Constant bandwidth server revisited

Abstract

The Constant Bandwidth Server (CBS) is an algorithm for providing temporal protection and real-time guarantees to real-time sporadic tasks. Recently, an implementation of this algorithm called SCHED_DEADLINE has been included in the Linux kernel. Therefore, the CBS algorithm is now used to serve more generic tasks than do not obey to the classical sporadic task model. One important type of tasks which was not considered by the original CBS algorithm is the so called "self-suspending task model", where a task instance can suspend itself waiting for an external event. Even if the original algorithm is adapted so that the temporal protection property continues to hold, it is difficult for developers to provide guarantees and to select the most appropriate server parameters for such tasks. This paper investigates the problem of using the CBS algorithm for serving self-suspending tasks, by analysing it from a theoretical point of view and showing how to select the server parameters (budget and periods) for self-suspending tasks. Finally, the effectiveness of these proposals is shown through both simulations and real experiments on Linux / SCHED_DEADLINE.

Deadline-Based Scheduling for GPU with Preemption Support

“…CBS was selected for its design simplicity and limited implementation overhead, seamlessly integrating with the deadline-based scheduling support we prototyped at GPU level. Schedulability analysis with respect to EDF with CBS, also in the presence of shared resources can be found in [19] and [20].…”

Efficient Hosting of Robust IoT Applications on Edge Computing Platform

“…The Real-Time group scheduling [3] extension allows to divide and allocate CPU time between real-time and non real-time tasks. (iii) Earliest Deadline First Scheduler (EDF) Uses Constant Bandwidth Server [6] and allows to associate to each task a budget and a period.…”

Constant bandwidth servers with constrained deadlines

“…While the traditional H-CBS algorithm tries to re-use the current budget (and generates a new scheduling deadline when the current budget is not usable), the "revised" H-CBS [1] (again, defined for D = P), instead, tries to re-use the current scheduling deadline as much as possible (at the cost of decreasing the budget) and is based on the following definition of the generate() function:…”

“…□ Improving the available budget. Alternatively, the first branch of generate could be defined to follow a principle similar to the one adopted in the revised CBS [1]: maintain the current scheduling deadline and use the maximum budget q j allowed by Lemma 5.4. Such a budget can be derived as described in Lemma 5.8.…”

“…According to this scheme the resulting chunks have always a constrained deadline, thus implicitly requiring reservation servers with a relative deadline shorter than their period. The need for a theoretically sound framework for deadline-constrained reservations has also recently come out 1 in the context of on-going developments of the SCHED_DEADLINE scheduling class of the Linux kernel [19].…”