Scheduling real-time transactions: a performance evaluation
ACM Transactions on Database Systems (TODS)
Eliminating receive livelock in an interrupt-driven kernel
ACM Transactions on Computer Systems (TOCS)
An admission control scheme for predictable server response time for web accesses
Proceedings of the 10th international conference on World Wide Web
Real-Time Systems
Session-Based Admission Control: A Mechanism for Peak Load Management of Commercial Web Sites
IEEE Transactions on Computers
Ensuring Latency Targets in Multiclass Web Servers
IEEE Transactions on Parallel and Distributed Systems
A QoS-Sensitive Approach for Timeliness and Freshness Guarantees in Real-Time Databases
ECRTS '02 Proceedings of the 14th Euromicro Conference on Real-Time Systems
Qos-aware real-time data management
Qos-aware real-time data management
A method for transparent admission control and request scheduling in e-commerce web sites
Proceedings of the 13th international conference on World Wide Web
Managing Deadline Miss Ratio and Sensor Data Freshness in Real-Time Databases
IEEE Transactions on Knowledge and Data Engineering
How to Determine a Good Multi-Programming Level for External Scheduling
ICDE '06 Proceedings of the 22nd International Conference on Data Engineering
Autonomic computing through analytic performance models
Autonomic computing through analytic performance models
Web server support for tiered services
IEEE Network: The Magazine of Global Internetworking
Improving throughput via slowdowns
Proceedings of the 32nd ACM/IEEE International Conference on Software Engineering - Volume 2
| Hi-index | 0.00 |
Congestion Control is a necessary tool in Transaction Processing Systems (TPS), to avoid excessive degradation of response times. But it should have an autonomic behavior, adapting automatically to request characteristics to deliver the best possible service. Given that every request has either explicit or implicit deadlines (maximum acceptable response times), we analyze strategies that target request deadlines and throughput. These include maximum throughput seeker strategies, strategies using feedback control on miss rate and an additional proposal for preventive control of miss rates and throughput. Another goal of our proposal is for the control to be external and it should not rely on analytic models for control (because the predictions may be erroneous due to physical system issues). We analyze and compare alternative designs for the control including our own proposals and related ones. Our experiments consider both varied request inter-arrival and duration distributions and a real transaction processing benchmark.