TY - GEN
T1 - Network optimizations on prediction server with multiple predictors
AU - Okuyama, Kaho
AU - Tokusashi, Yuta
AU - Iwata, Takuma
AU - Tsukada, Mineto
AU - Kishiki, Kazumasa
AU - Matsutani, Hiroki
N1 - Funding Information:
Acknowledgements This work was supported by JST CREST Grant Number JPMJCR1785, Japan.
Publisher Copyright:
© 2018 IEEE.
PY - 2018/7/2
Y1 - 2018/7/2
N2 - Toward machine learning based prediction services, the prediction server has multiple predictors and selects an appropriate one based on past feedbacks from the clients. In this case, three messages including request, reply, and feedback, are required for each prediction request. Packets are typically transmitted and received via a network protocol stack in OS kernel, and performance improvement can be expected by avoiding the protocol stack since it degrades the communication performance especially for small packets. We implement the prediction server using network optimization approaches including kernel-bypassing and in-NIC processing approaches. Evaluation results show that these network optimizations are beneficial to improve the prediction server performance compared to a baseline prediction server using a standard network protocol stack.
AB - Toward machine learning based prediction services, the prediction server has multiple predictors and selects an appropriate one based on past feedbacks from the clients. In this case, three messages including request, reply, and feedback, are required for each prediction request. Packets are typically transmitted and received via a network protocol stack in OS kernel, and performance improvement can be expected by avoiding the protocol stack since it degrades the communication performance especially for small packets. We implement the prediction server using network optimization approaches including kernel-bypassing and in-NIC processing approaches. Evaluation results show that these network optimizations are beneficial to improve the prediction server performance compared to a baseline prediction server using a standard network protocol stack.
KW - DPDK
KW - FPGA NIC
KW - Prediction Server
UR - http://www.scopus.com/inward/record.url?scp=85063893075&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85063893075&partnerID=8YFLogxK
U2 - 10.1109/BDCloud.2018.00155
DO - 10.1109/BDCloud.2018.00155
M3 - Conference contribution
AN - SCOPUS:85063893075
T3 - Proceedings - 16th IEEE International Symposium on Parallel and Distributed Processing with Applications, 17th IEEE International Conference on Ubiquitous Computing and Communications, 8th IEEE International Conference on Big Data and Cloud Computing, 11th IEEE International Conference on Social Computing and Networking and 8th IEEE International Conference on Sustainable Computing and Communications, ISPA/IUCC/BDCloud/SocialCom/SustainCom 2018
SP - 1044
EP - 1045
BT - Proceedings - 16th IEEE International Symposium on Parallel and Distributed Processing with Applications, 17th IEEE International Conference on Ubiquitous Computing and Communications, 8th IEEE International Conference on Big Data and Cloud Computing, 11th IEEE International Conference on Social Computing and Networking and 8th IEEE International Conference on Sustainable Computing and Communications, ISPA/IUCC/BDCloud/SocialCom/SustainCom 2018
A2 - Chen, Jinjun
A2 - Yang, Laurence T.
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 16th IEEE International Symposium on Parallel and Distributed Processing with Applications, 17th IEEE International Conference on Ubiquitous Computing and Communications, 8th IEEE International Conference on Big Data and Cloud Computing, 11th IEEE International Conference on Social Computing and Networking and 8th IEEE International Conference on Sustainable Computing and Communications, ISPA/IUCC/BDCloud/SocialCom/SustainCom 2018
Y2 - 11 December 2018 through 13 December 2018
ER -