Abstract
High performance GPU clusters are widely used for massive amount of concurrent dataflow processing, and have higher requirements for real-time, reliability and flexibility. However, the higher computational intensiveness and resources utilization lead to excessively high system temperature and power consumption, and even result in instantaneous failures. In this paper, we present a real-time and efficient dynamic taskflow migration approach (DTMA) based on a computing cluster. Firstly, we propose our basic theoretical models. Among them, the cluster communication model elaborates on all the communication paths and calculates the communication overhead of different migration modes. Secondly, on the basis of theoretical models and multiple instances analysis, our taskflow migration rules are summarized, and the rules help to balance cluster resources utilization and improve the overall performance of GPUs. Thirdly, the DTMA adjusts the cluster task allocation by utilizing performance and power consumption aware migration approach. This is done to reduce single node power consumption and enhance system reliability by shifting the current GPU load to other available GPU (GPUs). Moreover, the DTMA uses a circular queue to store resources information of available GPUs for better task scheduling. We evaluate the effect of DTMA through analyzing power consumption, temperature, fan speed and migration cost with different experiments. The experiment results demonstrate that DTMA is able to improve the performance and reliability of our cluster computing system, and reduce instantaneous failures.
Similar content being viewed by others
References
Marjani, M., Nasaruddin, F., Gani, A., Karim, A., Hashem, I.A.T., Siddiqa, A., Yaqoob, I.: Big iot data analytics: architecture, opportunities, and open research challenges. Big IoT Data Anal. Archit. Oppor. Open Res. Chall. 5(99), 5247–5261 (2017)
Mervis, J.: Agencies rally to tackle big data. Science 336(6077), 22 (2012)
Lv, Z., Song, H., Basanta-Val, P., Steed, A., Jo, M.: Next-generation big data analytics: State of the art, challenges, and future research topics. IEEE Trans. Ind. Inf. 13(4), 1891–1899 (2017)
Zhang, Y., Qiu, M., Tsai, C.W., Hassan, M.M., Alamri, A.: Health-CPS: Healthcare cyber-physical system assisted by cloud and big data. IEEE Syst. J. 11(1), 88–95 (2017)
Venkatesh, G., Arunesh K.: Map Reduce for big data processing based on traffic aware partition and aggregation. Clust. Comput. (2018). https://doi.org/10.1007/s10586-018-1799-6
Mmel, R.: Google’s mapreduce programming model revisited. Sci. Comput. Program. 70(1), 1–30 (2008)
Jiang, H., Chen, Y., Qiao, Z., Weng, T.-H., Li, K.-C.: Scaling up mapreduce-based big data processing on multi-gpu systems. Clust. Comput. 18(1), 369–383 (2015)
Ramírez-Gallego, S., Garca, S., Beítez, J.M., Herrera, F.: A distributed evolutionary multivariate discretizer for big data processing on apache spark. Swarm Evol. Comput. 38, 240–250 (2017)
Alsheikh, M.A., Niyato, D., Lin, S., Tan, H.P., Han, Z.: Mobile big data analytics using deep learning and apache spark. IEEE Netw. 30(3), 22–29 (2016)
Huang, W., Song, G., Hong, H., Xie, K.: Deep architecture for traffic flow prediction: Deep belief networks with multitask learning. IEEE Trans. Intell. Transp. Syst. 15(5), 2191–2201 (2014)
Li, P., Chen, Z., Yang, L.T., Zhang, Q., Deen, M.J.: Deep convolutional computation model for feature learning on big data in Internet of Things. IEEE Trans. Ind. Inf. 14(2), 790–798 (2017)
Chen, C.F.R., Lee, G.G.C., Xia, Y., Lin, W.S., Suzumura, T., Lin, C.Y.: Efficient multi-training framework of image deep learning on GPU cluster. In: IEEE International Symposium on Multimedia, pp. 489–494 (2016)
TOP500: Tp500list. https://www.top500.org/lists/2017/11/slides/
Li, K., Tang, X., Li, K.: Energy-efficient stochastic task scheduling on heterogeneous computing systems. IEEE Trans. Parallel Distrib. Syst. 25(11), 2867–2876 (2014)
Kreutzer, M., Thies, J., Pieper, A., Alvermann, A., Galgon, M., Rhrig-Zllner, M., Shahzad, F., Basermann, A., Bishop, A.R., Fehske, H.: Performance Engineering and Energy Efficiency of Building Blocks for Large. Sparse Eigenvalue Computations on Heterogeneous Supercomputers. Springer, Cham (2016)
Liu, W., Du, Z., Xiao, Y., Bader, D.A., Chen, X.: A waterfall model to achieve energy efficient tasks mapping for large scale GPU clusters. In: International Heterogeneity in Computing Workshop. Anchorage, pp. 82–92 (2011)
