Apache Cassandra

Apache Cassandra is an open source distributed 1.0, released Oct 17 2011, added integrated com-
database management system designed to handle large pression, leveled compaction, and improved read
amounts of data across many commodity servers, provid- performance[14]
ing high availability with no single point of failure. Cas-
sandra oers robust support for clusters spanning multi- 1.1, released Apr 23 2012, added self-tuning
ple datacenters,[1] with asynchronous masterless replica- caches, row-level isolation, and support for mixed
tion allowing low latency operations for all clients. ssd/spinning disk deployments[15]

Cassandra also places a high value on performance. In 1.2, released Jan 2 2013, added clustering across
2012, University of Toronto researchers studying NoSQL virtual nodes, inter-node communication, atomic
systems concluded that In terms of scalability, there is batches, and request tracing[16]
a clear winner throughout our experiments. Cassandra
achieves the highest throughput for the maximum number 2.0, released Sep 4 2013, added lightweight transac-
of nodes in all experiments although this comes at the tions (based on the Paxos consensus protocol), trig-
price of high write and read latencies.[2] gers, improved compactions
Cassandras data model is a partitioned row store with 2.0.4, released Dec 30 2013, added allowing spec-
tunable consistency.[3] Rows are organized into tables; the ifying datacenters to participate in a repair, client
rst component of a tables primary key is the partition encryption support to sstableloader, allow removing
key; within a partition, rows are clustered by the remain- snapshots of no-longer-existing CFs[17]
ing columns of the key.[4] Other columns may be indexed
separately from the primary key.[5] 2.1.0 released Sep 10 2014 [18]
Tables may be created, dropped, and altered at runtime
without blocking updates and queries.[6]
2 Licensing and support
Cassandra does not support joins or subqueries, except
for batch analysis via Hadoop. Rather, Cassandra empha-
sizes denormalization through features like collections.[7] Apache Cassandra is an Apache Software Foundation
project, so it has an Apache License (version 2.0).

1 History 3 Main features

Apache Cassandra was initially developed at Facebook to Decentralized Every node in the cluster has the same
power their Inbox Search feature by Avinash Lakshman role. There is no single point of failure. Data is
(one of the authors of Amazons Dynamo) and Prashant distributed across the cluster (so each node contains
Malik. It was released as an open source project on dierent data), but there is no master as every node
Google code in July 2008.[8] In March 2009, it became can service any request.
an Apache Incubator project.[9] On February 17, 2010 it
graduated to a top-level project.[10]
Supports replication and multi data center replication
Releases after graduation include Replication strategies are congurable.[19] Cas-
sandra is designed as a distributed system, for
0.6, released Apr 12 2010, added support for inte- deployment of large numbers of nodes across
grated caching, and Apache Hadoop MapReduce [11] multiple data centers. Key features of Cassandras
distributed architecture are specically tailored for
0.7, released Jan 08 2011, added secondary indexes multiple-data center deployment, for redundancy,
and online schema changes [12] for failover and disaster recovery.

0.8, released Jun 2 2011, added the Cassandra Scalability Read and write throughput both increase lin-
Query Language (CQL), self-tuning memtables, early as new machines are added, with no downtime
and support for zero-downtime upgrades[13] or interruption to applications.

1
2 6 PROMINENT USERS

Fault-tolerant Data is automatically replicated to mul- 1. RandomPartitioner (RP): This partitioner randomly
tiple nodes for fault-tolerance. Replication across distributes the key-value pairs over the network, re-
multiple data centers is supported. Failed nodes can sulting in a good load balancing. Compared to OPP,
be replaced with no downtime. more nodes have to be accessed to get a number of
keys.
Tunable consistency Writes and reads oer a tunable
2. OrderPreservingPartitioner (OPP): This partitioner
level of consistency, all the way from writes never
distributes the key-value pairs in a natural way so
fail to block for all replicas to be readable, with
that similar keys are not far away. The advantage is
the quorum level in the middle.[3]
that fewer nodes have to be accessed. The drawback
is the uneven distribution of the key-value pairs.
MapReduce support Cassandra has Hadoop integra-
tion, with MapReduce support. There is support
also for Apache Pig and Apache Hive.[20]
6 Prominent users
Query language Cassandra introduces CQL
@WalmartLabs[24] (previously Kosmix) uses Cas-
(Cassandra Query Language), a SQL-like alter-
sandra with SSD
native to the traditional RPC interface. Language
drivers are available for Java (JDBC), Python Apple uses 75,000 Cassandra nodes, as revealed at
(DBAPI2), Node.JS (Helenus), Go (gocql) and Cassandra Summit San Francisco 2014,[25] although
C++.[21] it has not elaborated for which products, services or
features.

