Automatically classifying database workloads
Proceedings of the eleventh international conference on Information and knowledge management
The dawning of the autonomic computing era
IBM Systems Journal
Recommending Materialized Views and Indexes with IBM DB2 Design Advisor
ICAC '04 Proceedings of the First International Conference on Autonomic Computing
Workload Models for Autonomic Database Management Systems
ICAS '06 Proceedings of the International Conference on Autonomic and Autonomous Systems
Adaptive self-tuning memory in DB2
VLDB '06 Proceedings of the 32nd international conference on Very large data bases
Automatic SQL tuning in oracle 10g
VLDB '04 Proceedings of the Thirtieth international conference on Very large data bases - Volume 30
Towards workload shift detection and prediction for autonomic databases
Proceedings of the ACM first Ph.D. workshop in CIKM
Markov Models for Pattern Recognition: From Theory to Applications
Markov Models for Pattern Recognition: From Theory to Applications
On-Line Index Selection for Shifting Workloads
ICDEW '07 Proceedings of the 2007 IEEE 23rd International Conference on Data Engineering Workshop
Finding and analyzing database user sessions
DASFAA'05 Proceedings of the 10th international conference on Database Systems for Advanced Applications
System models for goal-driven self-management in autonomic databases
Data & Knowledge Engineering
On predictive modeling for optimizing transaction execution in parallel OLTP systems
Proceedings of the VLDB Endowment
Hi-index | 0.00 |
Autonomic databases are intended to reduce the total cost of ownership for a database system by providing self-management functionality. The self-management decisions heavily depend on the database workload, as the workload influences both the physical design and the DBMS configuration. In particular, a database reconfiguration is required whenever there is a significant change, i.e. shift, in the workload.In this paper we present an approach for continuous, light-weight workload monitoring in autonomic databases. Our concept is based on a workload model, which describes the typical workload of a particular DBS using n-Gram-Models. We show how this model can be used to detect significant workload changes. Additionally, a processing model for the instrumentation of the workload is proposed. We evaluate our approach using several workload shift scenarios.