Database Reference
In-Depth Information
inappropriate for interactive or real-time data access purposes. Apache
HBase fills the gap in this space and is a NoSQL database built on top of
Hadoop and HDFS that provides real-time, random read/write access to
your big data.
The Apache HBase project is actually a clone modeled after the Google
BigTable project defined by Chang et al. (2006) in the paper “BigTable: A
Distributed Storage System for Structured Data.” You can review the whole
paper at http://research.google.com/archive/bigtable.html , but a summary
or overview of columnar databases in general will suffice to get you going.
NoSQL Database Types
The term NoSQL often refers to nonrelational databases. There are four
common types of NoSQL databases:
• Key/value: The simplest of the NoSQL databases, key/value
databases are essentially hash sets that consist of a unique key and a
value that is often represented as a schema-less blob.
• Document: Similar to key/value databases, document databases
contain structured documents (such as XML, JSON, or even HTML)
in place of the schema-less blob. These systems usually provide
functionality to search within the stored documents.
• Columnar: Instead of storing data in a row/column approach, data
in a columnar database is organized by column families which are
groups of related columns, as discussed in more detail in the
following section.
• Graph: The graph database consists of entities and edges, which
represent relationships between nodes. The relationships between
nodes can contain properties, which include items like direction of
the relationship. This type of NoSQL database is commonly used to
traverse organization or social network data.
Search WWH ::




Custom Search