On our network we have two domain controllers. (I can't remember the new terminology so clue me in . . .)

The primary DC is responsible for everything. AD is set up here,DHCP,DNS.

The backup DC has AD and DNS setup for backup.

The primary DC was rebooted today and odd things happened on the network (mostly everyone's PC locked up). I thought the point of the backup domain controller was to pick up where the primary DC left off.

What don't I understand?

Po

The is no such thing as a backup domain controller as you think about it. Server failover like that is referred to as Clustering.

You need to understand there are forest/domain functions and then file server/app server functions. Two domain controllers are greate for forest/domain functions but, unless you have set it up, don't replicate file shares or application access.

Both DC's should have the global catalog enabled. Here is how you do that on the second server
http://support.microsoft.com/kb/313994

Both servers provide name resolution [dns] and user authenication services.

You may want to consider implementing DFS and its subset FRS to build in more fault tolerance/failover to your system.

Imagine the power of knowing how to internet search
http://www.lib.berkeley.edu/Teachin...

Wanderer- thanks for a great reply. I wonder, why did we set up a backup domain controller if it can't fill in for the primary domain controller?

Is it that everything is in place should the primary DC go down?

In that case why do you think only DNS and authentication were configured on the backup.

The primary has so much more going on - it's the DHCP server (we lose that if it goes down?) and it's the schema master.

So I guess I don't understand what good the backup is. Maybe a little DC 101 would help.

Thanks

Po

MS recommends a second dc for one major reason. If you only have one DC and it dies you lost EVERYTHING and must create AD from scratch.

That is no small task. Two DCs also share the load as far as name resolution [dns] and user authenications [connecting to shares and file/folder rights].

You should copy the dhcp folder info to the other server but don't do anything with it unless dc1 dies. Then you can authorize and walla dhcp is working again.

You can seize the fsmo roles if dc1 dies to dc2.

Lot's of reasons for two DCs. You just have to redefine what you consider "backup". If you want complete fault tolerant failover you need to cluster the two servers but that gets expensive and there is a learning curve.

You should seriously consider engaging DFS/FRS so you have a replica of your files and shares on the 2nd dc. Might come in handy someday...

Imagine the power of knowing how to internet search
http://www.lib.berkeley.edu/Teachin...