CheCUDA: A Checkpoint/Restart Tool for CUDA Applications

Abstract

In this paper, a tool named CheCUDA is designed to checkpoint CUDA applications that use GPUs as accelerators. As existing checkpoint/restart implementations do not support checkpointing the GPU status, CheCUDA hooks a part of basic CUDA driver API calls in order to record the status changes on the main memory. At checkpointing, CheCUDA stores the status changes in a file after copying all necessary data in the video memory to the main memory and then disabling the CUDA runtime. At restarting, CheCUDA reads the file, re-initializes the CUDA runtime, and recovers the resources on GPUs so as to restart from the stored status. This paper demonstrates that a prototype implementation of CheCUDA can correctly checkpoint and restart a CUDA application written with basic APIs. This also indicates that CheCUDA can migrate a process from one PC to another even if the process uses a GPU. Accordingly, CheCUDA is useful not only to enhance the dependability of CUDA applications but also to enable dynamic task scheduling of CUDA applications required especially on heterogeneous GPU cluster systems. This paper also shows the timing overhead for checkpointing.

Supporting automatic recovery in offloaded distributed programming models through MPI-3 techniques

“…CheCUDA [12] and CheCL [13] are C/R solutions for CUDA and OpenCL applications, respectively. Snapify provides C/R, migration, and swapping services for Intel ® Xeon…”

A Fast and Iterative Migration for GPU Applications

“…The work is done by adding memory copying after the syncthreads() function. CheCUDA [5] hooks a part of CUDA driver API calls to record the modifications in main memory. VCCP and CheCUDA are not designed for general GPU applications.…”

Transparent Accelerator Migration in a Virtualized GPU Environment

“…Takizawa et al [13], [14] demonstrated the feasibility of migrating a GPU program from one node to another. This work is similar to ours.…”