Help creating failover fileshare

[hd2]

Hi all,

Thanks for reading.
I'm trying to create a failover solution that would allow users to keep accessing network file shares in the even of one server going down.

Here's a brief outline of our network:

Two buildings, linked by fiber (two pair being used, I have access to 6 pair if needed).
Each building has approx. 40 workstations in each, all running on 1000bT copper.

Building 1
Server: DC1 - Domain controller
Windows 2003 x64
Connected to the LAN through 1000bT copper.

There is a Drobopro connected to it, through a switch, via iSCSI that contains all the file shares all the workstations in both buildings pull from.

If anything happens to DC1, everyone looses their network drives.



Building 2
Server: DC2 - Domain controller
Windows 2003 x64
Conected to the LAN through 1000bT copper.

Note, DC1 and DC2 could have a direct connection to each other on their own fiber if needed.

Is it possible to add a second Drobopro (or other device) to DC2, via iSCSI, and have it mirroring the Drobopro connected to DC1, then, if something happened to DC1, DC2 takes over without major interruption?

Or, still only use the one drobopro, but have DC2 take over the authentication / sharing tasks until DC1 is fixed?

I hope that makes some sense.

Thanks a lot for any help or suggestions.
Take care.

 

[fugtruck]

One product I have had success with is Double-Take. It does replication & failover for file shares. I'm sure there are a ton of ways to accomplish this, but it's one way that worked well for me.

 

[RMGBELGIUM]

if your latency is less then 500 ms, why don't you use a MS cluster ? This takes care of the server side redundancy.
Double take is indeed a much used solution as well. You can setup a DFS, and replicate with double take your primary storage to your DR site. In case of disaster, you can easily ( through scripts ) let all shares point to the DR site.

rgds,

R.

NetApp Certified NCDA/NCIE-SAN