OpenGL application live migration with GPU acceleration in personal cloud

Live migration of virtual machine (VM) across physical machines with limited downtime provides many new benefits for users having multiple devices within a personal cloud such as resource sharing and composition. Meanwhile, there are many graphics acceleration solutions within virtual machine which could improve 3D performance for Guest OS, such as VMGL (VMM-independent GPU acceleration), Virtual Box, and so on. However, these solutions cannot support 3D application live migration with physical GPU acceleration, because of the lack of OpenGL context in the target machine. In this paper, we present the design and evaluation of a novel solution that enables OpenGL application live migration with host physical GPU acceleration.

[1]  Fabrice Bellard,et al.  QEMU, a Fast and Portable Dynamic Translator , 2005, USENIX ATC, FREENIX Track.

[2]  Eyal de Lara,et al.  VMM-independent graphics acceleration , 2007, VEE '07.

[3]  Andrew Warfield,et al.  Live migration of virtual machines , 2005, NSDI.

[4]  Roy Want Carry Small, Live Large , 2007, IEEE Pervasive Comput..

[5]  Wei Wang,et al.  Composable IO: a novel resource sharing platform in personal Clouds , 2011, The Journal of Supercomputing.

[6]  Gil Neiger,et al.  Intel virtualization technology , 2005, Computer.