Database Reference
In-Depth Information
AND ts = '2013-12-01T00:00:00';
... APPLY BATCH;
Reading from Cassandra
Like inserting data, reading data from Cassandra uses SQL-like
SELECT
statements. Cassandra does not have many query options, so
SELECT
statements tend to be fairly simple.
Other Storage Technologies
This chapter has focused on a few different persistent stores popular for
real-time applications. They were chosen to show some of the range of
available stores beyond the relational database. This section briefly
mentions some of the other technologies available, but not discussed in this
book.
Relational Databases
Many relational database implementations are backed by a
high-performance data store. The performance hit introduced by
transactions are maintained at the application level rather than being
intrinsic to the database itself. Some relational databases even allow access
to these high-performance stores. For example, PostgreSQL exposes a
key-value store through its
hstore
module, which allows access through
the SQL interface. It also allows for some indexing of the key-value store.
Other databases, especially commercial column stores, are designed to
ingest large amounts of raw data that can then be efficiently queried. In
cases where the access patterns are not well defined, it might be best to
stream data into a column store for access. The downside is usually related
to the cost required to achieve acceptable ingest performance, but this is
often the case when balancing query flexibility and performance.
Distributed In-Memory Data Grids
Data grids have been available in various forms for a number of years. They
aremostlymarketedascommercialproductsfromwell-knownvendorssuch
