Hoping someone can shed some light for me.
Suppose there was a web application (A) that needed a separate image server
(I). The image server (I) will need huge amount of storage and need to be
expanded later. I'm thinking to just map a drive on A that points to image
server I's storage folder. The web application A will take user-uploaded
files and place them in the shared drive, which points to I. For access,
though, visitors will goto the image server I's web server, and not A. This
is done for load balancing purposes.
Going a step further, suppose there was a need for many image servers for
load balancing, so the storage of the image files will need to be on a
dedicated storage server (separate from the image servers).
What would you propose for the storage server in this case? Obviously if
traffic picks up the load will increase, perhaps to a point where it becomes
unacceptable. Therefore I'll need to be able to add more CPU power to the
storage server, or array.
|