I use an exit in "node" DTCPARMS that checks the DR status using the CPU serial
number, e.g.
.*-----------------------------------------------------------------------
.* Add user exit to stack startup
.*-----------------------------------------------------------------------
:nick.TCPIP :type.server :class.stack
:exit.DRCHECK
.*-----------------------------------------------------------------------
.* Add user exit to MPROUTE startup
.*-----------------------------------------------------------------------
:nick.MPROUTE :type.server :class.mproute
:exit.DRCHECK
The exit is specific to our environment, so I won't post it. It calls another
EXEC that determines whether we're in DR by checking the CPU serial number.
The operator is also prompted at system startup to tell the system whether
we're in a DR test or real disaster. DR tests are conducted behind a firewall,
so we use different OSA's for test and real DR. For the stack, it copies a DR
version of PROFILE TCPIP to TCPIP 191. For MPROUTE, it returns ":Config." with
the name of the appropriate MPROUTE CONFIG file. We don't use SYSTEM NETID or
SYSTEM CONFIG to change the node name for DR, because we have things that are
dependent on the node name.
Dennis
"If I could not go to heaven but with a [political] party, I would not go there
at all." -- Thomas Jefferson
-----Original Message-----
From: The IBM z/VM Operating System [mailto:[email protected]] On Behalf
Of McDonough, George
Sent: Thursday, August 12, 2010 08:13
To: [email protected]
Subject: [IBMVM] Disaster Recovery TCPIP Address Issue
Hello, Everyone.
We are in the planning stages for the first disaster recovery test of our
new z/VM environment. This test will be done at an offsite vender
location. While discussing our requirements, the topic of TCPIP addresse
s
came up. Since the addresses at the offsite location will be different
from our local addresses, we are at a loss as to how to change them. In
our z/OS environment, we just use the vendor's floor system to make our
changes. However, in the z/VM restoration we will not have a floor syste
m
to make similar updates. We believe we're not going to have access to
their HMC where we could use the Integrated 3270 Console, but we're still
trying to get final confirmation on that.
My question is how do other folks get their DR vendor's IP information
into and active on the restored VM system?
Thanks for your help.
George.
