DMTCP: Checkpoint-Restart on the Desktop

摘要：本工作建立在现有的开源、用户空间的DMTCP包的基础上，用于透明、分布式的检查点。将实现三个目标：(i)检查点-重新启动桌面上长时间运行的计算；（ii）交互软件包的保存-恢复；（iii）通用可逆调试器。前两个目标将允许软件开发团队在他们的包中添加可靠的“保存工作区”特性——不需要内核模块或其他特权操作。第三个目标是增强任何调试器的可逆性（例如后退命令），并使用反向表达式watchpoint命令从软件错误向后移动到原始软件故障。智力优势：虽然检查点已经存在了20多年，但早期的软件包很难维护。dmtcp的非特权用户空间设计已经有5年的历史了。它非常适合集成到其他软件中，在这种情况下，任何终端用户对安装内核模块或其他管理权限的要求都与广泛分发不兼容。最后，DMTCP是第一个能够直接检查点gdb会话（gdb进程及其目标进程）的包——这是设想的新型可逆调试器的一个关键特性。更广泛的影响：检查点和过程迁移长期以来一直是科学和工程的兴趣所在，但是经常受到软件脆弱性或特殊需求的影响。DMTCP方法消除了这些障碍。此外，“时间旅行（可逆）调试器”的广泛使用将极大地加速软件开发，因为更容易发现错误。NIST的一份报告估计，软件漏洞每年给经济造成的损失为595亿美元。最后，检查点重新启动桌面的兴奋因素有助于吸引和激励学生学习这一关键技术中有时晦涩难懂的系统问题。

Title: DMTCP: Checkpoint-Restart on the DesktopPI: Gene CoopermanNSF Proposal Number: 0960978ABSTRACT: This work builds upon the existing open source, user-space DMTCP packagefor transparent, distributed checkpointing. Three goals will be accomplished:(i) checkpoint-restart of long-running computations on the desktop;(ii) save-restore of interactive software packages; and (iii) a universalreversible debugger. The first two goals will allow software developmentteams to add to their package a reliable "save workspace" feature --- withno requirement for a kernel module or other privileged operations.The third goal is to enhance any debugger with reversibility (e.g. aback-step command), and with a reverse expression watchpoint command tomove backwards from a software error to the original software fault.INTELLECTUAL MERIT:While checkpointing has existed for over 20 years, earlier packageswere difficult to maintain. The unprivileged, user-space design ofDMTCP has a five-year track record. It is ideal for integration intoother software, where any end-user requirement for installation of akernel module or other administrative privilege is incompatible withwidespread distribution. Finally, DMTCP is the first package able todirectly checkpoint a gdb session (the gdb process and its target process)-- a key feature for the envisioned new type of reversible debugger.BROADER IMPACT:Checkpointing and process migration have long been of interest for scienceand engineering, but too often suffered from software fragility or specialrequirements. The DMTCP approach removes these obstacles. Further,the wider use of ``time-traveling (reversible) debuggers'' will greatlyaccelerate software development due to the greater ease of finding bugs.A NIST report estimates the cost of software bugs to the economy at$59.5 billion per year. Finally, the excitement factor of checkpoint-restarton the desktop helps attract and motivate students toward the learningof sometimes arcane systems issues in this critical technology.