Jump to content

Apache Cassandra: Difference between revisions

From Wikipedia, the free encyclopedia
Browse history interactively
← Previous editNext edit →
Content deleted Content added
VisualWikitext
Slebresne (talk | contribs)
48 edits
mNo edit summary
→Data model: better arrows
Line 62: Line 62:


Map with four dimensions:
Map with four dimensions:
# Keyspace -> Column Family
# Keyspace → Column Family
# Column Family -> Column Family Row
# Column Family → Column Family Row
# Column Family Row -> Columns
# Column Family Row → Columns
# Column -> Data value
# Column → Data value


Map with five dimensions:
Map with five dimensions:
# Keyspace -> Super Column Family
# Keyspace → Super Column Family
# Super Column Family -> Super Column Family Row
# Super Column Family → Super Column Family Row
# Super Column Family Row -> Super Columns
# Super Column Family Row → Super Columns
# Super Column -> Columns
# Super Column → Columns
# Column -> Data value
# Column → Data value


The first dimension - Keyspace -> (Super) Column Family - is limited to a (small) set of schema-predefined keys. Cassandra documentation mentions similarities of this dimension with the table definition inside a relational database.
The first dimension - Keyspace → (Super) Column Family - is limited to a (small) set of schema-predefined keys. Cassandra documentation mentions similarities of this dimension with the table definition inside a relational database.


Furthermore, applications can specify the sort order of columns within a Super Column or Simple Column family. The system allows columns to be sorted either by time or by name. Time sorting of columns is exploited by applications like Facebook Inbox Search, where the results are always displayed in time-sorted order. Any column within a column family is accessed using the convention <code>column_family : column</code>, and any column within a column family that is of type super is accessed using the convention <code>column_family : super_column : column</code>.
Furthermore, applications can specify the sort order of columns within a Super Column or Simple Column family. The system allows columns to be sorted either by time or by name. Time sorting of columns is exploited by applications like Facebook Inbox Search, where the results are always displayed in time-sorted order. Any column within a column family is accessed using the convention <code>column_family : column</code>, and any column within a column family that is of type super is accessed using the convention <code>column_family : super_column : column</code>.

Revision as of 14:26, 22 September 2011

Apache Cassandra
Original author(s)Avinash Lakshman, Prashant Malik
Developer(s)Apache Software Foundation
Initial release2008
Stable release
0.8.6 / September 20, 2011 (2011-09-20)
Repository
Written inJava
Operating systemCross-platform
Available inEnglish
Typekey-value store
LicenseApache License 2
Websitehttp://cassandra.apache.org/

Apache Cassandra is an open source distributed database management system. It is an Apache Software Foundation top-level project[1] designed to handle very large amounts of data spread out across many commodity servers while providing a highly available service with no single point of failure. It is a NoSQL solution that was initially developed by Facebook and powered their Inbox Search feature until late 2010.[2][3] Jeff Hammerbacher, who led the Facebook Data team at the time, has described Cassandra as a BigTable data model running on an Amazon Dynamo-like infrastructure.[4]

Cassandra provides a structured key-value store with tunable consistency.[5] Keys map to multiple values, which are grouped into column families. The column families are fixed when a Cassandra database is created, but columns can be added to a family at any time. Furthermore, columns are added only to specified keys, so different keys can have different numbers of columns in any given family.

The values from a column family for each key are stored together. This makes Cassandra a hybrid data management system between a column-oriented DBMS and a row-oriented store.[6] Also, besides using the way of modeling of BigTable, it has properties like eventual consistency, the Gossip protocol, a master-master way of serving the read and write requests that are inspired by Amazon's Dynamo.[7]

History

Apache Cassandra was developed at Facebook to power their Inbox Search feature by Avinash Lakshman (one of the authors of Amazon's Dynamo) and Prashant Malik. It was released as an open source project on Google code in July 2008.[4] In March 2009, it became an Apache Incubator project.[8] On February 17, 2010 it graduated to a top-level project.[1]

Data model

A table in Cassandra is a distributed multidimensional map indexed by a key. The value is an object that is highly structured. The row key in a table is a string with no size restrictions, although typically 16 to 36 bytes long. Every operation under a single row key is atomic per replica no matter how many columns are being read or written into. Columns are grouped together into sets called column families very much similar to what happens in the BigTable system. Cassandra exposes two kinds of column families: Simple and Super. Super column families can be visualized as a column family within a column family. The top dimension in Cassandra is called Keyspace.

