I was talking with TeraTech’s server tuning consultant Mike Brunt about servers and SANs (Storage Area Networks). When should you use a SAN?
1/ As a SAN systems are typically simply a large number of hard-drives, optical drives etc in an enclosure, it is very easy to add capacity quickly as needed as opposed to adding a complete new server which takes time. So if you database is growing fast you might chose to put the data on a SAN.
2/ If there is a need to allow access to a large volumes of data from multiple systems. A typical case here would be large media files such as photographs-images, movies, music, games etc.
It is generally not a good idea to put CF code on a SAN to share among a cluster of servers - the network traffic will be much slower than accessing the CF code on a local hard drive. Instead use automated tools to copy updated code between servers in the cluster. More on that later.
But... when we turned trusted cache on, the differences disappeared. SAN access was virtually the same as local disk.
I'd really like to see the evidence backing up this claim that code should not go on a SAN!
I think "generally" and "much slower" are terms that should have been scrutinzed further before the post.