The Cross-node Task Migration System Based On SpaceWire Protocol

The continuous development of space science and technology makes the number and type of equipment in the spacecraft equipment system increase dramatically and the requirements of the adaptability,data forwarding efficiency and fault tolerance of the spacecraft equipment system network are also increasing gradually.The Spacewire bus protocol is a high-speed,efficient,and resilient spacecraft bus protocol that is currently of concern.China is currently successfully applied to Fengyun 4 meteorological satellite,which can be seen the initial stage in China's application,the use of spacewire protocal and the development in vxworks to achieve cross-node task migration system is of high value Meaning,whose essence is to improve the fault tolerance of space equipment system,and thus improve the entire system of spaceflight equipment real-time,making the spacecraft more able to adapt to the outer space of the dark environment,more flexible response to some unknown errors in space.Based on the understanding of Spacewire communication protocol,this paper analyzes the underlying hardware structure in detail,especially the design of the glue logic of the connection with SpaceWire equipment.On this basis,the design and implementation of the Space Wore device driver,then refer Vx Works character device driver design framework,design and package the underlying driver,so that the operating system can read the standard interface equipment,and then become the cross-node communication link to the upper application module,the application software is based on the cross-node underlying communication interface which has been designed and developed,a dynamic migration system with a recoverable task is designed.The software layer design of the whole task migration system is consist of the master module,the resource monitoring module,the checkpoint,a setup and rewind recovery module,a bottom task communication module,and a task migration module.The master control module is responsible for managing coordinating the work between the various modules,triggering and mobilizing the work of each module.The resource detection module is responsible for regularly checking the system resources on each node and collecting the resource usage.Checkpoint settings and rewind recovery module are responsible for regularly backing up system resource usage.The underlying task communication module is responsible for the task communication of different nodes.The task migration module is responsible for migrating the tasks on the faulty node according to the migration mapping table when the migration request is made.After the completion of the above-mentioned module design,do many tests for its function and performance with repeated debugging,and ultimately meet the expected operational results and successfully integrated into the entire project,the basic work meets the existing project needs.