 | This article'suse ofexternal links may not follow Wikipedia's policies or guidelines. Pleaseimprove this article by removingexcessive orinappropriate external links, and converting useful links where appropriate intofootnote references.(October 2023) (Learn how and when to remove this message) |
| Apache Cassandra | |
|---|---|
 | |
| Original authors | Avinash Lakshman, Prashant Malik /Facebook |
| Developer | Apache Software Foundation |
| Initial release | July 2008; 17 years ago (2008-07) |
| Stable release | |
| Repository | |
| Written in | Java |
| Operating system | Cross-platform |
| Available in | English |
| Type | NoSQLDatabase,data store |
| License | Apache License 2.0 |
| Website | cassandra |
Apache Cassandra is afree and open-sourcedatabase management system designed to handle large volumes of data across multiplecommodity servers. The system prioritizes availability andscalability overconsistency, making it particularly suited for systems with high write throughput requirements due to itsLSM tree indexing storage layer.[2] As awide-column database, Cassandra supports flexible schemas and efficiently handles data models with numerous sparse columns. The system is optimized for applications with well-defined data access patterns that can be incorporated into the schema design.[2] Cassandra supportscomputer clusters which may span multipledata centers,[3] featuringasynchronous and masterless replication. It enableslow-latency operations for all clients and incorporatesAmazon'sDynamodistributed storage and replication techniques, combined withGoogle'sBigtable data storage engine model.[4]
Avinash Lakshman, a co-author ofAmazon'sDynamo, and Prashant Malik developed Cassandra atFacebook to support theinboxsearch functionality. Facebook released Cassandra as open-source software onGoogle Code in July 2008.[5] In March 2009, it became an Apache Incubator project[6] and on February 17, 2010, it graduated to a top-level project.[7]
The developers atFacebook named their database afterCassandra, themythologicalTrojan prophetess, referencing her curse of making prophecies that were never believed.[8]
Cassandra uses adistributed architecture where all nodes perform identical functions, eliminating single points of failure. The system employs configurable replication strategies to distribute data across clusters, providing redundancy and disaster recovery capabilities. The system is capable of linear scaling, which increases read and write throughput with the addition of new nodes, while maintaining continuous service.
Cassandra is categorized as an AP (Availability and Partition Tolerance) system, emphasizing availability and partition tolerance overconsistency. While it offers tunable consistency levels for both read and write operations, its architecture makes it less suitable for use cases requiring strict consistency guarantees.[2] Additionally, Cassandra's compatibility withHadoop and related tools allows for integration with existing big data processing workflows. Eventual consistency is maintained usingtombstones to manage reads,upserts, and deletes.
The system's query capabilities have notable limitations. Cassandra does not support advanced query patterns such as multi-tableJOINs, ad hoc aggregations, or complex queries.[2] These limitations stem from its distributed architecture, which optimizes for scalability and availability rather than complex query operations.
As awide-column store, Cassandra combines features of both key-value and tabular database systems. It implements a partitioned row store model with adjustable consistency levels.[9] The following table compares Cassandra andrelational database management systems (RDBMS).
| Feature | Cassandra | RDBMS |
|---|---|---|
| Organization | Keyspace → Table → Row | Database → Table → Row |
| Row Structure | Dynamic columns | Fixed schema |
| Column Data | Name, type, value, timestamp | Name, type, value |
| Schema Changes | Runtime modifications | Usually requires downtime |
| Data Model | Denormalized | Normalized with JOINs |
The data model consists of several hierarchical components:
A keyspace in Cassandra is analogous to a database inrelational systems. It contains multiple tables and manages configuration information, including replication strategy and user-defined types (UDTs).[2]
Tables (formerly calledcolumn families prior to CQL 3) are containers for rows of data. Each table has a name and configuration information for its stored data. Tables may be created, dropped, or altered at run-time without blockingupdates and queries.[10]
Each row is identified by aprimary key and contains columns. The first component of a table's primary key is the partition key; within a partition, rows areclustered by the remaining columns of the key.[11]
Columns contain data belonging to a row and consist of:
Unlike traditional RDBMS tables, rows within the same table can have varying columns, providing a flexible structure. This flexibility distinguishes Cassandra from relational databases, as not all columns need to be specified for each row.[2] Other columns may be indexed separately from the primary key.[12]
Cassandra uses aLog Structured Merge Tree (LSM tree) index to optimize write throughput, in contrast to theB-tree indexes used by most databases.[2]
| Feature | Cassandra | RDBMS |
|---|---|---|
| Index Structure | LSM Tree | B-Tree |
| Write Process | Append-only with Memtable | In-place updates |