Therefore, Cassandra can handle maps with four or five dimensions:

Map with four dimensions:

  1. Keyspace → Column Family
  2. Column Family → Column Family Row
  3. Column Family Row → Columns
  4. Column → Data value

Map with five dimensions:

  1. Keyspace → Super Column Family
  2. Super Column Family → Super Column Family Row
  3. Super Column Family Row → Super Columns
  4. Super Column → Columns
  5. Column → Data value

The first dimension - Keyspace → (Super) Column Family - is limited to a (small) set of schema-predefined keys. Cassandra documentation mentions similarities of this dimension with the table definition inside a relational database.

Furthermore, applications can specify the sort order of columns within a Super Column or Simple Column family. The system allows columns to be sorted either by time or by name. Time sorting of columns is exploited by applications like Facebook Inbox Search, where the results are always displayed in time-sorted order. Any column within a column family is accessed using the convention column_family : column, and any column within a column family that is of type super is accessed using the convention column_family : super_column : column.

Clustering

When the cluster for Apache Cassandra is designed, an important point is to select the right partitioner. Two partitioners exist:[9]

  1. RandomPartitioner (RP): This partitioner randomly distributes the key-value pairs over the network, resulting in a good load balancing. Compared to OPP, more nodes have to be accessed to get a number of keys.
  2. OrderPreservingPartitioner (OPP): This partitioner distributes the key-value pairs in a natural way so that similar keys are not far away. The advantage is that fewer nodes have to be accessed. The drawback is the uneven distribution of the key-value pairs.

Features

Decentralized

Every node in the cluster has the same role. There is no single point of failure. Data is distributed across the cluster (so each node contains different data), but there is no master as every node can service any request.

Elasticity

Read and write throughput both increase linearly as new machines are added, with no downtime or interruption to applications.

Fault-tolerant

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.

Prominent users

  • AppScale uses Cassandra as a back-end for Google App Engine applications[10]
  • Cisco's WebEx uses Cassandra to store user feed and activity in near real time.[11]
  • Clearspring uses Cassandra "[keep] track of how many times a URL is shared and serves over 200M view requests daily."[12]
  • Cloudkick uses Cassandra to store the server metrics of their users.[13]
  • Cloudtalk's Platform contains API's for users to create messaging apps with Cassandra as it's data store.[14]
  • Constant Contact uses Cassandra in their social media marketing application.[15]
  • Digg, a large social news website, announced on Sep 9th, 2009 that it is rolling out its use of Cassandra[16] and confirmed this on March 8, 2010.[17] TechCrunch has since linked Cassandra to Digg v4 reliability criticisms and recent company struggles.[18] Lead engineers at Digg later rebuked these criticisms as red herring and blamed a lack of load testing.[19]
  • Digital Reasoning's Synthesys application with potential scale to over a 400-node Cassandra database was rolled out in late 2010.[20]
  • Facebook used Cassandra to power Inbox Search, with over 200 nodes deployed.[2] This was abandoned in late 2010.[3]
  • IBM has done research in building a scalable email system based on Cassandra.[21]
  • Isidorey is the creator of Cloudsandra: a multi-tenant platform built on Brisk (Hadoop + Cassandra).[22]
  • Martini Media Network moved from MySQL to Cassandra.[23]
  • Mollom uses Cassandra to track reputations from IP data[24]
  • Netflix uses Cassandra as their back-end database for their streaming services[25][26]
  • 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[27]
  • Mahalo.com uses Cassandra to record user activity logs and topics for their Q&A website[28][29]
  • Ooyala Built a scalable, flexible, real-time analytics engine using Cassandra[30]
  • At Openwave, Cassandra acts as a distributed database and serves as a distributed storage mechanism for Openwave’s next generation messaging platform[31]
  • OpenX is running over 130 nodes on Cassandra for their OpenX Enterprise product to store and replicate advertisements and targeting data for ad delivery[32]
  • Plaxo has "reviewed 3 billion contacts in [their] database, compared them with publicly available data sources, and identified approximately 600 million unique people with contact info."[33]
  • PostRank uses Cassandra as their backend database[34]
  • Rackspace is known to use Cassandra internally.[35]
  • Reddit switched to Cassandra from memcacheDB on March 12, 2010[36] and experienced some problems with overload handling in Cassandra in May.[37]
  • RockYou uses Cassandra to record every single click for 50 million Monthly Active Users in real-time for their online games[38]
  • ShopSavvy uses Cassandra as their storage technology for their barcode scanning application[39]
  • SimpleGeo built a scalable geospatial database on top of Apache Cassandra[40]
  • SoundCloud uses Cassandra to store user account information[41]
  • Twitter announced it is planning to use Cassandra because it can be run on large server clusters and is capable of taking in very large amounts of data at a time.[42][43] Twitter continues to use it but not for Tweets themselves.[44]
  • Urban Airship uses Cassandra with the mobile service hosting for over 160 million application installs across 80 million unique devices[45]
  • Walmart Labs (previously Kosmix) uses Cassandra with SSD[46]
  • Yakaz uses Cassandra on a five-node cluster to store millions of images as well as its social data.[47]

