2 servers accessing the same files via shared LUN?

[dai]

Hi all.

 

I am in the process of setting up a brand new Intel MFSYS 25 modular server system with Windows Web Server 2008 for the purpose of hosting redundant web and mail servers.

 

The MFSYS25 has a shared LUN feature which allows me to create a RAID5 drive and allow multiple servers to access it at the same time. Which is great, because I have created individual system partitions (drive C) for the OS and a shared data partition (D drive) for web files and email to be stored.

 

However, although all servers can see the D: drive it would seem that they can't access the same files at the same time. For example, I have several folders in there and while Server1 can read/write using folder1 and Server2 can read/write using folder2, if I create a new folder on Server1 it won't appear on Server2 until I reboot (sometimes it feels the need to run a disk check).

 

I'm assuming this is a limitation of Windows Web Server but am hoping someone can confirm this and help me with a workaround, as I'm only hosting simple web sites and email accounts so don't want the expense of Enterprise edition.

 

Any help or advice would be greatly appreciated.

 

Regards,

Dai

[AWS]

It shouldn't need a reboot to make new folders visible betwwen servers.

 

I don't know what would cause this. Mis-configuration of the raid array most likely.

[dai]

I'm not really sure where to start looking. The RAID is easily configured through the web interface of the MFSYS25 so there is very little to mis-configure.

 

From searching around I keep reading that Windows only supports a 'shared-nothing' configuration. I'm not sure what this means but maybe it's a limitation of Windows that menas I can't have the SAN shared between multiple servers as I want to.

 

Unfortunately I only have Web Server 2008 and not Enterprise, so don't have access to all the clustering tools that could help me out.

 

There may be another way around it, but all I want to acheive is for 2 mail servers to access the same data, so that if one fails NLB directs all requests to the remaining server and the same live data is still available. I don't want to set up any replication scripts to kick in every 5 minutes or something.

 

Thanks,

Dave

[AWS]

I did some searching and I cam to the same conclusion.

 

I bet somewhere there is a solution. Someone always come up with a work around.