General clustering questions
Hi,
One of my clients have asked me to draw up and implement a fail over
solution for their network, the network is will be having new Windows 2003
server implemented in the next six months.
The purposed set-up is to have the following:
2 X Windows 2003 – Acting as file servers
1 X Exchange 2003
1 X SQL 2000
Although not set in stone I think the servers will be HP ML370’s with two
network interfaces.
In an event of a server falling over the redundant server should provide
users with the following:
Access the server-stored files
DHCP, DNS and domain logon authentication
After reading through the Microsoft’s do***entation I feel more confused
than I did before I even started reading it, please could you help me by
clearing up a few questions I have.
1.For the above-proposed set-up would Network Load Balancing (NLB) be enough
or does it have to be Server clustering.
2.There are no plans for SAN type storage so does clustering work (at all)
if the data is stored on the servers themselves (In other words does it
replicate the data between servers), and if it does is it done in real-time
or is a utility like Robocopy needed
3.Are there any good references (other than Microsoft website), which you
recommend websites or books?
I’m feeling rather dense when it comes to clustering at the moment
Kind Regards
Richard Harlaar
|