Database Reference
In-Depth Information
services, if they're provided at all. In many ways, Redis is more closely
related to caching systems like Memcached than it is a database.
Unlike caching systems, Redis provides server-side support for high-level
data structures with atomic update capability. The basic structures are lists,
sets, hash tables, and sorted sets in addition to the basic key-value
functionality. It also includes the capability to expire keys and a
publish-subscription mechanism that can be used as a messaging bus
between, for instance, the real-time streaming processing system and a
front end. Later chapters provide more detail, but when keys are updated a
front-end server can be notified of the event so that the user interface may
be updated.
Getting Set Up
On most systems, Redis is available through the standard package manager.
However, if Redis is out-of-date or otherwise unavailable through a package
manager, it is easy to build it on Unix-like systems. After downloading the
appropriate archive, the usual
make && make install
will build a
suitable version of Redis. By default, this build is 64 bit, but it is possible to
build a 32-bit version. Building a 32-bit version will save some memory, but
limits Redis' total footprint to 4GB.
NOTE
For users of Amazon's EC2 service, Redis is now included as part of the
Elasticache. The version usually lags the most recent release, but it
makes it really easy to get started using Redis in either a development
or a production environment. Just select redis from the drop-down (it
shows memcache by default).
Redis ships with a fairly good set of default configuration parameters, but
there are some options that should be considered when using Redis. The
configuration entries are usually held in the
redis.conf
file passed to the
server when it starts.
