Linux中检查点(Checkpoint)的核心支持——ckpt文件系统的设计

检查点(Checkpoint)是一种软件容错机制,它的目的是提高系统可靠性、减少运算损失,同时检查点机制也是并行系统中进程迁移和负载平衡的基础。在一些检查点系统中,由于对进程的状态检查/状态恢复只具有用户级支持,所以有许多局限性,比如不能完成进程外部状态检查。而在作者的设计与实现中由于具有了核心级的支持,所以能够充分地克服这些局限性。

Supporting Checkpoint Inside the Linux Kernel ——Design of CKPT File System

Checkpoint is a technique used for many purpose,including fault-tolerant applications,rollback transactions and process migration.Many implementation of Checkpoint-system do checkpointing /restarting process just at user level,this bring them some limitations such as can't checkpoint a process outside it.With kernel surpport,this lack can be re-solved in our design and implementation.