Downloaded 66 times
Uploaded byIvan Zoratti
The Evolution of Open Source Databases
The document provides an overview of the evolution of open source databases from the past to present and future. It discusses the early days of navigational and hierarchical databases. It then covers the development of relational databases and SQL. It outlines the rise of open source databases like MySQL, PostgreSQL, and SQLite. It also summarizes the emergence of NoSQL databases and NewSQL systems to handle big data and cloud computing. The document predicts continued development and blending of features between SQL, NoSQL, and NewSQL databases.
The Evolution of Open Source Databases
- 1. The Evolution of Open Source Databases Past, Present and Future Ivan Zoratti V1310.01
- 2.
- 3. SkySQL •Leading provider of open source databases, services and solutions •Home for the founders and the original developers of the core of MySQL •The creators of MariaDB, the drop-‐in, innovative replacement of MySQL
- 4.
- 5. Databases and DBMSs •Database “An organized collection of data” “Databases are created to operate large quantities of information by inputting, storing, retrieving, and managing that information” •DBMS -‐ DataBase Management System “A suite of computer software providing the interface between users and a database or databases” Wikipedia -‐ hIp://en.wikipedia.org/wiki/Database PAGE 5
- 6. The early days •Navigational Databases •Objects are found by following references from other objects •Network and Hierarchical Databases •Data is organized into a network or a tree-‐like structure PAGE 6
- 7. The beginning of a relationship... •1970 -‐ Edgar Codd starts working on the relational model •1973 -‐ Ingres project at Berkeley University •1974 -‐ System R (DB/2 precursor) •1977 -‐ Larry Ellison founds Software Development Labs •1979 -‐ Oracle v2 becomes relational and implements SQL •1980 -‐ Informix is established as Relational Database Systems -‐ the first RDBMS is based on ISAM •1984 -‐ Mark Hoffman and Bob Epstein found Sybase to exploit the capabilities of client/server architectures •1985 -‐ Michael Stonebraker starts the POSTGRES project •1992 -‐ Micosoft starts JET (Joint Engine Technology) •1995 -‐ MySQL v1 from MySQL AB •2000 -‐ Richard Hipp designs SQLite PAGE 7
- 8.
- 9. Open source RDBMSs •PostgreSQL License • Similar to the MIT license • Greenplum • Open source versions: • Commercial versions: • • • EnterpriseDB Postgres Plus Netezza Red Hat Database • GPL v2 License • Open source versions: • Public domain • Operadng Systems • Commercial versions: • Web Browsers • • MariaDB Server Percona Server • • • • Oracle MySQL Enterprise InfiniDB Infobright ScaleDB • • • OS X IOS Windows Phone • • Firefox Chrome • Thunderbird • Others PAGE 9
- 10.
- 11. Online vs. Enterprise The Top 10 web sites 1. Google 2. Facebook 3. YouTube 4. Yahoo! 5. Baidu.com 6. Wikipedia 7. QQ.com 8. Linkedin 9. Windows Live 10. TwiIer 5% Oracle IBM Microsok SAP/Sybase Teradata 4% 18% 51% 22% ! Source: Gartner (2011) www.alexa.com PAGE 11
- 12. Understanding the CAP theorem •Consistency / Availability / Partition Tolerance •CA •Synchronous Replication •Two Phase Commit •MySQL, ACID/RDBMSs •CP •MongoDB, HBase, Redis, MemcacheD •AP C A P •Cassandra, Riak, CouchDB PAGE 12
- 13. The advent of NoSQL •First seen in 1998, but really active since 2009 •SQL is complex, boring and clunky •Not everybody needs ACID compliance •Relational databases are: •Slow •Not scalable •Not optimized for rich data •Difficult to change and maintain PAGE 13
- 14. NoSQL vs SQL NoSQL •Schema-‐less (or dynamic schema) •Dynamic horizontal scaling •Good to store and retrieve a great quantity of data ! •Great Flexibility •Full ACID not required -‐ “BASE is better” Basically Available, Soft state, Eventually consistent •Objects: Collections, Documents, Fields •NoSQL DBs: Key/Value BigTable Document Graph SQL •Rigid Schema design •Static or no horizontal scaling •Good to store and retrieve data that has relationship between the elements •Pretty inflexible •ACID as a given •Atomic, Consistent, Isolated, Durable •Objects: Tables, Rows, Columns •SQL DBs: •Row-‐based •Columnar •Object Relational PAGE 14
- 15. NoSQL players 2009 Key/Value Stores 2009 2008 Document Store 2005 2009 Graph DB Column Store 2007 PAGE 15
- 16.
