Apache Cassandra

Apache Cassandra is a free and open-source distributed NoSQL database management system designed to handle large amounts of data across many commodity servers , providing high availability with no single point of failure . Cassandra offers robust support for multiple spanning datacenter clusters , [1] with asynchronous masterless replication allowing low latency operations for all clients.

Cassandra also places a high value on performance. In 2012, University of Toronto researchers studying NoSQL systems concluded that “In terms of scalability, there is a clear winner of our experiments.” Cassandra achieves the highest throughput for the maximum number of nodes in all experiments “although” this comes at the price of high write and read latencies. ” [2]


Avinash Lakshman (one of the authors of Amazon’s Dynamo ) and Prashant Malik INITIALLY Developed Cassandra at Facebook to power the Facebook inbox search feature. Facebook released Cassandra as an open-source project on Google code in July 2008. [3] In March 2009 it became an Apache Incubator project. [4] On February 17, 2010 it graduated to a top-level project. [5]

Facebook mythologic prophet Cassandra – with classical allusions to a curse on an oracle . [6]

Releases after graduation include

  • 0.6, released Apr 12 2010, added support for integrated caching, and Apache Hadoop MapReduce [7]
  • 0.7, released Jan 08 2011, added secondary indexes and online schema changes [8]
  • 0.8, released Jun 2 2011, added the Cassandra Query Language (CQL), self-tuning memtables, and support for zero-downtime upgrades [9]
  • 1.0, released Oct 17 2011, added integrated compression, leveled compaction, and improved read-performance [10]
  • 1.1, released Apr 23 2012, added self-tuning caches, row-level isolation, and support for mixed ssd / spinning disk deployments [11]
  • 1.2, released Jan 2 2013, added clustering across virtual nodes, inter-node communication, atomic batches, and request tracing [12]
  • 2.0, released Sep 4 2013, added Paxos consensus protocol, triggers, improved compactions
  • 2.1 released Sep 10 2014 [13]
  • 2.2 released July 20, 2015
  • 3.0 released November 11, 2015
  • 3.1 through 3.10 Were releases monthly releases using a tick-tock -like release model, with even-numbered releases supplier providing Both new features and bug fixes while odd-numbered releases will include bug fixes only. [14]
  • 3.11 released June 23, 2017 as a stable 3.11 release series and bug fixes from the last tick-tock feature release.
Version Original release date Latest version Release date Status [15]
0.6 2010-04-12 0.6.13 2011-04-18 No longer supported
0.7 2011-01-10 0.7.10 2011-10-31 No longer supported
0.8 2011-06-03 0.8.10 2012-02-13 No longer supported
1.0 2011-10-18 1.0.12 2012-10-04 No longer supported
1.1 2012-04-24 1.1.12 2013-05-27 No longer supported
1.2 2013-01-02 1.2.19 2014-09-18 No longer supported
2.0 2013-09-03 2.0.17 2015-09-21 No longer supported
2.1 2014-09-16 2.1.19 2017-10-05 Still supported, critical fixed only
2.2 2015-07-20 2.2.11 2017-10-05 Still supported
3.0 2015-11-09 3.0.15 2017-10-10 Still supported
3.11 2017-06-23 3.11.1 2017-10-10 Latest release
Old version
Older version, still supported
Latest version
Latest preview version

