----------------------------<snip>---------------------------
To warm up (is that an English expression?) an old argument: The above
*assumes* that it is a 'real' test, as in "Let's take down the system on
the fly and see if it comes up again". I have seen this done once, and
it really pointed out a lot of problems that we could then fix.
----------------------------<unsnip>--------------------------
We always took incremental backups at several points through the day.
The assumption was that we lost power, so the only recovery mechanism we
had was those incremental backup tapes. We also played with XRC, from
our Shark box to the remote site, with the data mover operating from the
remote site.
-----------------------------<snip>--------------------------------
Unfortunately, the going practise is to "test DR" by doing an orderly
shutdown first and then just re-IPL in the other location. In my
opinion, this just tests that you defined hardware (and maybe some
infrastructure in iplparm/loadxx/ieasysxx) correctly), but it is not a
DR test.
------------------------------<unsnip>-----------------------------
Agreed: that's not a true "disaster" if you have time for an orderly
shutdown.
----------------------------------------------------------------------
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
