On Feb 20, 2006, at 6:11 PM, Ron and Jenny Hawkins wrote:
Ed,
If you think the main issue is distance then you seem to have
missed the
point. This is "In System Replication," meaning the source and the
target
are managed by the same Storage Control Unit.
If you are running XRC or TrueCopy then you send commands to the
Remote
Storage Controller to make an "In System Copy" of the Remote Copy
target.
The only additional traffic across that distance is the commands to
tell the
Remote CU to refresh and split the Shadowimage.
Dual Copy - there's a blast from the past. I agree that you could
think of
this as Dual Copy on steroids, even though Dual Copy was more of a
weak
attempt at RAID-1.
Ron
---------------------SNIP---------------
I believe the industry is misleading the user (our management) with
the term "instant".
I think*MAYBE* a better term would be point in time with an asterisk
explaining the exception(s).
For remote point in time copy, to me its disengenious to call it
instant. Fast probably is OK as well. But the remote "funnel" has to
be fairly big to handle the I/O load. I am not involved in the
purchasing decisions but I would think a few t-4's or dark fibre
might be able to handle it.
For a local it is still tied how fast the control unit (shark, stk
etc) can actually duplicate the data.
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
