Weird ways to use application checkpoint/restore

Checkpoint/restore is mostly used for live migration of containers, i.e. moving tasks from one machine to another without need to shut them down. If we let ourselves think outside the box, there are a lot of other possible usage scenarios with checkpoint/restore, some of them are quite intriguing. Task and system cloning. Reversing a process execution. Remote fork() for a task. Slow-to-boot services speed-up. That magical "save" button in a game that lacks one...

The talk gives a brief introduction to CRIU, and describes some of such less known ways to use it.