See also

References

  1. ^ a b "Cassandra is an Apache top level project". Mail-archive.com. 2010-02-18. Retrieved 2010-03-29.
  2. ^ a b "Niet compatibele browser". Facebook. Retrieved 2010-03-29.
  3. ^ a b Kannan Muthukkaruppan. "The Underlying Technology of Messages".
  4. ^ a b James Hamilton (July 12, 2008). "Facebook Releases Cassandra as Open Source". Retrieved 2009-06-04.
  5. ^ http://www.cs.cornell.edu/projects/ladis2009/papers/lakshman-ladis2009.pdf
  6. ^ Vishal Shinde (2011-01-19). "Apache Cassandra". http://www.linuxquestions.in/: Linux Questions, Help, Howtos and Tutorials. Retrieved 2010-03-09. The values from a column family for each key are stored together, making Cassandra a hybrid between a column-oriented DBMS and a row-oriented store {{cite web}}: External link in |location= (help)
  7. ^ Olivier Mallassi (2010-06-09). "Let's play with Cassandra… (Part 1/3)". http://blog.octo.com/: OCTO Talks. Retrieved 2010-03-22. Hybrid firstly because Cassandra uses a column-oriented way of modeling data (inspired by the BigTable) and permit to use Hadoop Map/Reduce jobs and secondly because it uses patterns inspired by Dynamo like Eventually Consistent, Gossip protocols, a master-master way of serving both read and write requests… {{cite web}}: External link in |location= (help)
  8. ^ "Is this the new hotness now?". Mail-archive.com. 2009-03-02. Retrieved 2010-03-29.
  9. ^ Dominic Williams. "Cassandra: RandomPartitioner vs OrderPreservingPartitioner". http://wordpress.com/: WordPress.com. Retrieved 2011-03-23. When building a Cassandra cluster, the "key" question (sorry, that's weak) is whether to use the RandomPartitioner (RP), or the OrderPreservingPartitioner (OPP). These control how your data is distributed over your nodes. Once you have chosen your partitioner, you cannot change without wiping your data, so think carefully! The problem with OPP: If the distribution of keys used by individual column families is different, their sets of keys will not fall evenly across the ranges assigned to nodes. Thus nodes will end up storing preponderances of keys (and the associated data) corresponding to one column family or another. If as is likely column families store differing quantities of data with their keys, or store data accessed according to differing usage patterns, then some nodes will end up with disproportionately more data than others, or serving more "hot" data than others. {{cite web}}: External link in |location= (help); line feed character in |quote= at position 325 (help)
  10. ^ cite web|url=http://appscale.cs.ucsb.edu/datastores.html#cassandra
  11. ^ "Re: Cassandra users survey". Mail-archive.com. 2009-11-21. Retrieved 2010-03-29.
  12. ^ Matt Abrams (2011-05-05). "Clearsprings Big Data Architecture Part 1".
  13. ^ https://www.cloudkick.com/blog/2010/mar/02/4_months_with_cassandra/