Hong, S., Kim, H.: An integrated GPU power and performance model. In: International Symposium on Computer Architecture, pp. 280–289 (2010)
Alonso, P., Dolz, M.F., Igual, F.D., Mayo, R., Quintanaor, E.S.: Reducing energy consumption of dense linear algebra operations on hybrid CPU–GPU platforms. In: IEEE International Symposium on Parallel and Distributed Processing with Applications, pp. 56–62 (2012)
Padoin, E.L., Pilla, L.L., Boito, F.Z., Kassick, R.V., Velho, P., Navaux, P.O.: Evaluating application performance and energy consumption on hybrid CPU+GPU architecture. Clust. Comput. 16(3), 511–525 (2013)
Hong, S., Kim, H.: An analytical model for a GPU architecture with memory-level and thread-level parallelism awareness. ACM SIGARCH Comput. Architect. News 37(3), 152–163 (2009)
Ge, R., Feng, X., Song, S., Chang, H.C., Li, D., Cameron, K.W.: Powerpack: energy profiling and analysis of high-performance systems and applications. IEEE Trans. Parallel Distrib. Syst. 21(5), 658–671 (2010)
Defour, D., Petit, E.: GPUburn: a system to test and mitigate GPU hardware failures. In: International Conference on Embedded Computer Systems: Architectures, Modeling, and Simulation, pp. 263–270 (2013)
Rech, P., Aguiar, C., Ferreira, R., Silvestri, M.: Neutron-induced soft errors in graphic processing units. In: IEEE Radiation Effects Data Workshop, pp. 1–6 (2012)
Guilhemsang, J., Hron, O., Ventroux, N., Goncalves, O., Giulieri, A.: Impact of the application activity on intermittent faults in embedded systems. In: VLSI Test Symposium, pp. 191–196 (2011)
Sun, D., Zhang, G., Yang, S., Zheng, W., Khan, S.U., Li, K.: Re-stream: real-time and energy-efficient resource scheduling in big data stream computing environments. Inf. Sci. 319, 92–112 (2015)
Lin, S., Xie, Z.: A Jacobi\(\_\)PCG solver for sparse linear systems on multi-GPU cluster. J. Supercomput. 73(1), 1–22 (2016)
Fang, Y., Chen, Q., Xiong, N.N., Zhao, D., Wang, J.: RGCA: a reliable gpu cluster architecture for large-scale internet of things computing based on effective performance-energy optimization. Sensors 17(8), 1799 (2017)
Cook, S.: CUDA Programming: A Developer’s Guide to Parallel Computing with GPUs, vol. 44. Elsevier, Amsterdam (2012)
Wikipedia: PCI express. https://www.top500.org/lists/2017/11/slides/
Laosooksathit, S., Nassar, R., Leangsuksun, C., Paun, M.: Reliability-aware performance model for optimal gpu-enabled cluster environment. J. Supercomput. 68(3), 1630–1651 (2014)
Zhang, L., Li, K., Li, C., Li, K.: Bi-objective workflow scheduling of the energy consumption and reliability in heterogeneous computing systems. Inf. Sci. 379, 241–256 (2016)
Thanakornworakij, T., Nassar, R., Leangsuksun, C.B., Paun, M.: Reliability model of a system of k nodes with simultaneous failures for high-performance computing applications. Int. J. High Perform. Comput. Appl. 27(4), 474–482 (2013)
NVIDIA GeForce GTX680: The Fastest, Most Efficient GPU Ever Built. NVIDIA, Santa Clara (2012)
NVIDIA GeForce GTX980: Featuring Maxwell, The Most Advanced GPU Ever Made. NVIDIA Corporation, White Paper (2014)
Liu, B., Chen, Q.: Implementation and optimization of intra prediction in H264 video parallel decoder on CUDA. In: IEEE Fifth International Conference on Advanced Computational Intelligence, pp. 119–122 (2012)
Vacavant, A., Chateau, T., Wilhelm, A.: A benchmark dataset for outdoor foreground/background extraction. In: International Conference on Computer Vision, pp. 291–300 (2012)
Lecun, Y.: LeNet-5, Convolutional Neural Networks
Krizhevsky, A., Sutskever, I., Hinton, G.E.: Imagenet Classification with Deep Convolutional Neural Networks, pp. 1097–1105 (2012)
Yuan, Z.W., Zhang, J.: Feature Extraction and Image Retrieval Based on Alexnet, p. 100330E(2016)
Acknowledgements
The authors gratefully acknowledge the support of the National Natural Science Foundation of China (61572325 and 60970012); Ph.D. Programs Foundation of Ministry of Education of China (Grant No. 20113120110008); Shanghai Key Programs of Science and Technology (14511107902 and 16DZ1203603); Shanghai Leading Academic Discipline Project (No. XTKX2012); Shanghai Engineering Research Center Project (GCZX14014 and C14001).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Fang, Y., Chen, Q. A real-time and reliable dynamic migration model for concurrent taskflow in a GPU cluster. Cluster Comput 22, 585–599 (2019). https://doi.org/10.1007/s10586-018-2866-8
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10586-018-2866-8