A Lazy Bailout Approach for Dual-Criticality Systems on Uniprocessor Platforms

Abstract

A challenge in the design of cyber-physical systems is to integrate the scheduling of tasks of different criticality, while still providing service guarantees for the higher critical tasks in the case of resource-shortages caused by faults. While standard real-time scheduling is agnostic to the criticality of tasks, the scheduling of tasks with different criticalities is called mixed-criticality scheduling. In this paper, we present the Lazy Bailout Protocol (LBP), a mixed-criticality scheduling method where low-criticality jobs overrunning their time budget cannot threaten the timeliness of high-criticality jobs while at the same time the method tries to complete as many low-criticality jobs as possible. The key principle of LBP is instead of immediately abandoning low-criticality jobs when a high-criticality job overruns its optimistic WCET estimate, to put them in a low-priority queue for later execution. To compare mixed-criticality scheduling methods, we introduce a formal quality criterion for mixed-criticality scheduling, which, above all else, compares schedulability of high-criticality jobs and only afterwards the schedulability of low-criticality jobs. Based on this criterion, we prove that LBP behaves better than the original Bailout Protocol (BP). We show that LBP can be further improved by slack time exploitation and by gain time collection at runtime, resulting in LBPSG. We also show that these improvements of LBP perform better than the analogous improvements based on BP.

Analysis-Runtime Co-design for Adaptive Mixed Criticality Scheduling

“…Further, the time spent in the degraded mode can be reduced via online budget accounting resulting in a faster bailout [39], [40] and recovery. Finally, by using a separate background priority queue, low-criticality jobs that would have been dropped in degraded mode can be run in what would otherwise have been idle time, providing a last chance to meet their deadlines [41]. This mechanism is orthogonal to the fixed priority scheduling scheme used and can be applied to both AMC [4] and the Bailout Protocol [39], [40].…”

“…The performance of AMC-RHS and AMC-RAS can be further improved by running LO-criticality jobs that would otherwise be abandoned in degraded mode at background priorities to give them a last chance to complete by their deadlines. This approach was proposed by Iacovelli and Kirner [41] as a way of augmenting the Bailout Protocol [39], [40], but can be applied to any scheme based on fixed priorities that abandons jobs in degraded mode.…”

“…In this section, we present a simulation and hence scenariobased evaluation of the performance of the variants of the AMC scheme introduced in this paper and compare them to prior work in this area, specifically to variants of the Bailout Protocol [39], [40] and the original AMC scheme [4], using an experimental framework with configurations and metrics similar to those that have previously been used to evaluate the Bailout Protocol [39], [40] and Lazy Execution [41].…”

“…AMC-RAS -AMC-RA enhanced by off-line increases in execution time budgets, see Section IV-C. 3. AMC-RASL -AMC-RAS enhanced via lazy execution [41] of jobs that would otherwise be abandoned, see Section IV-D. 4. AMC-RH -The modified AMC scheme, as described in Section IV, with transitions to degraded mode when any HI-criticality task τ j reaches, without completing, its LO-criticality response time R j (LO) since the start of the busy period in which it was released.…”

ODRE Workshop: Using SIL Arithmetic to Design Safe and Secure Systems

“…Depending on the stage of a mission or external events, the system might switch between different modes. For example, in previous work on mixed-criticality scheduling, fault events like the overrun of the optimistic WCET estimate have been used to switch from a normal operation mode into emergency modes where tasks of higher criticality are given preference [3], [4], [5], [12].…”