4 Data model AppScale uses Cassandra as a back-end for Google

App Engine applications[26]
Cassandra is essentially a hybrid between a key-value and CERN uses Cassandra for its ATLAS experiment
a column-oriented (or tabular) database. to archive the online DAQ systems monitoring
information[27]
A column family (called table since CQL 3)
resembles a table in an RDBMS. Column fam- Cisco's WebEx uses Cassandra to store user feed
ilies contain rows and columns. Each row is and activity in near real time.[28]
uniquely identied by a row key. Each row has
Cloudkick uses Cassandra to store the server metrics
multiple columns, each of which has a name,
of their users.[29]
value, and a timestamp. Unlike a table in an
RDBMS, dierent rows in the same column Constant Contact uses Cassandra in their email and
family do not have to share the same set of social media marketing applications.[30] Over 200
columns, and a column may be added to one nodes are deployed.
or multiple rows at any time.[22]
Digg, a large social news website, announced on
Each key in Cassandra corresponds to a value which is an Sep 9th, 2009 that it is rolling out its use of
object. Each key has values as columns, and columns are Cassandra[31] and conrmed this on March 8,
grouped together into sets called column families. 2010.[32] TechCrunch has since linked Cassandra to
Digg v4 reliability criticisms and recent company
Thus, each key identies a row of a variable number of el-
struggles.[33] Lead engineers at Digg later rebuked
ements. These column families could be considered then
these criticisms as red herring and blamed a lack of
as tables. A table in Cassandra is a distributed multi di-
load testing.[34]
mensional map indexed by a key.
Furthermore, applications can specify the sort order of Facebook used Cassandra to power Inbox Search,
columns within a Super Column or Simple Column fam- with over 200 nodes deployed.[35] This was aban-
ily. doned in late 2010 when they built Facebook Mes-
saging platform on HBase as they found Cassan-
dras eventual consistency model to be a dicult
pattern.[36]
5 Clustering
Formspring uses Cassandra to count responses, as
When the cluster for Apache Cassandra is designed, an well as store Social Graph data (followers, following,
important point is to select the right partitioner. Two par- blockers, blocking) for 26 Million accounts with 10
titioners exist:[23] million responses a day[37]
 3

IBM has done research in building a scalable email Facebook moved o its pre-Apache Cassandra deploy-
system based on Cassandra.[38] ment in late 2010 when they replaced Inbox Search with
the Facebook Messaging platform.[36] In 2012, Facebook
Mahalo.com uses Cassandra to record user activity began using Apache Cassandra in its Instagram unit.[59]
logs and topics for their Q&A website[39][40]
Cassandra is the most popular wide column store,[60] and
Netix uses Cassandra as their back-end database in September 2014 surpassed Sybase to become the 9th
for their streaming services[41][42] most popular database, close behind Microsoft Access
and SQLite.[61]
Ooyala built a scalable, exible, real-time analytics
engine using Cassandra[43]

Openwave uses Cassandra as a distributed database 7 See also

and as a distributed storage mechanism for their next
generation messaging platform[44]
7.1 Academic Background
OpenX is running over 130 nodes on Cassandra
for their OpenX Enterprise product to store and BigTable - Original distributed database by Google
replicate advertisements and targeting data for ad
delivery[45] Distributed database

Plaxo has reviewed 3 billion contacts in [their]

Distributed hash table (DHT)
database, compared them with publicly available
data sources, and identied approximately 600 mil-
Dynamo (storage system) - Cassandra borrows
lion unique people with contact info.[46]
many elements from Dynamo
PostRank used Cassandra as their backend
database[47] NoSQL

Rackspace is known to use Cassandra internally.[48]

