Locating and accessing large datasets using Flower Index Approach. (27th February 2019)

• Record Type: Journal Article
• Authors: Kvet, Michal; Krsak, Emil; Matiasko, Karol
• Other Names: Hodon Michal guestEditor.; Furtak Janusz guestEditor.; Fahrnberger Güenter guestEditor.; Awad Ali I. guestEditor.; Cui Xiaohui guestEditor.; Suman Bilial guestEditor.; Wang Zhibo guestEditor.
• Is Part Of: Concurrency and computation. Volume 32:Number 13(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2019-02-27
• Subjects: attribute granularity temporal architecture -- Flower Index Approach -- full table scan -- index data pointer -- query processing -- volatility; Parallel processing (Electronic computers) -- Periodicals; Parallel computers -- Periodicals; 004.35
• DOI: 10.1002/cpe.5209
• Languages: English
• ISSNs: 1532-0626

Abstract

Summary: Information system core part is just the data stored in the database. Over the decades, the number and structure of the data have been changed. Nowadays, data must reflect not only current valid data but also historical and future images as well. Each data tuple is therefore delimited by the validity timeframe forming a temporal paradigm. Several temporal models have been developed with an emphasis on the data structure, the frequency of changes, and synchronization processes. Although the system stores time delimited data during the object lifecycle, it is not efficient, even useful to store data in the main system indefinitely. Reliability is another significant aspect of the processing covered by the purging processes. Query processing is based on the accessing data in the memory buffer cache of the database instance preceded by the loading process from the physical database. This paper proposes a Flower Index Approach as the main contribution. It removes the impact of the High Water Mark, removes useless block loading with no relevant data, and provides effective data access stream using a specific index. Full Table Scan is then not used and data are accessed directly using index ROWID locators.