CHAPTER 29

NoSQL Databases

IN THIS CHAPTER

Key/Value   Stores

Document    Stores

Wide    Column  Stores

Graph   Stores

References

If you read Chapter 28, “Administering Relational Database Services,” you

have already read a brief description of the databases in this chapter. NoSQL
is a broad term covering a large number of database styles. There are some
similarities, but each of these databases was developed for a specific purpose.
They are therefore not necessarily interchangeable, although it might be
possible to force one to serve a task for which it is not designed. That is rarely
a good idea. Also, although there has been a lot of press and hype about
NoSQL over the past few years, NoSQL will not and should not be
considered a replacement for relational databases. Rather, this is a new set of
databases designed to excel in specific situations, especially in large-scale,
high-traffic-volume applications. If you have a need for storing and
interacting with specific types of data, as described in this chapter, only then
do we recommend using one of the databases listed in this chapter.

NOT ONE SIZE    FITS    ALL

“It’s   not a   one size    fits    all anymore.    One will    use multiple    technologies.   If  I

were    a   CTO,    I’d want    to  use NoSQL   for scalable    high    performance

operational data    access, lots    of  reads   and writes  at  high    speed   and [for]

semi    real-time   and low latency for end users.  And you need    another for

reporting   and BI. These   [NoSQL] technologies    are not optimal for that.   In

general,    a   classic data    warehouse   is  a   good    solution    for those   things.”