|
I
would add that the Exchange Server Disaster Recovery Operations Guide covers a
number of specifics that should be of interest (http://microsoft.com/exchange/library).
In general, it is becoming somewhat rare that I recommend for people to “recover”
an Exchange server. In a disaster situation, it is generally quicker to restore
service using dial-tone principles and then merge in user data from a recovery
storage group…but YMMV and it depends on the goals of the recovery, the hardware
you have available, the failure mode, etc. From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Al Mulnick You're not missing something basic; that's part of the
design you have. You'll need a root domain controller and a domain
controller for the north american domain (assuming your Exchange server is part
of that forest and not a separate forest implementation.) Exchange relies
heavily on AD and AD will need the root domain and the domain that your
Exchange servers are in. From there, it's just a restore of the
individual server. Unfortunately, you'll have to do the same thing each
time which may break the security model that was intended. You'll have to
check. While you *could* put up a new forest etc, I don't advise it
because it doesn't contain the same software, settings, schema mods, etc.
It also doesn't give you a true representation of the actual process and so you
won't be able to pull all of the resources together as you would in the case of
a real melt-down. This test, IMHO should be a regular part of your testing as
you'll rarely want to put the full server back in a production setting
anyway. It's the data and the service you want, vs. the restored server. You may also want to check with the folks that architected
your solution and see what their plan was in the first place. This should have
come up many times during the planning phase. On 2/21/06, [EMAIL PROTECTED]
<[EMAIL PROTECTED]
> wrote: Hello all. I'm a longtime
lurker, and very infrequent poster. I take the advice of this
group to heart and apply it regularly every day. This might be
slightly off topic post, but I'm curious how you perform disaster recovery
testing on your Exchange environment. We are a
medium-sized part of a very large company. Our local site hosts a 2003
Domain Controller for the North American domain and the Exchange 2003 server
for our local users. My coworker and I are Domain Admins for
North America, but not for the root domain, and not for the forest. We
are also Exchange admins for our server, but not the entire Exchange
organization. We are attempting to test our Disaster Recovery Plan
for our Exchange server. One of our success criteria is to not interfere
with Production email or users. To this end, w e created a separate
network that has no connection to our Production network and moved
a Production DC for the NA domain onto this network. Although we
have been able to get DNS resolution to work from this DC, AD authentication is
not working and we cannot join any of our Test network servers to the domain
hosted by it. So far the recovery is going no where. Back in the
good ole days, we could have restored Exchange 5.5 independent of our domain,
whether it was NT 4.0 or Active Directory. Also, if this was a true
disaster, the recovery would be simpler since we would have access to the
Production network. Because of our success criteria of not
interfering with Production, we are in this limbo where we have to have some
Production data (represented in the DC and GC) to restore Exchange, but we
don't have access to the entire forest or Exchange org since the
Test has to be on a separate network. I curious
if any of you have encountered this issue before and how you got around
it. Building a separate Test forest on a separate Test network that
mimics Production is not out of the question, but it would require much more
support and cooperation from the corporate levels above us to complete.
This is something that would probably take months to setup in our
company. Are we missing something basic? Thanks in
advance for all of your help. jasonjordan MCSE, MCP+I, MCP Manager Security,
Audit, and Recovery Team Data
Center Services Emerson
Process Management, LLLP |