Reddit switched to Cassandra from memcacheDB 7.2 Popular Alternatives

on March 12, 2010[49] and experienced some prob-
lems in May due to insucient nodes in their Aerospike
[50]
cluster.
Apache HBase - Apache Hadoop based distributed
RockYou uses Cassandra to record every single click database. Very similar to BigTable
for 50 million Monthly Active Users in real-time for
their online games[51] MongoDB
SoundCloud uses Cassandra to store the dashboard
of their users[52]
7.3 Other Alternatives
Talentica Software uses Cassandra as a back-end
for Analytics Application with Cassandra cluster of Apache Accumulo - Secure Apache Hadoop based
30 nodes and inserting around 200GB data on daily distributed database.
basis.[53]
Berkeley DB
Twitter announced it was planning to move entirely
from MySQL to Cassandra,[54][55] though soon af-
ter retracted this, keeping Tweets in MySQL while Druid (open-source data store)
using Cassandra for analytics.[56]
Hypertable - Apache Hadoop based distributed
Urban Airship uses Cassandra with the mobile ser- database. Very similar to BigTable
vice hosting for over 160 million application installs
across 80 million unique devices[57] Riak

Wikimedia uses Cassandra as backend storage for

its public-facing REST Content API.[58]
7.4 Commercial Companies
Zoho uses Cassandra for generating the inbox pre-
view in their Zoho#Zoho Mail service DataStax
4 7 SEE ALSO

7.5 References [17] Eric Evans. "[Cassandra-User] [RELEASE] Apache Cas-

sandra 2.0.4. qnalist.com. Retrieved 11 December 2014.
[62]
[18] Sylvain Lebresne (10 September 2014). "[VOTE
SUCCESS] Release Apache Cassandra 2.1.0. mail-
[1] Casares, Joaquin (2012-11-05). Multi-datacenter Repli- archive.com. Retrieved 11 December 2014.
cation in Cassandra. DataStax. Retrieved 2013-07-25.
Cassandras innate datacenter concepts are important as [19] Deploying Cassandra across Multiple Data Centers.
they allow multiple workloads to be run across multiple DataStax. Retrieved 11 December 2014.
datacenters
[20] Hadoop Support article on Cassandras wiki
[2] Rabl, Tilmann; Sadoghi, Mohammad; Jacobsen, Hans-
Arno; Villamor, Sergio Gomez-; Mulero -, Victor [21] DataStax C/C++ Driver for Apache Cassandra. DataS-
Muntes; Mankovskii, Serge (2012-08-27). Solving Big tax. Retrieved 15 December 2014.
Data Challenges for Enterprise Application Performance [22] DataStax. Apache Cassandra 0.7 Documentation - Col-
Management. VLDB. Retrieved 2013-07-25. In terms umn Families. Apache Cassandra 0.7 Documentation.
of scalability, there is a clear winner throughout our ex- Retrieved 29 October 2012.
periments. Cassandra achieves the highest throughput for
the maximum number of nodes in all experiments... this [23] Williams, Dominic. Cassandra: RandomPartitioner
comes at the price of high write and read latencies vs OrderPreservingPartitioner. http://wordpress.com/:
WordPress.com. Retrieved 2011-03-23. When building a
[3] DataStax (2013-01-15). About data consistency. Re- Cassandra cluster, the key question (sorry, thats weak)
trieved 2013-07-25. is whether to use the RandomPartitioner (RP), or the Or-
[4] Ellis, Jonathan (2012-02-15). Schema in Cassandra derPreservingPartitioner (OPP). These control how your
1.1. DataStax. Retrieved 2013-07-25. data is distributed over your nodes. Once you have chosen
your partitioner, you cannot change without wiping your
[5] Ellis, Jonathan (2010-12-03). Whats new in Cassandra data, so think carefully! The problem with OPP: If the
0.7: Secondary indexes. DataStax. Retrieved 2013-07- distribution of keys used by individual column families
25. is dierent, their sets of keys will not fall evenly across
the ranges assigned to nodes. Thus nodes will end up
[6] Ellis, Jonathan (2012-03-02). The Schema Management storing preponderances of keys (and the associated data)
Renaissance in Cassandra 1.1. DataStax. Retrieved corresponding to one column family or another. If as is
2013-07-25. likely column families store diering quantities of data
with their keys, or store data accessed according to dif-
[7] Lebresne, Sylvain (2012-08-05). Coming in 1.2: Col- fering usage patterns, then some nodes will end up with
lections support in CQL3. DataStax. Retrieved 2013- disproportionately more data than others, or serving more
07-25. hot data than others.
[8] Hamilton, James (July 12, 2008). Facebook Releases [24] "@WalmartLabs. walmartlabs.com. Retrieved 11 De-
Cassandra as Open Source. Retrieved 2009-06-04. cember 2014.
[9] Is this the new hotness now?". Mail-archive.com. 2009- [25] Lev Walker: A denition of Massive scale from Apple.
03-02. Archived from the original on 25 April 2010. Re- #CassandraSummit on Twitter
trieved 2010-03-29.
[26] Datastores on Appscale.
[10] Cassandra is an Apache top level project. Mail-
archive.com. 2010-02-18. Archived from the original on [27] A Persistent Back-End for the ATLAS Online Informa-
28 March 2010. Retrieved 2010-03-29. tion Service (P-BEAST)".

