Alot of this depends on how much $ get devoted to DR, which of course shows 
how much of a commitment management is willing to make to it and how 
serious they are about having true DR processes in place.  I'm putting a DR 
system together right now for a customer, and we are building the box 
absolutely identical to the production box so as to minimize "gotchas".  It 
will cost maybe $20,000 more than patching together a 9x9 system (it's an 
rp box), but the SLT tape won't know the difference if and when it's 
needed, and you can imagine the panic that will be in the air when that day 
comes.

The next decision you have to make is how current you want the DR box 
running with the Prod. box, which can vary from a DR box that sits in a 
closet unplugged to one that is running concurrent with the Prod box.  The 
latter is, of course, better but it's also the most expensive.  There are 
some in-between options, also, which is what your budget committee must 
decide.  If you need true DR, then you need to run concurrent.  It probably 
comes down to  how much it costs you per-hour to be down...that's what I 
would use to determine it.

Lastly, you need to check the data pipes in and out of your DR location and 
run processes through them, make sure they can connect to any remote 
locations, web orders go through as normal, etc.    Set up your DR system, 
then start playing with it to find any holes.

John Lee
Vaske Computer Solutions

At 09:10 AM 9/19/06 -0700, donna hofmeister wrote:
>--- Jim McDermott <[log in to unmask]> wrote:
>
> > I've been away from the HPe3000 for nearly 2 years and just recently
> > got a contract with a client to create a Disaster Recovery scenario
> > at an IBM facility with their HPe3000/979.  I used the client's CSLT
> > to do an install and tried to add volume sets and members after a
> > successful boot but have had major problems with drives "configured
> > but not available" errors.  I assumed that the client and the DR site
>
> > HPe3000 were identical but found out that the DR site had multiple
> > EMC drives which the client did not.  I used SYSGEN to change the
> > drive id's which worked but I couldn't get the paths changed
> > using the MAPPER listing data.  Could anyone help me and tell me what
> > I need to do.
>
>craig is right, you need to do either search or mapper while the box is
>down to figure out what the DR system is seeing (or not).
>
>this brings up a topic that i'd like to collectively discuss -- what to
>do (or not) for dr systems.
>
>as jim has rather painfully discovered, if you're doing true DR (that
>is you're expected to bring up your mpe production system on a
>unfamiliar box) there's a real challenge getting the hardware
>configuration established.  (you can pretty well figure that you're
>going to have a different architecture box (eg, you run 9x9 at
>home...but your DR box is 99x or an n-class))  so that's one...
>
>another discussion point is -- then what? you've got the hardware
>straightened out and it's time to start restoring -- are there any more
>gotcha's?
>
>for folks doing modified-dr (where you have a distinct system with an
>established config), i know there are files (especially in the sys
>account) that you are *not* going to want to disturb.  again, are there
>gotcha's?  (i'm in this situation, and i'm wondering about how to keep
>the two systems insync for patches and other hp-delivered software).
>- d
>
>Donna Hofmeister, HP-CSA  Sr. System Programmer
>dgarverick -at- longs -dot- com
>925-210-6631              Longs Drug Stores
>
> >>>MY opinions, not Longs Drug Stores'<<<
>
>__________________________________________________
>Do You Yahoo!?
>Tired of spam?  Yahoo! Mail has the best spam protection around
>http://mail.yahoo.com
>
>* To join/leave the list, search archives, change list settings, *
>* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *