Schema evolution

From HandWiki
Revision as of 13:44, 10 July 2021 by imported>John Stpola (simplify)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

In computer science, schema versioning and schema evolution, deal with the need to retain current data and software system functionality in the face of changing database structure.[1] The problem is not limited to the modification of the schema. It, in fact, affects the data stored under the given schema and the queries (and thus the applications) posed on that schema. A database design is sometimes created as a "as of now" instance and thus schema evolution is not considered. (This is different but related to where a database is designed as a "one size fits all" which doesn't cover attribute volatility). This assumption, almost unrealistic in the context of traditional information systems, becomes unacceptable in the context of systems that retain large volumes of historical information or those such as Web Information Systems, that due to the distributed and cooperative nature of their development, are subject of an even stronger pressure toward change (from 39% to over 500% more intense than in traditional settings).[2] Due to this historical heritage the process of schema evolution is nowadays a particularly taxing one. It is, in fact, widely acknowledged that the data management core of an applications is one of the most difficult and critical components to evolve. The key problem is the impact of the schema evolution on queries and applications. As shown in [3] (which provides an analysis of the MediaWiki evolution) each evolution step might affect up to 70% of the queries operating on the schema, that must be manually reworked consequently.

The problem has been recognized as a pressing one by the database community for more than 12 years.[4][5] Supporting Schema Evolution is a difficult problem involving complex mapping among schema versions and the tool support has been so far very limited. The recent theoretical advances on mapping composition[6] and mapping invertibility,[7] which represent the core problems underlying the schema evolution remains almost inaccessible to the large public. The issue is particular felt by temporal databases.[8]

Related works

References

  1. Roddick, John F (1995). "A survey of schema versioning issues for database systems". Information and Software Technology 37 (7): 383–393. doi:10.1016/0950-5849(95)91494-K. 
  2. "Schema Evolution Benchmark - Schema Evolution". yellowstone.cs.ucla.edu. http://yellowstone.cs.ucla.edu/schema-evolution/index.php/Schema_Evolution_Benchmark. 
  3. "Schema Evolution in Wikipedia: toward a Web Information System Benchmark,". 2008. http://web.cs.ucla.edu/~zaniolo/papers/ICEIS2008.pdf. 
  4. "An Online Bibliography on Schema Evolution". http://www.sigmod.org/sigmod/record/issues/0612/p30-article-rahm.pdf. 
  5. Topor, Rodney; Salem, Kenneth; Gupta, Amarnath; Goda, Kazuo; Gehrke, Johannes; Palmer, Nathaniel; Sharaf, Mohamed; Labrinidis, Alexandros et al. (2009). "Schema Versioning" (in en). Encyclopedia of Database Systems. Springer, Boston, MA. pp. 2499–2502. doi:10.1007/978-0-387-39940-9_323. ISBN 978-0-387-35544-3. 
  6. Nash, Alan; Bernstein, Philip A.; Melnik, Sergey (2007). "Composition of mappings given by embedded dependencies". ACM Transactions on Database Systems 32: 4–es. doi:10.1145/1206049.1206053. 
  7. "Quasi-inverses of Schema Mappings". http://www.almaden.ibm.com/cs/people/fagin/quasi.pdf. 
  8. Roddick, John F.; Snodgrass, Richard T. (1995). "Schema Versioning" (in en). The TSQL2 Temporal Query Language. The Springer International Series in Engineering and Computer Science. Springer, Boston, MA. pp. 427–449. doi:10.1007/978-1-4615-2289-8_22. ISBN 9781461359661. 
  9. "deltasql, Database Evolution Under Control". Deltasql Development Team. 2013-04-20. http://deltasql.sourceforge.net.