Differences
This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision | ||
|
projects:fc-orb [2009/02/27 16:06] gsmith28 |
projects:fc-orb [2019/12/18 23:54] (current) |
||
|---|---|---|---|
| Line 1: | Line 1: | ||
| + | ~~NOCACHE~~ | ||
| + | ~~NOTOC~~ | ||
| + | |||
| ====== FC-ORB ====== | ====== FC-ORB ====== | ||
| ===== A Feedback Controlled-ORB Middleware to Enhance the Robustness of Distributed Real-Time Middleware via End-to-End Utilization Control ===== | ===== A Feedback Controlled-ORB Middleware to Enhance the Robustness of Distributed Real-Time Middleware via End-to-End Utilization Control ===== | ||
| - | A Feedback Controlled-ORB Middleware to Enhance the Robustness of Distributed Real-Time Middleware via End-to-End Utilization Control | ||
| {{:fc-orb.jpg |}} | {{:fc-orb.jpg |}} | ||
| A key challenge for distributed real-time and embedded (DRE) middleware is maintaining both system reliability and desired real-time performance in unpredictable environments where system workload and resources may fluctuate significantly. This paper presents FC-ORB, a real-time Object Request Broker (ORB) middleware that employs end-to-end utilization control to handle fluctuations in application workload and system resources. The contributions of this paper are three-fold. First, we present a novel utilization control service that enforces desired CPU utilization bounds on multiple processors by adapting the rates of end-to-end tasks within user-specified ranges. Second, we describe a set of middleware-level mechanisms designed to support end-to-end tasks and distributed multi-processor utilization control in a real-time ORB. Finally, we present extensive experimental results on a Linux testbed. Our results demonstrate that our middleware can maintain desired utilizations in face of uncertainties and variations in task execution times, resource contentions from external workloads, and permanent processor failure. FC-ORB demonstrates that the integration of utilization control, end-to-end scheduling and fault-tolerance mechanisms in DRE middleware is a promising approach for enhancing the robustness of DRE applications in unpredictable environments. | A key challenge for distributed real-time and embedded (DRE) middleware is maintaining both system reliability and desired real-time performance in unpredictable environments where system workload and resources may fluctuate significantly. This paper presents FC-ORB, a real-time Object Request Broker (ORB) middleware that employs end-to-end utilization control to handle fluctuations in application workload and system resources. The contributions of this paper are three-fold. First, we present a novel utilization control service that enforces desired CPU utilization bounds on multiple processors by adapting the rates of end-to-end tasks within user-specified ranges. Second, we describe a set of middleware-level mechanisms designed to support end-to-end tasks and distributed multi-processor utilization control in a real-time ORB. Finally, we present extensive experimental results on a Linux testbed. Our results demonstrate that our middleware can maintain desired utilizations in face of uncertainties and variations in task execution times, resource contentions from external workloads, and permanent processor failure. FC-ORB demonstrates that the integration of utilization control, end-to-end scheduling and fault-tolerance mechanisms in DRE middleware is a promising approach for enhancing the robustness of DRE applications in unpredictable environments. | ||
