Apache Cassandra vs Apache HBase

In the last article we have compared Cassandra with MongoDB. That comparison made it obvious that we need to compare Cassandra with HBase, and MongoDB with HBase to complete the understanding. Cassandra and HBase are similar at many things but Cassandra’s consistency model is like DynamoDB to provide eventual consistency.

HBase is a part of HDFS which runs on top of Hadoop (based on Zookeeper). HBase is not a typical database. Cassandra is a distributed database. On a distributed database, data can be accessed by a client from any cluster or node. Cassandra was developed knowing there is HBase. All the nodes work as an Active Node in case of Cassandra whereas HBase follows Master-Slave Node model. Cassandra supports JavaScript and Ruby while HBase does not support them.

Hbase itself a key/value store which allows range scans of keys. We need to write code to access Hbase and it does not use some SQL-like language like Cassandra. Apache Phoenix is Hbase with SQL-like language. So it is obvious, expected to find about Apache Phoenix and compare with Cassandra. Apache Phoenix is Hadoop dependent while Cassandra is purposefully built to make easier to install, configure and manage.

Cassandra and HBase have different deployment profiles, so the answer which is more suitable in a particular situation that depends on the nature of the task. Cassandra’s write performance, consistency and replication capability are attractive. However, HBase is likely a better possibly a better choice for larger-scale scans to those who want a tight coupling with Hadoop. More companies are supporting the development of HBase than does Cassandra. Yet, Cassandra is more popular and widely used.

Both Cassandra and HBase are similarly licensed NoSQL database. Both databases can manage images, videos, audio, etc i.e. both Cassandra and HBase are made for Big Data. HBase is designed for the Data warehouse whereas Cassandra is designed for running web and mobile applications.

This Article Has Been Shared 8822 Times!

Facebook Twitter Google+ Pinterest