Parallelism and recovery in database systems
ACM Transactions on Database Systems (TODS)
Guest Editor's Introduction: The Development of Data-Base Technology
ACM Computing Surveys (CSUR)
Recovery Techniques for Database Systems
ACM Computing Surveys (CSUR)
SIGMOD '76 Proceedings of the 1976 ACM SIGMOD international conference on Management of data
Restart and recovery in a transaction-oriented information processing system
SIGFIDET '74 Proceedings of the 1974 ACM SIGFIDET (now SIGMOD) workshop on Data description, access and control
A Parity Structure for Large Remotely Located Replicated Data Files
IEEE Transactions on Computers
| Hi-index | 0.00 |
Database recovery techniques in a real-time environment for so called single-division databases are investigated. A classification of database recovery goals and a classification of database system crashes is presented. It is shown that the (best) recovery goal is a function of a crash category against which the system is to be protected. In particular, for the broadest category of hidden hard crashes an actual past state is an attainable recovery goal. It is described how to reach this goal using a generic recovery technique, based on an idea of a database recovery block. The specific recovery techniques implementing the generic technique are described. Then the representation of each specific recovery technique in terms of atomic "primitives" is demonstrated. The claim is made that this "divide-and-conquer" approach can facilitate the analysis of the database recovery techniques.