Social Locality in Data Storage - Data Storage for Social Networks: A Socially Aware Approach - page 15

is limited by some upper bound. Therefore, there are different design problems
depending on the specific requirements of the storage system under consideration.
For a system that requires perfect social locality, we may want to minimize the
number of replicas needed. For a system that has a fixed disk budget for the replicas,
and hence, imperfect social locality is the only choice, we may want to maximize
the (partial) preservation of social locality. In either case, server workloads (read
and write) and load balancing are additional factors that need to be addressed.
In the following sections, we discuss a select set of optimization problems for
socially aware storage. We start in the next section with some assumptions and
notations.

2.2 Assumptions and Notations

Consider a storage system with M servers to store data for a social graph of
N users. We assume that each user's data is of small size and as such it is
desirable to minimize the number of servers required to access a given number
of data records. Illustrated in Fig. 2.2 , we assume a three-tier system architecture,

Fig. 2.2 Three-tier storage system architecture