On Aug 24, 2005, at 4:28 PM, Mark Zelden wrote:
-----------------------
SNIP-----------------------------------------------------------
I didn't ever set it up that way myself, but no. You only defined
new SSAs at install time and they pointed to the maintenence
sysres(es) via catalog. I guess if you had to move the data sets for
some reason you would need to DELETE (noscratch) and DEFINE the
data sets to the new volumes, but how often would one change volsers
within a single release. When's the last time you changed the
devicetype of your maintenance sysres mid-release?!
A simple example would to have data sets cataloged using SSAs
like this (to support multiple maintainable levels of the OS):
TGTZ14.SYS1.LINKLIB
TGTZ15.SYS1.LINKLIB
TGTZ16.SYS1.LINKLIB
I personally like the other method better because my cloning procedures
include creating a tgt zone that matches each sysres set. This has 2
advantages that I like:
1) I could apply a PTF in an emergency to a live sysres (such as an
ISPF fix that could be implemented via dynamic LPA or LLA update):
2) When looking at a problem and IBM asks "do you have PTF UAnnnn
on?"
I can trust what SMP/E tells me instead of having to refer to a
chart or some other method that tells me when the sysres was
cloned.
BTW, my target zone names all are prefixed with the actual volser
of the primary sysres. For example (these aren't my actual
names):
zones: RSM60A1 primary sysres: RSM60A (maintenance sysres)
RSM60A2
zones: RSZ61A1 primary sysres: RSZ61A
RSZ61A2
zones: RSZ62A1 primary sysres: RSZ62A
RSZ62A2
etc.
Regards,
Mark
Mark,
You have some good (and valid) points.
Over the years I have had to change device type 2 or maybe 3 times.
Changing the volser 10-15 times . Changing the volser was done to
conform to local "standards".
Using your methodology still has issues for device type mass change of
all references in a catalog. I think the same applies to volser change.
I usually write a quick and dirty rexx or clist to change such items.
The DDDEF seems (to me) a quick and dirty way of changing the
information without the need for a clist or rexx.
I think I like the idea of not cataloging the datasets and using the
DDEF way. The only issue I have run into is that at times I have had
conflicts with the DSN (dataset in use).
I have gone as far as unidexing the volume and zapping the vtoc to get
access to a dataset. Its usually in a pinch and I am semi desperate. to
get access to a dataset.
Ed
----------------------------------------------------------------------
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
