprise Extender!
Stuart originally presented a log for his Communications Server (CS) IP
address space.
I believe the problem he wanted sorting out was the fact that the code in
message EZZ4310I was not documented in even the latest CS IP and SNA
Codes manual in Chapter 3. "Data link control (DLC) status codes" which is
where the explanation of message EZZ4310I says the code should be found.
Indeed quoting Stuart's original post there are codes starting "30nn and 32nn
but no 31nn".
Now Ed Finnell jumped in and provided Stuart with the explanation of the
EZZ0331I message straight out of LookAt! which, as far as I can tell, hadn't
fazed Stuart at all!
Meantime I indicated to Stuart that the explanation of the code in the
EZZ4310I message was documented in APAR OA25064 - and I hope he found
my post explaining these updates to the IP and SNA Codes manual.
Because of Ed's contribution Stuart proved that there was a HOME list IP
address for the OSA interface.
The EZZ0331I refers to a link name invented by CS for communication
between two instances of CS IP address spaces running with a z/OS system in
the same LPAR. It is also used for communication between an instance of a CS
IP address space and the instance of a CS SNA (VTAM) address space
necessarily running with a z/OS system in the same LPAR for use by the
Enterprise Extender function.
IUTSAMEH ("same host") is indeed something cooked up by "VTAM" but only in
the sense that the old VTAM development shop have been given responsibility
for all of the more recent - in the last 10 years or so - communication
functions within the CS pair of products. Note that, where two or more CS IP
instances are using IUTSAMEH interfaces, VTAM as the CS SNA component is
just not involved.
IUTSAMEH is a bit messy. My patchy experience with the CS products in the
last few years has always involved specifying DYNAMICXCF in the CS IP
PROFILE data set. From APAR PQ38618 this necessarily causes the IUTSAMEH
essentially internal interface to be created and assigns the address specified
in the DYNAMICXCF statement as the HOME address for the IUTSAMEH
interface. There's a lot of "under-the-covers" supposed help going on which
can be very confusing! If I had the choice I'd prefer just defining what was
needed using first principles. Take a look at the section "Dynamic XCF" in
Chapter 8, "TCP/IP in a sysplex" in the CS IP Configuration Guide in order to
see all the "under-the-covers" tricks CS IP gets up to to try to reduce so
much tedious definition work!
Stuart must not have DYNAMICXCF specified since that would avoid an
IUTSAMEH interface being defined without a corresponding HOME list address.
Maybe CS IP is so used to DYNAMICXCF being defined that - at some point -
the "under-the-covers" IUTSAMEH interface is *always* defined and, if you
don't have DYNAMICXCF specified, you just have to put up with that
EZZ0331I message and you should - simply - ignore - it!
If there is someone still reading this who has experience with CS IP *without*
DYNAMICXCF specified who can state whether or not IUTSAMEH is defined
automatically, I for one would like to hear about it - perhaps Stuart himself.
-
Now to the rest of your post.
I wonder what you mean by "TCPIP's HOME IP address". IP addresses are
associated with *interfaces* in the HOME list. There is no "TCPIP HOME IP
address" as such. I have for very many years now recommended that, in order
to act as an IP address for all applications which are necessarily associated
with a particular instance of CS IP, a static VIPA is defined for that very
purpose. The standard application I give as an example is the SNMP agent. I
call this the "node" static VIPA but a customer where I work from time to time
calls it the "source" static VIPA since it is defined as a source address for
some applications. Such a static VIPA can also be used as the local address
for Enterprise Extender. Perhaps such a static VIPA is what you had in mind
as "TCPIP's HOME IP address" - or perhaps not!?!
In the past it was required that the IP address was specified as the IPADDR
start option in the ATCSTRxx member. Best practice today is to specify the
IPADDR operand (or HOSTNAME operand) on the GROUP statement following
the PORT statement with MEDIUM=HPRIP in the XCA major node used for
Enterprise Extender. The IPADDR or HOSTNAME start options then become the
default values to be used in case the values are not specified on the GROUP
statement.
"Along with NODETYPE=NN" is just wrong. What is correct is that the
NODETYPE start option must be specified - definitely along with HPR=RTP
although this is the default - but the value can be EN as well as NN.
Chris Mason
On Fri, 20 Feb 2009 07:59:39 -0600, Bruce Richardson
<[email protected]> wrote:
>I think the key is the VTAM construct IUTSAMEH - this is used in Enterprise
>Extender (EE). You need a HOME IP address for VTAM to use, different from
>TCPIP's HOME IP address. It is defined in the VTAMLST(ACTSTRxx), as
>IPADDR=xxx.xxx.xxx.xxx (dotted IP number address) along with
NODETYPE=NN.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
