Misconceptions About Real-time Databases

Stankovic, John A.; Son, Sang H.; Hansson, Jörgen

doi:10.1007/0-306-46988-x_2

Cited by 37 publications

(46 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…ΣupdateFrequency denotes the sum of update frequency of all the cold data items. incrementFactor=(1.00-((ΣupdateFrequency-X)/ΣupdateFrequency) (2) All the cold data items' update period is then incremented by incrementFactor. When the QoS metrics is not found within acceptable limits even after changing the update policy of all the cold data items, admission control is then applied.…”

Section: Miss Ratio Adjustmentmentioning

confidence: 99%

“…Real-time data such as, sensor data, stock market prices and location of moving objects are valid for a specified interval. A realtime database system has not only to be fast, but also has to execute transactions under specified timing constraints maintaining the value(s) of data item(s), reflecting the current real world state and performing transactions within their deadlines [2] .…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Adaptive Update Policy for Proactive Management of Deadline Miss Ratio and Data Freshness in Real-Time Database Models

Tanwani¹,

Ramani²

2005

J. of Computer Science

View full text Add to dashboard Cite

Important applications, like e-commerce, online stock trading, traffic control demand real-time data services. Conventional database perform poor at these applications. A database for real-time data services has to support timing constraints and temporal consistency in addition to supporting characteristics of a conventional database system. In other words, it is desirable to execute transactions within its deadline using updated data reflecting the current world status. In order to achieve these objectives, the Quality of Service (QoS) metrics like miss ratio, data freshness, perceived freshness and Quality of Data (QoD) management schemes along with architectures for effective QoS management are introduced in Literature. These architectures accommodate the unpredictable change in user transactions and maintain QoS metrics by changing the update pattern of data items, which are periodically updated to reflect the real world status. In this study, we present architecture with an improved update policy and propose an algorithm to change the update pattern in such a manner that the system is able to maintain QoS metrics even under busty traffic conditions. The simulation results show that the proposed model can keep QoS metrics, like miss ratio and perceived freshness within limits under busty traffic conditions [1] .

show abstract

Section: Miss Ratio Adjustmentmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Adaptive Update Policy for Proactive Management of Deadline Miss Ratio and Data Freshness in Real-Time Database Models

Tanwani¹,

Ramani²

2005

J. of Computer Science

View full text Add to dashboard Cite

show abstract

“…Different from traditional data stored in databases, real-time data have time semantics. A sampled value is valid only for a certain time interval [1]- [3] and its quality degrades with time until refreshed. The concept of temporal validity is first introduced in [1] to define the correctness of real-time data.…”

Section: Introductionmentioning

confidence: 99%

Online Scheduling Switch for Maintaining Data Freshness in Flexible Real-Time Systems

Han

Chen

Xiong³

et al. 2009

2009 30th IEEE Real-Time Systems Symposium

View full text Add to dashboard Cite

Abstract-Maintaining the temporal validity of real-time data is one of the crucial issues in a real-time database system. Past studies focus on designing algorithms to minimize imposed workload by a fixed set of update transactions while maintaining data freshness within validity intervals. In this paper we revisit this problem by investigating the cost of data freshness maintenance and online scheduling overhead in the presence of mode changes in real-time systems. We propose to apply periodic scheduling policies when the imposed update workload is low to maintain high data freshness. When the update workload becomes high, we propose to switch to more sophisticated algorithms to improve schedulability. In the latter case, not only each scheduling policy must be able to schedule the task set in the corresponding mode, temporal validity must also be maintained during the mode changes. To address this problem, two algorithms, named search-based switch (SBS) and adjustmentbased switch (ABS) are proposed to search for the proper switch point online. SBS checks the temporal validity at the beginning time slot of each idle period while ABS further relaxes this restriction through schedule adjustment. Our experimental results demonstrate the correctness and efficiency of these two algorithms. Our results also show that scheduling switch according to the runtime processor workload can significantly outperform a single fixed scheduling policy in terms of data freshness while incurring only limited online switch overhead.

show abstract

“…These devices communicate the system with its external environment. Many of these embedded systems perform control activities demanding strict timing restrictions and hence predictability [11]. As a result, the interrupt mechanism needs to meet this predictability.…”

Section: Introductionmentioning

confidence: 99%