[11] The Apache Software Foundation Announces Apache [28] Re: Cassandra users survey. Mail-archive.com. 2009-
Cassandra Release 0.6 : The Apache Software Founda- 11-21. Archived from the original on 17 April 2010. Re-
tion Blog trieved 2010-03-29.

[12] The Apache Software Foundation Announces Apache [29] 4 Months with Cassandra, a love story |Cloudkick, man-
Cassandra 0.7 : The Apache Software Foundation Blog age servers better

[13] [Cassandra-user] [RELEASE] 0.8.0 - Grokbase [30] Finley, Klint (2011-02-18). This Week in Consolidation:
HP Buys Vertica, Constant Contact Buys Bantam Live and
[14] Cassandra 1.0.0. Is Ready for the Enterprise More. Read Write Enterprise.
[15] The Apache Software Foundation Announces Apache [31] Eure, Ian. Looking to the future with Cassandra.
Cassandra v1.1 : The Apache Software Foundation
Blog [32] Quinn, John. Saying Yes to NoSQL; Going Steady with
Cassandra.
[16] The Apache Software Foundation Announces Apache
Cassandra v1.2 : The Apache Software Foundation [33] Schonfeld, Erick. As Digg Struggles, VP Of Engineering
Blog. apache.org. Retrieved 11 December 2014. Is Shown The Door.
 5

[34] Is Cassandra to Blame for Digg v4s Failures?". [56] King, Ryan (2010-07-10). Cassandra at Twitter Today.
blog.twitter.com. San Fransisco, CA, USA: Twitter. Re-
[35] Niet compatibele browser. Facebook. Retrieved 2010- trieved 2014-06-20.
03-29.
[57] Onnen, Erik. From 100s to 100s of Millions.
[36] Muthukkaruppan, Kannan. The Underlying Technology
of Messages. [58] Wicke, Gabriel. Wikimedia REST content API is now
available in beta.
[37] Cozzi, Martin (2011-08-31). Cassandra at Formspring.
[59] Rick Branson (2013-06-26). Cassandra at Instagram.
[38] BlueRunner: Building an Email Service in the Cloud.
DataStax. Retrieved 2013-07-25.
ieee.org. 2009-07-20. Retrieved 2010-03-29.

[39] Mahalo.com powered by Apache Cassandra". DataS- [60] DB-Engines. DB-Engines Ranking of Wide Column
tax.com. Santa Clara, CA, USA: DataStax. 2012-04-10. Stores.
Retrieved 2014-06-13.
[61] DB-Engines. DB-Engines Ranking.
[40] Watch Cassandra at Mahalo.com |DataStax Episodes |Blip
[62] Introduction to Cassandra Architecture. Edureka.co.
[41] Cockcroft, Adrian (2011-07-11). Migrating Netix from Retrieved 11 December 2014.
Datacenter Oracle to Global Cassandra. slideshare.net.
Retrieved 2014-06-13.

[42] Izrailevsky, Yury (2011-01-28). NoSQL at Netix. 8 Bibliography

[43] Ooyala (2010-05-18). Designing a Scalable Database for
Online Video Analytics. DataStax.com. Mountain View
Hewitt, Eben (December 15, 2010). Cassandra:
CA, USA. Retrieved 2014-06-14. The Denitive Guide (1st ed.). O'Reilly Media. p.
300. ISBN 978-1-4493-9041-9.
[44] Mainstay LLC (2013-11-11). DataStax Case Study of
Openwave Messaging. DataStax.com. Santa Clara, CA, Capriolo, Edward (July 15, 2011). Cassandra High
USA: DataStax. Retrieved 2014-06-15. Performance Cookbook (1st ed.). Packt Publishing.
p. 324. ISBN 1-84951-512-3.
[45] Ad Serving Technology - Advanced Optimization, Fore-
casting, & Targeting |OpenX

[46] Smalley, Preston (2011-03-20). An important milestone 9 External links

- and its only the beginning!".