- 17.
- 18. Here is what is going on in the NoSQL world PAGE 18
- 19. Here is what is going on in the NoSQL world ! ! ! •New in CQL with Cassandra 2.0 •Lightweight transacdon for INSERT and UPDATE with opdmisdc locking •Inidal support for triggers •Prepared statements and cursors •User authendcadon PAGE 19
- 20. Here is what is going on in the NoSQL world ! ! ! ! ! ! •Transacdons with MVCC and ACID reliability with TokuMX •Fractal indexing •Replicadon •Compression PAGE 20
- 21. Welcome to the Cloud •Some cloud-‐focused DB players •Herokypostgres -‐ Open source -‐ PostgreSQL •Clustrix -‐ Closed source -‐ MySQL client compatible •GenieDB -‐ Closed source -‐ MySQL client compatible •Amazon DynamoDB -‐ Closed source -‐ From the Dynamo project •Amazon RDS/MySQL -‐ Open Source -‐ Pure MySQL •Cloud lock-‐in / Service lock-‐in / Vendor lock-‐in PAGE 21
- 22. ...and welcome to NewSQL •NoSQL performance for OLTP systems requiring ACID features •Shared nothing •Sharding •In-‐memory storage •Local and geographical replication •NewSQL players •NuoDB •VoltDB •Clustrix •Spanner/Google F1 •Translattice •MariaDB PAGE 22
- 23. ...and welcome to NewSQL ! 10 •Multi-‐source Replication •Handler Socket •Dynamic Columns •Text Search Engines •Sphinx •Sharding Engines •Spider •ScaleDB •Integration Engines •Cassandra •CONNECT •Optimistic Locking / Sync Replication •Galera •RESTFul API PAGE 23
- 24. Risks and old tricks •Vendors who Open Source DB compatibility, but they are closed source •Vendors who start open sources, then they move to a closed source license •Only DBs available from non-‐ commercial organisations seem to be safe PAGE 24
- 25. Databases backed by foundations •Apache Software Foundation •Accumulo -‐ Key/value written in Java, based on Google BigTable (on top of Hadoop) •Cassandra -‐ Key/valye written in Java •Derby -‐ RDBMS written in Java •HBase -‐ Big Data columnar engine written in Java •MariaDB Foundation •MariaDB -‐ MySQL compatible, GPL v2, written in C, with multiple storage engine options •SPI -‐ Software in the Public Interest •Postgresql PAGE 25
- 26. Watch what is coming •Flash Storage and SSD optimized •Interoperability among multiple •NoSQL and SQL integration > •More Big Data features databases NewSQL as a standard •Sharding •Eventual Consistency as an option •Complex queries + in-‐DB logic vs. performance •Data has changed, databases must databases •Columnar storage •Document handling and indexing •Databases as part of the IaaS layer •Or what we used to say in the old days: The database is the operating system handle the new and the old data PAGE 26
- 27. Pictures courtesy of: • www.interflora.co.uk • www.bubblenews.com • www.fullhdwpp.com • www.hortonworks.com • www.cloudera.com Thank You! www.skysql.com ivan@skysql.com izoratti.blogspot.com www.slideshare.net/izoratti