Hi guys,
i am running a mission critical financial system on sql
2000 active/passive cluster.
right now i have only cluster failover and transactional
replication for reporting as disaster recovery mechanism.
i want to implement a solution which will allow me to
recover 100 % DATA with minimum down time in case of any
kind of disasters like database crash,server crash,raid
array crash,database corruption,fire etc.
can you advice me the best tool available in the market to
acheive this.
pls advice me.
Thanks
Biju

Well, the thing is ... you can never guarantee 100% data
recovery. You certainly need a robust backup and restore
plan, which I hope you have, but that is a point in time.

Transactional replication is about as real time as it
gets, but it's not the way to create a disaster recovery
site as it only send a subset of your data.

Log shipping is based upon backup/recovery, and is the
most common way to set up disaster recovery, however there
is latency involved. Even if you have the ability to make
tran log backups every minute, it still needs to get
copied and applied.

In a complete failure, what if your backups were bad? How
far back will you have to go? Where are the backups
located? On the disks that fail?

Having, say, a triple mirrored disk solution you can split
off may or may not work for you (again, it's point in
time), but it's certainly quick especially if you have a
large database.

So as you can tell, it's very complicated.

What you're asking applies whether or not you'd have a
cluster. You can prepare for most situations, b ut I'd
never guaratee 100% data as it depends on what you have
available to you and the type of failure. YOu can get
close, so you really need to figure out how much data loss
you can tolerate.

You may need to put some things in at the application
layer as well to help mitigate this.