Abstract
Task migration is a well-established method for improving the reliability and achieving the load balancing of systems. Nowadays, multiple processes may be collaborative and they are usually dependent, increasing the difficulty of migrating them. Besides, a process may be required to be migrated to a heterogeneous computing platform according to the current usage and reliability of computer resources. This makes the process migration across heterogeneous platforms essential. This paper focuses on migrating the dependent processes that communicate through shared memory and migrating independent processes across heterogeneous platforms. Linux platform and Windows platform are considered. For the dependent processes, we migrate them mainly by tackling the problem of preserving and restoring their communication states. For the independent process migration across heterogeneous platforms, a user-level opaque checkpointing tool called CKPT is designed and implemented. CKPT strips a process from operating systems by abstracting the calculation state of the process into pure data. Its operations are simple, making it lightweight in migrating the process. Experimental results show that our proposed approach achieves the dependent process migration and heterogeneous process migration. The time overhead and space overhead of preserving the communication state on Linux platform are 8.5% and 0.55% respectively. The results on Windows platform are 7.5% and 25%. The time overhead and space overhead of CKPT in migrating heterogeneous processes are 43% and 20%. These results show the effectiveness of our approach.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.