On Thu, 30 May 2013 16:15:42 -0500, Tom Marchant <[email protected]> 
wrote:

>On Thu, 30 May 2013 11:44:32 -0700, Lizette Koehler wrote:
>
>>So do I have overkill?  .
>
>Software disasters can be the hardest ones to plan for.  What do 
>you do if one of your critical applications has a program change that 
>causes it to start corrupting data?  How long will it take before it is 
>noticed?  This can be a lot harder than a hardware failure.
>

Or "human" disasters, Tom. Someone deletes a data set, and because the DASD is 
mirrored everywhere, all your online copies are gone instantly. Oh, and if you 
didn't have any real backup copies of the DASD, then all copies of that data 
set are gone.

That's one reason that IBM recommends using RACF's duplexing of it's database, 
rather than depending on hardware mirror copies, and also recommend taking 
nightly backups of the database. When an administrator makes a mistake it can 
save a lot of hassle.

And, if RACF itself makes a mistake, there's a good chance that only the 
primary (or the duplex) copy will be damaged. But if you were depending on the 
hardware mirroring they're all broken.

-- 
Walt (former RACF Designer)

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to