  14. ^ http://cloudtalk.com/the-cloudtalk-platform.html
  15. ^ Klint Finley (2011-02-18). "This Week in Consolidation: HP Buys Vertica, Constant Contact Buys Bantam Live and More". Read Write Enterprise.
  16. ^ Ian Eure. "Looking to the future with Cassandra".
  17. ^ John Quinn. "Saying Yes to NoSQL; Going Steady with Cassandra".
  18. ^ Erick Schonfeld. "As Digg Struggles, VP Of Engineering Is Shown The Door".
  19. ^ "Is Cassandra to Blame for Digg v4's Failures?".
  20. ^ http://www.datastax.com/wp-content/uploads/2011/03/CS-DigitalReasoning.pdf
  21. ^ "Powered by Google Docs" (PDF). Docs.google.com. Retrieved 2010-03-29.
  22. ^ http://www.cloudsandra.com/
  23. ^ Manicka Babu (2011-05-22). "Cassandra Part 1".
  24. ^ Todd Hoff (2011-02-08). "Mollom Architecture - Killing Over 373 Million Spams at 100 Requests per Second". High Scalability.
  25. ^ cite web|url=http://www.slideshare.net/adrianco/migrating-netflix-from-oracle-to-global-cassandra
  26. ^ Yury Izrailevsky (2011-01-28). "NoSQL at Netflix".
  27. ^ Martin Cozzi (2011-08-31). "Cassandra at Formspring".
  28. ^ {{cite web|url=http://www.datastax.com/wp-content/uploads/2011/06/DataStax-CaseStudy-Mahalo.pdf
  29. ^ http://blip.tv/datastax/cassandra-at-mahalo-com-4030941
  30. ^ http://www.datastax.com/wp-content/uploads/2011/04/WP-Ooyala.pdf
  31. ^ http://www.datastax.com/wp-content/uploads/2011/05/DataStax-CaseStudy-Openwave.pdf
  32. ^ http://openx.com/publisher/technology
  33. ^ Preston Smalley (2011-03-20). "An important milestone - and it's only the beginning!".
  34. ^ Ilya Grigorik (2011-03-29). "Webpulp TV: Scaling PostRank with Ilya Grigorik".
  35. ^ "Hadoop and Cassandra (at Rackspace)". Stu Hood. 2010-04-23. Retrieved 2011-09-01.
  36. ^ Posted by david [ketralnis] (2010-03-12). "what's new on reddit: She who entangles men". blog.reddit. Retrieved 2010-03-29.
  37. ^ Posted by the reddit admins at (2010-05-11). "blog.reddit -- what's new on reddit: reddit's May 2010 "State of the Servers" report". blog.reddit. Retrieved 2010-05-16.
  38. ^ Dathan Vance Pattishall (2011-03-23). "Cassandra is my NoSQL Solution but".
  39. ^ Alexander Muse (2011-07-18). "Shopsavvy leverages Hadoop and Cassandra".
  40. ^ Klint Finley. "How SimpleGeo Built a Scalable Geospatial Database with Apache Cassandra". Read Write Cloud.
  41. ^ "Cassandra at SoundCloud" (PDF).
  42. ^ Popescu, Alex. "Cassandra @ Twitter: An Interview with Ryan King". myNoSQL. Retrieved 2010-03-29.
  43. ^ Babcock, Charles. "Twitter Drops MySQL For Cassandra - Cloud databases". InformationWeek. Retrieved 2010-03-29.
  44. ^ "Cassandra at Twitter Today".
  45. ^ Erik Onnen. "From 100s to 100s of Millions".
  46. ^ Karl Mueller. "Cassandra on SSD".
  47. ^ "Yakaz Technologies".

Bibliography

Retrieved from "https://en.wikipedia.org/enwiki/w/index.php?title=Apache_Cassandra&oldid=451853478"