DRO+: a systemic and economical approach to improve availability of massive database systems

Authors:
Kai Du;Huaimin Wang;Bo Deng;Yongheng Wang;Shuqiang Yang
Affiliations:
National Laboratory for Parallel and Distributed Processing, Changsha, China;National Laboratory for Parallel and Distributed Processing, Changsha, China;National Laboratory for Parallel and Distributed Processing, Changsha, China;National Laboratory for Parallel and Distributed Processing, Changsha, China;National Laboratory for Parallel and Distributed Processing, Changsha, China
Venue:
WISE'06 Proceedings of the 7th international conference on Web Information Systems
Year:
2006

Citing 11
Cited 0

The dangers of replication and a solution

SIGMOD '96 Proceedings of the 1996 ACM SIGMOD international conference on Management of data
Update propagation protocols for replicated databates

SIGMOD '99 Proceedings of the 1999 ACM SIGMOD international conference on Management of data
Designing and mining multi-terabyte astronomy archives: the Sloan Digital Sky Survey

SIGMOD '00 Proceedings of the 2000 ACM SIGMOD international conference on Management of data
Database Replication Techniques: A Three Parameter Classification

SRDS '00 Proceedings of the 19th IEEE Symposium on Reliable Distributed Systems
Recovery Oriented Computing (ROC): Motivation, Definition, Techniques,

Recovery Oriented Computing (ROC): Motivation, Definition, Techniques,
The Google file system

SOSP '03 Proceedings of the nineteenth ACM symposium on Operating systems principles
Testing the Dependability and Performance of Group Communication Based Database Replication Protocols

DSN '05 Proceedings of the 2005 International Conference on Dependable Systems and Networks
Optimized Data Loading for a Multi-Terabyte Sky Survey Repository

SC '05 Proceedings of the 2005 ACM/IEEE conference on Supercomputing
Integrating compression and execution in column-oriented database systems

Proceedings of the 2006 ACM SIGMOD international conference on Management of data
Using fault injection and modeling to evaluate the performability of cluster-based services

USITS'03 Proceedings of the 4th conference on USENIX Symposium on Internet Technologies and Systems - Volume 4
Self-tuning database technology and information services: from wishful thinking to viable engineering

VLDB '02 Proceedings of the 28th international conference on Very Large Data Bases

Quantified Score

Hi-index	0.00

Visualization

Abstract

Because of the huge scale and numerous components, a massive database system's availability has become a serious challenge. Many database replication technologies are used to increase the MTTF, but few are provided to decrease MTTR in massive database systems where the traditional backup methods are not feasible for expensive human cost. Based on analyzing the characteristics of the data in massive databases, we propose a novel approach called Detaching Read-Only (DRO) mechanism and its variation DRO+. It decreases MTTR through reducing the size of physically changing data in every database by detaching data on node granularity. The analysis and experiment results show that our approach can not only reduce MTTR by an order of magnitude, but also reduce the expensive human cost without extra hardware cost.