Abstract
This paper proposes a new approach to checkpointing MPI applications that use long-running CUDA kernels. It becomes possible to take snapshots of data residing on the GPUs without waiting for kernels to complete. The proposed technique is implemented in the context of the state of the art high performance fault tolerance library FTI. As a result we get an elegant solution to the problem of developing resilient MPI applications where GPU kernels run longer than the mean time between hardware failures. We describe in detail how we checkpoint/restart collaborative MPI-CUDA applications, and we provide an initial evaluation of the proposed approach using the Livermore Unstructured Lagrangian Explicit Shock Hydrodynamics (LULESH) application as a case study.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Available at https://bitbucket.org/maxbaird/cuda_backup.
- 2.
The FTI extension is available at https://github.com/leobago/fti/tree/cuda-dev-kernel-interrupt.
References
Baird, M., Fensch, C., Scholz, S.-B., Šinkarovs, A.: A lightweight approach to gpu resilience. In: Mencagli, G., et al. (eds.) Euro-Par 2018. LNCS, vol. 11339, pp. 826–838. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-10549-5_64
Bautista-Gomez, L., Tsuboi, S., et al.: FTI: high performance fault tolerance interface for hybrid systems. In: SC 2011, pp. 1–12 (2011). https://doi.org/10.1145/2063384.2063427
Duato, J., Peña, A.J., et al.: rCUDA: reducing the number of GPU-based accelerators in high performance clusters. In: 2010 International Conference on High Performance Computing Simulation, pp. 224–231 (2010). https://doi.org/10.1109/HPCS.2010.5547126
Garg, R., Mohan, A., et al.: CRUM: checkpoint-restart support for CUDA’s unified memory. In: CLUSTER 2018, pp. 302–313 (2018). https://doi.org/10.1109/CLUSTER.2018.00047
Giunta, G., Montella, R., Agrillo, G., Coviello, G.: A GPGPU transparent virtualization component for high performance computing clouds. In: D’Ambra, P., Guarracino, M., Talia, D. (eds.) Euro-Par 2010, Part I. LNCS, vol. 6271, pp. 379–391. Springer, Heidelberg (2010). https://doi.org/10.1007/978-3-642-15277-1_37
Gupta, V., Gavrilovska, A., et al.: GViM: GPU-accelerated virtual machines. In: ACM Workshop on System-level Virtualization for High Performance Computing, pp. 17–24. ACM (2009), https://doi.org/10.1145/1519138.1519141
Hargrove, P.H., Duell, J.C.: Berkeley lab checkpoint/restart (BLCR) for linux clusters. J. Phys. Conf. Ser. 46, 494–499 (2006). https://doi.org/10.1088/1742-6596/46/1/067
Kannan, S., Farooqui, N., et al.: HeteroCheckpoint: efficient checkpointing for accelerator-based systems. In: 2014 44th Annual IEEE/IFIP International Conference on Dependable Systems and Networks, pp. 738–743 (2014). https://doi.org/10.1109/DSN.2014.76
Karlin, I., et al.: LULESH Programming Model and Performance Ports Overview. Technical report. LLNL-TR-608824, December 2012. https://computing.llnl.gov/projects/co-design/lulesh_ports1.pdf
Lagar-Cavilla, H.A., et al.: VMM-independent graphics acceleration. In: Proceedings of the 3rd International Conference on Virtual Execution Environments, pp. 33–43. ACM (2007). https://doi.org/10.1145/1254810.1254816
Nukada, A., Takizawa, H., et al.: NVCR: a transparent checkpoint-restart library for NVIDIA CUDA. In: 2011 IEEE International Symposium on Parallel and Distributed Processing Workshops and PhD Forum, pp. 104–113 (2011). https://doi.org/10.1109/IPDPS.2011.131
NVIDIA Corporation: NVIDIA CUDA Compute Unified Device Architecture Programming Guide version 10.1.105 (2019). https://bit.ly/2EcQ4hN
Oikawa, M., Kawai, A., et al.: DS-CUDA: a middleware to use many GPUs in the cloud environment. In: 2012 SC Companion: High Performance Computing, Networking Storage and Analysis, pp. 1207–1214 (2012). https://doi.org/10.1109/SC.Companion.2012.146
Peña, A.J., Bland, W., et al.: VOCL-FT: introducing techniques for efficient soft error coprocessor recovery. In: SC 2015: Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, pp. 1–12 (2015). https://doi.org/10.1145/2807591.2807640
Pourghassemi, B., Chandramowlishwaran, A.: cudaCR: an In-Kernel application-level checkpoint/restart scheme for CUDA-enabled GPUs. In: CLUSTER 2017, pp. 725–732 (2017). https://doi.org/10.1109/CLUSTER.2017.100
Shi, L., Chen, H., Sun, J., et al.: vCUDA: GPU-accelerated high-performance computing in virtual machines. IEEE Trans. Comput. 61(6), 804–816 (2012). https://doi.org/10.1109/TC.2011.112
Suzuki, T., Akira Nukada, S.M.: Transparent Checkpoint and Restart Technology for CUDA applications (2016). https://bit.ly/2DzHGbO. Accessed 25 April 2019
Takizawa, H., Sato, K., et al.: CheCUDA: a checkpoint/restart tool for CUDA applications. In: 2009 International Conference on Parallel and Distributed Computing, Applications and Technologies, pp. 408–413 (2009). https://doi.org/10.1109/PDCAT.2009.78
Takizawa, H., et al.: CheCL: transparent checkpointing and process migration of OpenCL applications. In: 2011 IEEE International, IPDPS. IEEE (2011). https://doi.org/10.1109/IPDPS.2011.85
Xu, X., et al.: HiAL-Ckpt: a hierarchical application-level checkpointing for CPU-GPU hybrid systems, pp. 1895–1899 (2010). https://doi.org/10.1109/ICCSE.2010.5593819
Acknowledgement
This work was supported in part by grants EP/N028201/1 and EP/L00058X/1 from the Engineering and Physical Sciences Research Council (EPSRC) and partially sponsored by the European Union’s Horizon 2020 Programme under the LEGaTO Project (www.legato-project.eu), grant agreement 780681.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Baird, M., Scholz, SB., Ĺ inkarovs, A., Bautista-Gomez, L. (2020). Checkpointing Kernel Executions of MPI+CUDA Applications. In: Schwardmann, U., et al. Euro-Par 2019: Parallel Processing Workshops. Euro-Par 2019. Lecture Notes in Computer Science(), vol 11997. Springer, Cham. https://doi.org/10.1007/978-3-030-48340-1_53
Download citation
DOI: https://doi.org/10.1007/978-3-030-48340-1_53
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-48339-5
Online ISBN: 978-3-030-48340-1
eBook Packages: Computer ScienceComputer Science (R0)