Article Details

Title · Authors · Abstract · Keywords · DOI · References

Automatic Extension Method of Big Data Analysis Service Based on Reinforcement Learning

LI Jinfeng, SUO Qiang, REN Zhou

(Shanghai Prothentic Co., Ltd., Shanghai 201210, China)

Section：Artificial Intelligence · Vol. 2025, Issue 9 · Pages：32-36,42 · 2096-4706（2025）09-0032-06 · CLC number：TP18 · Document code：A · DOI：10.19850/j.cnki.2096-4706.2025.09.006

Abstract

As the demand for big data analytics services continues to grow, AaaS (Analysis as a Service) providers often rely on leasing IaaS (Infrastructure as a Service) virtual machine resources to build their service platforms. However, the dynamic changes in user requirements make how to flexibly adjust the type and number of virtual machines a key issue that needs to be resolved. In this paper, an automatic expansion method based on Q-Learning is proposed. By using the number of virtual machine instances as the action space of learning, the ε-greedy strategy is used to select actions, and the status and Q table values are dynamically updated. Through repeated iterations of this process, this method can maximize the utilization of virtual machine resources while minimizing the cost of AaaS providers. The experimental results show that this method can not only effectively reduce operating costs, but also ensure that the virtual machine CPU maintains high utilization for a long time.

Keywords

Big Data analysis; reinforcement learning; Artificial Intelligence

References

  1. [1] Wen Congmin, Liu Naixin Design and Research of Automatic Synchronization and Update Management System Based on Big Data Platform [J]. Modern Information Technology, 2023, 7 (14): 7-11
  2. [2] Zhang Shenglun, Zhang Nan Application of Cloud Computing in Network Communication Data Intelligent Operation and Maintenance System [J]. Modern Information Technology, 2022, 6 (6): 125-127+132
  3. [3] LI X，ZHOU J S，WEI X，et al. Topology-Aware Scheduling Framework for Microservice Applications in Cloud [J].IEEE Transactions On Parallel And Distribution Systems，2023，34（5）：1635-1649.
  4. [4] JANNAPUREDDY R，VIEN Q T，SHAH P，et al. An Auto-Scaling Framework for Analyzing Big Data in the Cloud Environment [J].Applied Science，2019，9（7）：1417-1429.
  5. [5] SI W，PAN L，LIU S J. A Cost-Driven Online Auto-Scaling Algorithm for Web Applications in Cloud Environments [J].Knowledge-Based Systems，2022，244：1-13.
  6. [6] GARÍ Y，MONGE D，MATEOS C，et al. Learning Budget Assignment Policies for Autoscaling Scientific Workflows in The Cloud [J].Cluster Computing，2020，23（1）：87-105.
  7. [7] GEORGE F，AROKIA R. A Novel Approach on Auto-Scaling for Resource Scheduling Using AWS [C]//2nd International Virtual Conference on Industry.Chennai：Springer，2021：99-109.
  8. [8] HOROVITZ S，ARIAN Y. Efficient Cloud Auto-Scaling with SLA Objective Using Q-Learning [C]//2018 IEEE 6th International Conference on Future Internet of Things and Cloud.Barcelona：IEEE，2018：85-92.
  9. [9] WEI Y，KUDENKO D，LIU S J，et al. A Reinforcement Learning Based Auto-Scaling Approach for SaaS Providers in Dynamic Cloud Environment [J/OL].Mathematical Problems in Engineering，2019.[2024-10-18].https://onlinelibrary.wiley.com/doi/10.1155/2019/5080647.
  10. [10] DING Z J，WANG S，JIANG C J. Kubernetes-Oriented Microservice Placement With Dynamic Resource Allocation [J].IEEE Transactions on Cloud Computing，2023，11（2）：1777-1793.

About the authors

Li Jinfeng (1983-), male, Han ethnicity, Shanghai native, PhD, research direction: Computer Science and Technology.