Re: How would you handle this installation...

It would help if you were a bit more specific about what name you are
referring to. I get the impression you mean the name that appears on
the signon display which is the System Name specified via CHGNETA.

Assuming that's the case, if it was me what I would (probably) do is
proceed as planned, giving the new machine a temporary IP address
while testing. Naming may not be an issue at that point depending on
what you are testing. if you want to give it a different name during
testing (and its probably a good idea) then:
- do a CHGNETA to set the new system name
- delete the existing directory entries using WRKDIRE
- CFGTCP and change the host name
- WRKRDBDIRE and delete and recreate the *LOCAL directory entry; local
directory name should be the same as the unqualified hostname

At cutover time I would yank the cable on the current machine and do
an unload/reload from the old machine to the new; this will take
across IP address, system names, directory entries etc. You may need
to modify your ethernet lines to align them with the resource names on
the new box.

After the machine was signed off as configured I would scratch the old
box and build it with a new name as you have described. If you use a
save from the new box then the biggest hurdle you will face will be
rebuilding your SNADS directories - presuming you need snads
connectivity. I find it handy at the system level for SAVRSTOBJ type
operations but beyond that it's use is limited IMO.

If you only have one machine now (and I get the impression you do)
then if you restore from the new machine back to the old you will
(once again) need to:
- do a CHGNETA to set the system name
- delete the existing directory entries using WRKDIRE
- CFGTCP and change the host name
- WRKRDBDIRE and delete and recreate the *LOCAL directory entry; local
directory name should be the same as the unqualified hostname

Hoep this helps


On Wed, Jun 6, 2012 at 9:52 AM, Graap, Kenneth
<Kenneth.Graap@xxxxxxxxxxxxx> wrote:

I'm wondering how to best handle the following situation:


1.)    I have a Power6 V7R1 system configured with the name "S02" . It is currently my Production System.

2.)    I'm purchasing a new Power7 V7R1 system, which will eventually be "S02" my Production System.

3.)    I want to configure the Power7 system using iASP's in preparation for a System Mirror implementation, while having it attached to the same network as our production Power6 system. I'm assuming it needs to be named something other than "S02".

4.)    Once the Power7 system is configured and tested, I want to backup Power6 application data and restore it to the Power7 system and rename it "S02".

5.)    I then want to scratch and configure the Power6 system into one named S02BU and use it as a CBU/System Mirror for the Power7.

6.)    I don't completely understand all the implications associated with changing the system name of the Power7 once it becomes the Production System.


I know that changing the System Name of an IBM i can be problematic, especially when BRMS is installed. I also believe that there can be issues with System Directory Names (WRKDIRE kind, not IFS directories) and even the RDB (relational Database Directory).

How would you go about coordinating this name swapping game with the intent of eventually having the Power6 named S02BU and the Power7 named S02 ???

Thanx in advance... Any suggestions and tips will be greatly appreciated...

Kenneth
Kenneth E. Graap
System Administrator
503.226.4211 x5537
http://www.linkedin.com/in/kennethgraap