[47] Grigorik, Ilya (2011-03-29). Webpulp TV: Scaling Pos- Lakshman, Avinash (2008-08-25). Cassandra - A
tRank with Ilya Grigorik. structured storage system on a P2P Network. En-
gineering @ Facebooks Notes. Retrieved 2014-06-
[48] Hadoop and Cassandra (at Rackspace)". Stu Hood.
17.
2010-04-23. Retrieved 2011-09-01.

[49] david [ketralnis] (2010-03-12). whats new on reddit: The Apache Cassandra Project. Forest Hill, MD,
She who entangles men. blog.reddit. Archived from the USA: The Apache Software Foundation. Retrieved
original on 25 March 2010. Retrieved 2010-03-29. 2014-06-17.

[50] Posted by the reddit admins at (2010-05-11). blog.reddit Project Wiki. Forest Hill, MD, USA: The Apache
-- whats new on reddit: reddits May 2010 State of the Software Foundation. Retrieved 2014-06-17.
Servers report. blog.reddit. Archived from the original
on 14 May 2010. Retrieved 2010-05-16. Hewitt, Eben (2010-12-01). Adopting Apache
Cassandra. infoq.com. InfoQ, C4Media Inc. Re-
[51] Pattishall, Dathan Vance (2011-03-23). Cassandra is my
NoSQL Solution but.
trieved 2014-06-17.

[52] Cassandra at SoundCloud. Lakshman, Avinash; Malik, Prashant (2009-08-15).

Cassandra - A Decentralized Structured Storage
[53] cite web|url=http://www.talentica.com[] System. cs.cornell.edu. The authors are from
Facebook. Retrieved 2014-06-17.
[54] Popescu, Alex. Cassandra @ Twitter: An Interview with
Ryan King. myNoSQL. Archived from the original on 1
Ellis, Jonathan (2009-07-29). What Every De-
March 2010. Retrieved 2010-03-29.
veloper Should Know About Database Scalabil-
[55] Babcock, Charles. Twitter Drops MySQL For Cassandra ity. slideshare.net. Retrieved 2014-06-17. From
- Cloud databases. InformationWeek. Archived from the the OSCON 2009 talk on RDBMS vs. Dynamo,
original on 2 April 2010. Retrieved 2010-03-29. BigTable, and Cassandra.
6 9 EXTERNAL LINKS

Cassandra-RPM - Red Hat Package Manager

(RPM) build for the Apache Cassandra project.
code.google.com. Menlo Park, CA, USA: Google
Project Hosting. Retrieved 2014-06-17.
Roth, Gregor (2012-10-14). Cassandra by ex-
ample - the path of read and write requests.
slideshare.net. Retrieved 2014-06-17.

Mansoor, Umer (2012-11-04). A collection of

Cassandra tutorials. Retrieved 2015-02-08.

Bushik, Sergey (2012-10-22). A vendor-

independent comparison of NoSQL databases: Cas-
sandra, HBase, MongoDB, Riak. NetworkWorld.
Framingham, MA, USA and Staines, Middlesex,
UK: IDG. Retrieved 2014-06-17.
 7

