Recovery and transaction management in write optimized database systems
Abstract: "In this paper, we present an efficient kernel system for update-intensive database systems. The database is represented as a single access structure -- the database super tree -- which is also used for block management on disk. The recovery algorithm combines the fast commit processi...
Gespeichert in:
1. Verfasser: | |
---|---|
Format: | UnknownFormat |
Sprache: | eng |
Veröffentlicht: |
München
1993
|
Schriftenreihe: | Technische Universität <München>: TUM-I
9340 |
Schlagworte: | |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | Abstract: "In this paper, we present an efficient kernel system for update-intensive database systems. The database is represented as a single access structure -- the database super tree -- which is also used for block management on disk. The recovery algorithm combines the fast commit processing of REDO logging with the fast crash recovery of UNDO logging, however without using any log at all. The recovery component only handles a plain transaction model. Concurrency of transactions is realized by mapping deferred update transactions by to [sic] this plain model." |
---|---|
Beschreibung: | 18 S. graph. Darst. |