Engineering the Future of Software
16–18 October 2017: Conference & Tutorials
18–19 October 2017: Training
London, UK

Application scaling over the edge: Microservice architecture in industrial applications

Fei Li (Siemens)
16:5017:40 Monday, 16 October 2017
Microservices, pros and cons
Location: King's Suite - Balmoral
Secondary topics:  Best Practice, Case Study
Average rating: **...
(2.57, 14 ratings)

Prerequisite Knowledge

  • A basic understanding of microservice architecture and edge computing

What you'll learn

  • Understand how to build a scalable application on resource-constrained devices by dynamically utilizing resources in the cloud


Edge computing is aimed at enabling analytics on devices that are close to physical data sources. Edge devices are generally characterized by limited computational resources and near-real-time responses to the needs of users. Application scaling on edge devices is a key success factor for edge computing and requires a software architecture that regards cloud and edge as a continuum of resources whose use is determined dynamically at runtime.

Fei Li explores the challenges of application scaling in edge computing with an analytics application in the context of the industrial internet of things (IIoT). Edge devices that are directly connected to industrial equipment, such as sensors, robots, and meters, are utilized to provide near-real-time analytics. The cloud is leveraged on the fly to offload analytics jobs when resources on the devices do not suffice. Microservice architecture is adopted on both cloud and edge with a strong focus on critical requirements in the IIoT—safety, performance, and reliability—which cannot be compromised during application execution.

Topics include:

  • Novel architectural patterns needed to effectively apply microservice architecture in edge computing
  • How service invocations are dynamically routed between cloud and edge to balance the need for maximal scalability and minimal latency
  • How containerized applications are continuously deployed to heterogeneous hosting environments in the cloud and on edge devices
  • Lessons learned from addressing these problems in the context of critical industrial applications
Photo of Fei Li

Fei Li


Fei Li is a software architect and research scientist at Siemens AG in Austria, where he designs and develops prototypes in the fields of edge computing, industrial automation, and manufacturing management, working closely with various production teams to transfer practical knowledge into production. Fei has more than 10 years of experience designing and developing large-scale distributed software. He has authored more than 30 publications on the subjects of service-oriented computing, the internet of things, and smart city solutions.

Comments on this page are now closed.


27/10/2017 13:45 BST

Hi Fei, I looked through presentation once more and on slide 23 from diagram I guess that jenkins deploys new updates to the edge. I wonder how it is possible, when jenkins is the cloud and edge is in the client’s local network.