Abstract
SNAPSHOT isolation provides a protocol for dealing with concurrent transactions in transaction processing applications. Because applications using SNAPSHOT isolation can have very high throughput, many applications are designed using this protocol.
However, SNAPSHOT isolation can produce non-serializable and incorrect schedules. One interesting question is: why are there not more complaints from users that they are getting incorrect results when they use SNAPSHOT isolation?
The most likely answer to that question is that application designers use some design pattern that produces correct schedules at SNAPSHOT isolation. The designers are not necessarily selecting a design pattern because it will guarantee correctness at SNAPSHOT isolation. They are just using design patterns that seem natural to them, and those patterns happen to produce correct executions at SNAPSHOT isolation.
This paper reviews some previous work on sufficient conditions for correctness at SNAPSHOT isolation and then presents an example of a design pattern that is guaranteed to produce correct schedules at SNAPSHOT isolation.
Keywords
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
H. Berenson, P. Bernstein, J. Gray, J. Melton, E. O’Neil, and P. O’Neil. A critique of ANSI SQL isolation levels. In Proc. ACM SIGMOD International Conference on Management of Data, pages 1–10, San Jose, CA, 1995.
A. Bernstein, P. Lewis, and S. Lu. Semantic conditions for correctness at different isolation levels. In Proc. IEEE Intl. Conference on Data Engineering, pages 57–66, 2000.
A. Fekete, E. O’Neil, and P. O’Neil. A read-only transaction anomaly under snapshot isolation. SIGMOD Record, 33(3):12–14, 2004.
A. Fekete, D. Lierokapis, E. O’Neil, and P. O’Neil. Making snapshot isolation serializable. ACM Trans. Database Syst., 30(2):492–528, 2005.
K. Jacobs. Concurrency control: Transaction isolation in SQL92 and Oracle 7. Oracle White Paper Part No. A33745, 1995. (Additional contributors: R. Bamford, G. Doherty, K. Haas, M. Holt, F. Putzolu and B. Quigley).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer Science + Business Media B.V.
About this chapter
Cite this chapter
Lewis, P. (2009). SNAPSHOT Isolation: Why Do Some People Call it SERIALIZABLE?. In: Ravi, S.S., Shukla, S.K. (eds) Fundamental Problems in Computing. Springer, Dordrecht. https://doi.org/10.1007/978-1-4020-9688-4_13
Download citation
DOI: https://doi.org/10.1007/978-1-4020-9688-4_13
Publisher Name: Springer, Dordrecht
Print ISBN: 978-1-4020-9687-7
Online ISBN: 978-1-4020-9688-4
eBook Packages: Computer ScienceComputer Science (R0)