Main features

Every node in the cluster has the same role. There is no single point of failure. Data is distributed across the cluster, but there is no need for any request.
Replication and multi data center replication media
Replication strategies are configurable. [16] Cassandra is designed as a distributed system for multiple numbers of nodes across multiple data centers. Key features of Cassandra’s distributed architecture are for multiple-data center deployment, for redundancy, for failover and disaster recovery.
Designed to read and write throughput both linearly and with new or improved applications.
Data is automatically replicated to multiple nodes for fault-tolerance . Replication across multiple data centers is supported. Failed nodes can be replaced with no downtime.
Tunable consistency
Writes and reads offer a tunable level of consistency, all the way from “writes never fail” to “block for all replicas to be readable,” with the quorum level in the middle. [17]
MapReduce support
Cassandra has Hadoop integration, with MapReduce support. There is support for Apache Pig and Apache Hive . [18]
Query language
Cassandra introduced the Cassandra Query Language (CQL). CQL is a simple interface for accessing Cassandra, as an alternative to the traditional Structured Query Language (SQL). CQL adds an abstraction layer that provides a framework for this structure and provides for native syntaxes for collections and other common encodings. Language drivers are available for Java (JDBC), Python (DBAPI2), Node.JS (Helenus), Go (gocql) and C ++. [19]

Below an example of keyspace creation, including a family column in CQL 3.0: [20]

 WITH REPLICATION =  'class' : 'SimpleStrategy' , 'replication_factor' : 3  ;
USE MyKeySpace ;
CREATE COLUMNFAMILY MyColumns ( text id , Last text , First text , PRIMARY KEY ( id )); INSERT INTO MyColumns ( id , Last , First ) VALUES (
 '1' , 'Doe' , 'John' );
SELECT * FROM MyColumns ;

Which gives:

id | first | last
---- + ------- + ------
 1 | John | Doe ( 1 rows )

Known issues

Cassandra is not row level consistent, [21] meaning that inserts and updates into the table that affect the same time that affect the non-key columns in inconsistent ways. One result may be affected by one of the following effects, resulting from the fact that they were never specified or intended.

Data model

Cassandra is essentially a hybrid between a key-value and a column-oriented (or tabular) database management system. Its data model is a partitioned row store with tunable consistency. [17] Rows are organized into tables ; the first component of a table is primary key is the partition key; within a partition, rows are clustered by the remaining columns of the key. [22] Other columns may be indexed from the primary key. [23]

Tables can be created, dropped, and altered at run-time without blocking updates and queries. [24]

Cassandra can not join or subqueries . Rather, Cassandra emphasizes denormalization through features like collections. [25]

A family column (called “table” since CQL 3) resembles table in an RDBMS. Column families contain rows and columns. Each row is uniquely identified by a row key. Each row has multiple columns, each of which has a name, value, and a timestamp. Unlike a table in an RDBMS, different rows in the same column do not have to share the same set of columns, and a column can be added to one or more rows at any time. [26]

Each key in Cassandra corresponds to a value which is an object. Each key has values ​​as columns, and columns are grouped together in sets called column families. Thus, each key identifies a row of a variable number of elements. These columns could be considered then as tables. A table in Cassandra is a distributed multi-dimensional map indexed by a key. Furthermore, applications can specify the order of columns within a Super Column or Simple Column family.


When the cluster for Apache Cassandra is designed, an important point is to select the right partition. Two partitioners exist: [27]

  1. OrderPreservingPartitioner (OPP): This partitioner distributes the key-value in a similar way. The advantage is that fewer nodes have to be accessed. The drawback is the uneven distribution of the key-value peers.
  2. RandomPartitioner (RP): This partitioner randomly distributes the key-value on the network, resulting in a good load balancing. Compared to OPP, more nodes have to be reached to get a number of keys.

Management and monitoring

Cassandra is a Java-based system that can be managed and managed through Java Management Extensions (JMX). The JMX-compliant nodetool utility, for instance, can be used to manage a clustered Cassandra (adding nodes to a ring, draining nodes, decommissioning nodes, and so on). [28] Cassandra metrics pertaining to disk usage, latency, compaction, garbage collection, and more. [29] Additional metrics are available through JMX tools such as JConsole and via pluggable metrics reporters for external monitoring tools, which became available with Cassandra version 2.0.2. [30]

Prominent users

Cassandra is the most popular online store , [31] and in September 2014 surpassed Sybase to become the 9th most popular database, close behind Microsoft Access and SQLite. [32]

