Semi-partitioned Mixed-Criticality Scheduling

Scheduling isolation in mixed-criticality systems is challenging without sacrificing performance. In response, we propose a scheduling approach that combines server-based semi-partitioning and deadline scaling. Semipartitioning (whereby only some tasks migrate, in a carefully managed manner), hitherto used in single criticality systems, offers good performance with low overheads. Deadline-scaling selectivelyprioritizes high-criticality tasks in parts of the schedule to ensure theirdeadlines are met even in rares case of execution time overrun. Ournew algorithm NPS-F-MC brings semi-partitioning to mixed-criticalityscheduling and uses Ekberg and Yi 19s state-ofthe-art deadline scaling approach. It ensures scheduling isolation among different-criticality tasksand only allows low-criticality task migration. We also explore variantsthat disallow migration entirely or relax the isolation between differentcriticalities (SP-EKB) in order to evaluate the performance tradeoffs associated with more flexible or rigid safety and isolation requirements. Semi-partitioned mixed-criticality scheduling Muhammad Ali Awan, Konstantinos Bletsas, Pedro F. Souto and Eduardo Tovar CISTER/INESC-TEC Research Centre, Porto, Portugal ISEP/IPP, Porto Faculty of Engineering, University of Porto Abstract. Scheduling isolation in mixed-criticality systems is challenging without sacrificing performance. In response, we propose a scheduling approach that combines server-based semi-partitioning and deadlinescaling. Semi-partitioning (whereby only some tasks migrate, in a carefully managed manner), hitherto used in single criticality systems, offers good performance with low overheads. Deadline-scaling selectively prioritise high-criticality tasks in parts of the schedule to ensure their deadlines are met even in rares case of execution time overrun. Our new algorithm NPS-F-MC brings semi-partitioning to mixed-criticality scheduling and uses Ekberg and Yi’s state-of-the-art deadline scaling approach. It ensures scheduling isolation among different-criticality tasks and only allows low-criticality task migration. We also explore variants that disallow migration entirely or relax the isolation between different criticalities (SP-EKB) in order to evaluate the performance tradeoffs associated with more flexible or rigid safety and isolation requirements. Scheduling isolation in mixed-criticality systems is challenging without sacrificing performance. In response, we propose a scheduling approach that combines server-based semi-partitioning and deadlinescaling. Semi-partitioning (whereby only some tasks migrate, in a carefully managed manner), hitherto used in single criticality systems, offers good performance with low overheads. Deadline-scaling selectively prioritise high-criticality tasks in parts of the schedule to ensure their deadlines are met even in rares case of execution time overrun. Our new algorithm NPS-F-MC brings semi-partitioning to mixed-criticality scheduling and uses Ekberg and Yi’s state-of-the-art deadline scaling approach. It ensures scheduling isolation among different-criticality tasks and only allows low-criticality task migration. We also explore variants that disallow migration entirely or relax the isolation between different criticalities (SP-EKB) in order to evaluate the performance tradeoffs associated with more flexible or rigid safety and isolation requirements.