Posted 29 November 2012 - 08:16 AM
SAN can be in general anything - from JBOD units to OS Based block sharing over iSCSI, iFC etc... So it is rather difficult to answer simply.
In a pure hardware form SAN is a Network - this means many devices are concerned.
1 Device by itself cannot form a SAN.
You could build a SAN using FC, SAS, and TCP based implementations like TCP over InfiniBand, iSCSI in 1G and 10G Ethernet...
10 years ago SAN was limited to Fibre Channel
so in this form all you needed was a firmware and a SAN switch. Now however you have the so-called hybrid storage which has a SAN part (Nexenta, OpenFiler, even typical Server OS with iSCSI Target installed could be considered SAN member...) and a NAS part (Sharing files over SMB/CIFS ; NFS ; AFP ...).
The difference is that in a SAN you have a number of blocks to be accessed which then you have to control using the client to form a volume and install an file system over it, whereas in a NAS application, the Filesystem and OS should not matter - it is a different protocol which is in the base of the whole thing.
In Windows Networks the network file sharing protocol is calles System Message Block (SMB) also known as Common Internet File System (CIFS), and the linux daemon for this type of access is usually samba or smbd.
So go back to my first post - there is not a simple way to explain SAN and NAS with diagrams and examples. You have to get the whole picture to know what is the one and what is the other.
What you see above is my personal opinion. Don't take it as the holy bible and the one-and-only truth :)