---------------------------------<snip>------------------------------------
Would people agree with the following?
If a vendor were shipping a product that resided in two datasets,
FOO.THIS and FOO.THAT, best practice would be to recommend that the
installer either create a user catalog named FOO or a user catalog with
an alias of FOO so that the FOO.xxxx datasets would not take up space in
the master catalog?
--------------------------------<unsnip>-----------------------------------
Charles, there are several reasons I can think of to discourage this idea.
1. You don't want any more updates to you Master Catalog than absolutely
necessary. ANY update leaves open the possibility of daamage to the MCAT
and you can't IPL a system without a Master Catalog that contains the
required System datasets. So the safety of your master catalog is paramount.
2. You don't really want to create too many user catalogs, else the
management becomes a serious problem. How much time are you willing to
spend reorging ICF catalogs?
3. Aliases allow you to share the catalogs between application sets more
efficiently. I generally recommend no more than 4 UCATs: Production,
Testing, TSO and whatever you choose to call your "Quality Assurnance"
or "Production Support", for moving programs from test to production as
appropriate.
4. Recovering a lot of small catalogs during Disaster Recovery, either
as an exercise or the real thing, can be a major stumbling block,
delaying the process and possibly preventing successful completion.
5. By using defined aliases, you eliminate any possible ambiguity in
determining which catalog will contain which datasets.
Rick
----------------------------------------------------------------------
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
