Improving ARINC 653 System Reliability by Using Fault-Tolerant Partition Scheduling

摘要

The ARINC 653 specifies multiple operating system components to provide isolation between partitions. This means failure on one partition does not affect any other partition. While each partition cannot affect the other partitions, the failure still happens and possibly leads to failure to the whole system. In this paper, we extend the partition scheduler on ARINC 653 compliant hypervisor called ARLX to use primary-backup scheme. The resulting scheduler will be able to provide the needed service on failed partitions through their backup partitions.