| Storage Components | Commit Log, Memtable, SSTable | Data files, Transaction Log |
| Update Strategy | New entry for each change | Modify existing data |
| Delete Handling | Tombstone markers | Direct removal |
| Read Optimization | Secondary | Primary |
| Write Optimization | Primary | Secondary |
The storage architecture consists of three main components:[2]
Write operations follow a two-stage process:
Read operations:
Every operation (create/update/delete) generates a new entry, with deletes handled via "tombstones". While common in many databases, tombstones can cause performance degradation in delete-heavy workloads.[13]
Compaction consolidates multiple SSTables to:
Cassandra Query Language (CQL) is the interface for accessing Cassandra, as an alternative to the traditionalStructured Query Language (SQL). CQL adds anabstraction layer that hides implementation details of this structure and provides native syntaxes for collections and other common encodings. Language drivers are available forJava (JDBC),Python (DBAPI2),Node.JS (DataStax),Go (gocql), andC++.[14]
The key space in Cassandra is a namespace that defines data replication across nodes. Therefore, replication is defined at the key space level. Below is an example of key space creation, including a column family in CQL 3.0:[15]
CREATEKEYSPACEMyKeySpaceWITHREPLICATION={'class':'SimpleStrategy','replication_factor':3};USEMyKeySpace;CREATECOLUMNFAMILYMyColumns(idtext,lastNametext,firstNametext,PRIMARYKEY(id));INSERTINTOMyColumns(id,lastName,firstName)VALUES('1','Doe','John');SELECT*FROMMyColumns;
Which gives:
id | lastName | firstName----+----------+---------- 1 | Doe | John(1 rows)
Cassandra uses a peer-to-peer gossip protocol for cluster communication. Nodes routinely exchange information about cluster state, including:
The system usesvector clocks to track information currency and ignore outdated state data.[2]
The architecture designates certain nodes as "seed" nodes that:
This design eliminates single points of failure while maintaining cluster-wide consistency of operational knowledge.[2]
Cassandra employs the Phi Accrual Failure Detector to manage node failures during cluster operation.[16] Through this system, each node independently assesses the availability of other nodes during gossip communication. When a node fails to respond, it is "convicted" and removed from write operations, though it can rejoin the cluster upon resuming heartbeat signals.[2]
To maintain data integrity during node outages, Cassandra uses a "hinted handoff" mechanism. When writing to an offline node, the coordinator node temporarily stores the write data as a "hint." Once the offline node returns to service, these hints are forwarded to restore data consistency. Notably, Cassandra only permanently removes nodes through explicit administrative decommissioning or rebuilding, preventing temporary communication failures or restarts from triggering unnecessary data rebalancing.[2]
Cassandra is a Java-based system that can be managed and monitored viaJava Management Extensions (JMX). The JMX-compliantNodetool utility, for instance, can be used to manage a Cassandra cluster.[17] Nodetool also offers a number of commands to return Cassandra metrics pertaining to disk usage, latency, compaction, garbage collection, and more.[18]
Since the release of Cassandra 2.0.2 in 2013, measures of several metrics are produced via the Dropwizard metrics framework,[19] and may be queried via JMX using tools such asJConsole or passed to external monitoring systems via Dropwizard-compatible reporter plugins.[20]
Releases after graduation include:
| Version | Original release date | Latest version | Release date | Status[21] |
|---|---|---|---|---|
| Unsupported: 0.6 | 2010-04-12 | 0.6.13 | 2011-04-18 | No longer maintained |
| Unsupported: 0.7 | 2011-01-10 | 0.7.10 | 2011-10-31 | No longer maintained |
| Unsupported: 0.8 | 2011-06-03 | 0.8.10 | 2012-02-13 | No longer maintained |
| Unsupported: 1.0 | 2011-10-18 | 1.0.12 | 2012-10-04 | No longer maintained |
| Unsupported: 1.1 | 2012-04-24 | 1.1.12 | 2013-05-27 | No longer maintained |
| Unsupported: 1.2 | 2013-01-02 | 1.2.19 | 2014-09-18 | No longer maintained |
| Unsupported: 2.0 | 2013-09-03 | 2.0.17 | 2015-09-21 | No longer maintained |
| Unsupported: 2.1 | 2014-09-16 | 2.1.22 | 2020-08-31 | No longer maintained |
| Unsupported: 2.2 | 2015-07-20 | 2.2.19 | 2020-11-04 | No longer maintained |
| Unsupported: 3.0 | 2015-11-09 | 3.0.29 | 2023-05-15 | No longer maintained |
| Unsupported: 3.11 | 2017-06-23 | 3.11.15 | 2023-05-05 | No longer maintained |
| Supported: 4.0 | 2021-07-26 | 4.0.18 | 2025-05-28 | Maintained until 5.1.0 release |
| Supported: 4.1 | 2022-06-17 | 4.1.9 | 2025-05-19 | Maintained until 5.2.0 release |
| Latest version:5.0 | 2024-09-05 | 5.0.5 | 2025-08-05 | Latest release. Maintained until 5.3.0 release |
Legend: Unsupported Supported Latest version Preview version Future version | ||||
{{cite web}}:Missing or empty|title= (help)Cassandra's innate datacenter concepts are important as they allow multiple workloads to be run across multiple datacenters...
Apache Cassandra is named after the Greek mythological prophet Cassandra. [...] Because of her beauty Apollo granted her the ability of prophecy. [...] When Cassandra of Troy refused Apollo, he put a curse on her so that all of her and her descendants' predictions would not be believed. [...] Cassandra is the cursed Oracle[.]
