Statistical Guarantee in QoS for Temporal Consistency of Real-Time Data Objects
Chuan Zhang · Chinese Journal of Computers · 2006
This paper proposes General Statistical More-Less(GSML) method that take tradeoff between quality of service(QoS) of temporal consistency and the number of update transactions that can be scheduled in the system with guaranteed deadlines.GSML-BA and GSML-OPT method that provide the requested QoS in temporal consistency and maximize the degree of guarantee in QoS by using the remaining processor utilization is constructed.Using adding a reclaiming slack scheme(GSML-SR) can improve further the QoS.The reclaimed slacks can be reallocated for processing the jobs whose computation times are higher than the guaranteed value.Extensive performance evaluation experiments have been done to compare the GSML schemes(GSML-BA and GSML-SR) with the deterministic schemes,More-Less and Half-Half.The performance experiments results have shown that the GSML schemes are effective one for trading between approaching power for the set of transactions and the QoS guaranteed.GSML-SR provides a significant improvement in QoS than that from GSML-BA,GSML-OPT and scheduling power.