Full Disaster Recovery

[Guest]

Need to see how others are doing DR. We have exisiting
Dell Servers. We have now taken the tapes and tried to
recover the servers to other hardware or seperate lan.

The idea being to simulate total failure (Fire ) in
office. No matter what we do, Repair etc we can not get
the servers to function reliably of the "disimilar"
hardware. I have read many of the Howto and followed
many of the procedures. I can get to Active DIrectory
Users and Computers, I can see restored data etc. But
the thing is unreliable on all flavours of hardware that
we have tried as recovery vehicles.

Therefore my question is. ......
Is it truly practical to attempt a full recovery of
servers in event of disaster (On supplied hardware at
time)or are we better rebuilding from scratch and using
ldifde csvde and exchange database imports to get us up
and running with backup tapes of data. We have no second
site with reliable links therefore a replicated server is
not an option.
After all in the main the management may have burnt in
our fire and sid will be largely unimportant as the PC`s
will have fuelled it
Thanks in advance
Tim

 

[Ray Lava [MSFT]]

Tim,

You are absolutely correct that doing full restores including system state to dissimilar hardware is a very iffy prospect. You have obviously already read kb article 263532 which deals with this topic and explains why Microsoft cannot guarantee that restoring domain controllers to dissimilar hardware will work. In fact, the more dissimilar the hardware, the less likely the restore will be successfully.

Your best bet would be to try to have two servers with exactly the same hardware. One of those servers would be a functioning DC, then other would be kept offsite. While this can be an expensive proposition, you could always purchase two low end servers for this purpose. As long as the low end domain controller does not hold any FSMO roles or run any apps such as exchange, terminal server or hold any FSMO roles, it authenticate users and replicate with your other DCs without any problems. Make sure that you are getting system state backups of your low end DC so that in a worst case scenario, you could do your restore to your low end server that you keep offsite.

I hope this helps.

Ray Lava
Microsoft Corporation

This posting is provided "AS IS" with no warranties, and confers no rights

 

[Jim]

I maintain the root domain, and one child domain. For each domain I keep two domain controllers running on site with FSMO roles split out per MS. I also have a third DC off site (25 miles away) just for DR reasons. I reused older servers that would normally be retired, but have enough memory, raid, etc. They run in a site that belongs to another child domain and has a good network connection. I count on hardware constantly changing, so in a real disaster I would need to buy current model servers. My off site DC goes a long way to bring things back to normal again since I can sieze the FSMO roles, clean up machine accounts, etc. I will still have to restore app servers, etc. but having the domain intact is a huge help. If both sites were lost to a MAJOR disaster, I probably won't care anyway... Just my two cents on DR.