论文信息 - Dynamic Data Distribution (D3) in a Shared-Nothing Multiprocessor Data Store

Dynamic Data Distribution (D3) in a Shared-Nothing Multiprocessor Data Store

Redundant storage of data can make a multiprocessor database system tolerant of single processor failures. The D3 algorithm described here enables such a system to dynamically reorganize after a failure in order to protect itself against further failures, without interruption of service. The algorithm also permits addition or removal of a processor, and data migration for load balancing, without interruption of service.

Donald D. Chamberlin | Frank B. Schmuck | D. Chamberlin | Frank B. Schmuck

[1] Jim Gray,et al. Notes on Data Base Operating Systems , 1978, Advanced Course: Operating Systems.

[2] David K. Gifford,et al. Weighted voting for replicated data , 1979, SOSP '79.

[3] Irving L. Traiger,et al. The Recovery Manager of the System R Database Manager , 1981, CSUR.

[4] Andrea J. Borr. Transaction Monitoring in ENCOMPASS: Reliable Distributed Transaction Processing , 1981, VLDB.

[5] Irving L. Traiger,et al. Transactions and consistency in distributed database systems , 1982, TODS.

[6] Philip A. Bernstein,et al. The failure and recovery problem for replicated databases , 1983, PODC '83.

[7] Flaviu Cristian,et al. An efficient, fault-tolerant protocol for replicated data management , 1985, Fault-Tolerant Distributed Computing.

[8] Flaviu Cristian,et al. An efficient, fault-tolerant protocol for replicated data management , 1985, PODS '85.

[9] Randy H. Katz,et al. A case for redundant arrays of inexpensive disks (RAID) , 1988, SIGMOD '88.

[10] David J. DeWitt,et al. Chained declustering: a new availability strategy for multiprocessor database machines , 1990, [1990] Proceedings. Sixth International Conference on Data Engineering.

[11] Hamid Pirahesh,et al. ARIES: a transaction recovery method supporting fine-granularity locking and partial rollbacks using write-ahead logging , 1998 .