Database Workload Management - Characterization to Idleness Detection
Database Management Systems experience workloads that consist of user and system workloads which include structured query language statements, transactions, sessions, jobs etc. These workloads need various resources including CPU cycles, memory, storage etc for their successful execution. The dynamic nature of the database workload leads researchers, vendors and other stack holders towards new problems and challenges. The research here introduces the concept of an efficient technique for the management of the database workload. Three modules of the proposed approach are characterization technique based on Case Base Reasoning, impact based scheduling using Fuzzy Logic and idleness detection to improve the performance of the DBMS.