10 Text and image sources, contributors, and licenses

10.1 Text
Apache Cassandra Source: http://en.wikipedia.org/wiki/Apache%20Cassandra?oldid=651804256 Contributors: Enchanter, Frecklefoot,
Edward, Ronz, Stefan, Ehn, Hashar, Cleduc, Bearcat, Alan Liefting, Msiebuhr, Neilc, Pgan002, Euphoria, Rich Farmbrough, Ben-
der235, Anthony Appleyard, PatrickFisher, YPavan, Runtime, Swaroopch, Mindmatrix, Timendum, Deansfa, Jmhodges, Jweiss11, Ve-
gaswikian, Intgr, Tas50, FrankTobia, SamJohnston, Mipadi, Grafen, Wainstead, JLaTondre, Tommymorgan, Chris Chittleborough, Dan-
craggs, SmackBot, Timoey, Aardvark92, Jdorner, Thumperward, OrangeDog, Frap, Cybercobra, Mwtoews, Acdx, Daniel.Cardenas, Argle-
bargleIV, Cydebot, Mblumber, Cinderblock63, Fyedernoggersnodden, Anupam, Nemnkim, Gstein, GreyTeardrop, Cander0000, Krotty,
McSly, Woodjr, Mercurywoodrose, Bcantoni, Andy Dingley, Peter.vanroose, Drmies, Alexbot, The-verver, XLinkBot, Kolyma, Max-
imgr, Deineka, Addbot, Mortense, S4saurabh, Jncraton, MrOllie, Mdnahas, Jbryanscott, Luckas-bot, Yobot, Midinastasurazz, Vanger13,
AnomieBOT, Jim1138, Materialscientist, Pmiossec, Citation bot, ArthurBot, Xqbot, Santiagobasulto, Yadavjpr, Cgraysontx, Ciges, Fres-
coBot, FalconL, Ksato9700, Sae1962, Tdmackey, Freenerd, Jadave234, Biktora, Rollins83, DASHBot, Dewritech, GoingBatty, Peaceray,
Driftx, Arthurjulian, Werieth, ZroBot, Al3xpopescu, Kylemurph, Grossenhayn, Dstainer, Billmantisco, ClueBot NG, Nmilford, Ben mor-
phett, Elisiariocouto, Kinglarvae, Helpful Pixie Bot, YogiWanKenobi, Slebresne, BG19bot, Mark Arsten, IluvatarBot, Andrewllavore,
Mmozum, Diglio.simoni, BattyBot, RichardMills65, Perlscriptgurubot, ChrisGualtieri, TheJJJunk, Viocomnetworks, Clydewylam, Hoest-
melankoli, Samarthgahire, Jamesx12345, CJGarner, RobinUS2, Jonathanbellis, Wikiuser298, Jimtarber, Stuartmccaul, Stather, Mguiere,
Dough34, Sebastibe, Khyryll, Dabron, Spiesche, Textractor, Wehanw, FLGMwt, Jericevans, Kevin at aerospike and Anonymous: 155

10.2 Images
File:ASF-logo.svg Source: http://upload.wikimedia.org/wikipedia/commons/c/cd/ASF-logo.svg License: Apache License 2.0 Contribu-
tors: http://www.apache.org/ Original artist: Apache Software Foundation (ASF)
File:Ambox_important.svg Source: http://upload.wikimedia.org/wikipedia/commons/b/b4/Ambox_important.svg License: Public do-
main Contributors: Own work, based o of Image:Ambox scales.svg Original artist: Dsmurat (talk contribs)
File:Commons-logo.svg Source: http://upload.wikimedia.org/wikipedia/en/4/4a/Commons-logo.svg License: ? Contributors: ? Original
artist: ?
File:Edit-clear.svg Source: http://upload.wikimedia.org/wikipedia/en/f/f2/Edit-clear.svg License: Public domain Contributors: The
Tango! Desktop Project. Original artist:
The people from the Tango! project. And according to the meta-data in the le, specically: Andreas Nilsson, and Jakub Steiner (although
minimally).
File:Folder_Hexagonal_Icon.svg Source: http://upload.wikimedia.org/wikipedia/en/4/48/Folder_Hexagonal_Icon.svg License: Cc-by-
sa-3.0 Contributors: ? Original artist: ?
File:Free_Software_Portal_Logo.svg Source: http://upload.wikimedia.org/wikipedia/commons/3/31/Free_and_open-source_
software_logo_%282009%29.svg License: Public domain Contributors: FOSS Logo.svg Original artist: Free Software Portal Logo.svg
(FOSS Logo.svg): ViperSnake151
File:Wiki_letter_w_cropped.svg Source: http://upload.wikimedia.org/wikipedia/commons/1/1c/Wiki_letter_w_cropped.svg License:
CC-BY-SA-3.0 Contributors:
Wiki_letter_w.svg Original artist: Wiki_letter_w.svg: Jarkko Piiroinen

10.3 Content license

Creative Commons Attribution-Share Alike 3.0