I have administered RHEL a few years ago, where I manually backup data using large HDD every Friday afternoon. Other than this, I have never been exposed to large enterprise environment, and can only image what the real situation is, so please forgive me for my knowledge.

I have setup multiple Centos servers through virtualization. I have four nodes and build GlusterFS storage, each node has 500G, (here I'm setting up things to simulate and practise SAN/NAS, this isn't REAL JOB) the volume is Distributed Replicated type. I have a main server where I generate 10-30M data per day, before going to bed, I mounted the Gluster volume onto the main server using NFS, and I zip/copy the data to the GlusterFS volume, then disconnect NFS mount, is this how the SAN how supposed to work?

In a large enterprise, I understand you have dedicated Storage server, with 20,40, maybe 120 HDD, my question are:
1). Do you write some small scripts to automate the task at 12AM? Since every vendor (IBM, EMC, Dell, HP, Oracle, RedHat) have storage products, if you used EMC in the past, what kind of learning curve when you use Dell products (say after job change)
2). What software allow you to backup only newly created/modified data onto the server? is it command-line or GUI based? Also RedHat is software company, so what's the point of using RedHat Storage Server when each hardware vendor already supply all-in-one solution (software, hardware, etc)?
3). If your data is generated in NYC, and storage servers in Nebraska, obviously the data transfer MUST be encrypted, do you use VPN or dedicated internet?
4). Any website where NAS/SAN are explained in great detail? I can't find any Something like this is too generic.
http://www.dummies.com/how-to/conten...-networks.html

Thank you so much for your patience&understanding!

In short. A NAS is just a service that creates a shared folder on a network so people can share their data.

A SAN on the other hand acts as if it is the computer hdd. So that means you can have a client with no hdd and boot from the SAN with pxe. The client will then mount the configured SAN and use it as the primary HDD.

So to summarize. Nas is for sharing, San is for hdd replacement.

So theoretically you would have one monster rack with tons of hdds running raid. So you dont have to worry about the backup. Since the raid should take care of that and all attached system would use the san for data.

For specifically checking new data backup programmes search google for differential backup software. Tons of tools are available for linux.

The benefit of buying only software is tat sometimes hardware vendors charge a hefty premium for their goods and if you have time it is almost always cheaper to build your own.

With seperate locations, use VPN it will work just fine as long as you do not try to run a san remotely, unless you have a really thick pipe and some pretty solid hardware to run the vpn over, fitted with a hardware encryption card. Cost of bandwidth required with vpn vs a private pipe, not sure, you will have to get in touch with some telcos/isps.

Googling "use case san vs nas" produced some interesting results