The last storage hierachy in modern computers has to consists of durable,
permanent storage devices that can keep the data over a long period of time.
Obviously, this hierachy has the slowes access time and the larges capacity.
Nowadays, magnetic hard disks are used to store the large amounts of data
permanently. But even the ever growing capacity of disk
drives
cannot satisfy
the ever growing demands for flexible and efficient storage systems. The
simplest way to provide almost indefinite storage capacity is the use of disk
drives in parallel. The underlying architecture of such a
storage
network can vary significantly. All we require is a number of storage devices
(e.g. disk drives) that are connected and can be accessed from the outside.
In this paper we give an overview of existing techniques and implementations
tackling
the problems arising in such systems. Starting point will be an introduction
of the problems originated in storage systems, like data distribution or
data access balance. This leads directly to the definition and discussion of
quality factors imposed on these networks. Using these factors will enable us
to compare rather different systems even if a
practical comparison is not possible. We describe the main ideas underlying
the introduced systems and analyse them using the defined quality measures.