  • @WalmartLabs [33] (previously Kosmix ) uses Cassandra with SSD
  • Amadeus IT Group uses Cassandra for some of their back-end systems.
  • Apple uses 100,000 Cassandra nodes, as revealed at Cassandra Summit San Francisco 2015, [34] but it is not elaborated for which products, services or features.
  • AppScale uses Cassandra as a back-end for Google App Engine applications [35]
  • BlackRock uses Cassandra in their Aladdin investment management platform [36] [37]
  • CERN used Cassandra-based prototype for its ATLAS experiment to archive the DAQ system’s monitoring information [38]
  • Cisco ‘s WebEx uses Cassandra to store user feedback. [39]
  • Cloudkick uses Cassandra to store the server metrics of their users. [40]
  • Constant Contact uses Cassandra in their email and social media marketing applications. [41] Over 200 nodes are deployed.
  • Digg , a social news website broad, annoncé on Sep 9th, 2009 That It is rolling out ict use of Cassandra [42] and confirmed this on March 8, 2010. [43] TechCrunch HAS since Cassandra linked to Digg v4 reliability and recent Criticisms company struggles. [44] Lead engineers at Digg later rebuked these criticisms as red herring and blamed a lack of load testing. [45]
  • Discord uses Cassandra to store over 120 million messages per day. [46]
  • Facebook used Cassandra to power Inbox Search, with over 200 nodes deployed. [47] This was abandoned in late 2010 when they built Facebook Messaging platform on HBase as they “found Cassandra’s eventual consistency model to be a difficult pattern”. [48] Facebook moved off its pre-Apache Cassandra deployment in late 2010 when they replaced Inbox with the Facebook Messaging platform. [48] In 2012, Facebook began using Apache Cassandra in its Instagram unit. [49]
  • Formspring uses Cassandra to count responses, as well as store Social Graph data (followers, following, blockers, blocking) for 26 million accounts with 10 million responses a day [50]
  • Globo.com uses Cassandra as a back-end database for their streaming services [51]
  • IBM has done research in building a scalable email system based on Cassandra. [52]
  • Mahalo.com uses Cassandra to record user activity logs and topics for their Q & A website [53] [54]
  • Netflix uses Cassandra as their back-end database for their streaming services [55] [56]
  • Nutanix appliances use Cassandra to store metadata and stats. [57]
  • Ooyala built a scalable, flexible, real-time analytics engine using Cassandra [58]
  • Openwave uses Cassandra as distributed computing mechanism for their next generation messaging platform [59]
  • OpenX is running over 130 nodes is for Cassandra Their OpenX Enterprise product to store and replicate data for targeting advertisements and ad delivery [60]
  • Plaxo has “reviewed 3 billion contacts in [their] database, compared with 600 million unique people with contact info.” [61]
  • Plexistor for Apache Cassandra delivers high capacity storage at near-memory speed, reducing the need for expensive storage and dedicated servers. Plexistor can be used in Amazon or on Docker containers. [62]
  • PostRank used Cassandra as their backend database [63]
  • Rackspace uses Cassandra internally. [64]
  • Reddit Switched to Cassandra from memcacheDB on March 12, 2010 [65] and experienced some problems in their cluster. [66]
  • RockYou uses Cassandra to record every single click for 50 million Monthly Active Users in real-time for their online games [67]
  • SoundCloud uses Cassandra to store the dashboard of their users [68]
  • Talentica Software uses Cassandra as a back-end for Analytics Application with Cassandra cluster of 30 nodes and inserting around 200GB data on a daily basis. [69]
  • Tibbo Systems uses Cassandra as configuration and event storage for AggreGate Platform .
  • Twitter annoncé It was schedule to move Entirely from MySQL to Cassandra, [70] [71] though soon after-retracted this, keeping in MySQL while using Tweets Cassandra for analytics. [72]
  • Urban Airship uses Cassandra with the mobile service hosting for over 160 million applications installed across 80 million unique devices [73]
  • Wikimedia uses Cassandra as backend storage for its public facing REST Content API. [74]