RESEARCH AND INNOVATION

Managing IOT Services the DevOps Way

 
November 27, 2018

The Internet of Things (IoT) is now all pervasive and disrupting every industry, right from travel, transportation and hospitality to manufacturing. We live in an era where the IoT is influencing all aspects of our lives, driving continuous digital transformation, where speed of enterprise innovation is critical to delivering new business models and customer value.

The business case for any IoT project starts with optimization, personalization and quality of experience. The IoT’s development experience has to be simple and resource efficient to further accelerate adoption of IoT technology among enterprises.  For IoT projects, delivering quickly, rolling out frequent updates, ensuring safety and also driving customer adoption - all assume new complexity.

DevOps teams in software companies have been accelerating the software development life cycle (SDLC) by embedding agile processes in the workflow. The DevOps approach provides framework and tools for process automation of the existing SDLC and to rapidly scale up infrastructure. IoT DevOps will embed the lean startup methodology in organizations looking to drive new services and revenue streams.

Here are four key reasons why IoT services should be managed the DevOps way:

1: To manage complexity and enable rapid prototyping in a constrained environment:

In an IoT project lifecycle, devices have to connect over a variety of transportation layers and protocols and also support configurations that are complex in nature. In addition, it is challenging to build IoT applications and roll out new product features in real-time. Developers working in an IoT environment face various constraints such as low-memory, small processors and network bandwidth that makes rapid prototyping a nightmare.

IoT DevOps is critical to ensuring discipline among developers, testers or QA, and IoT infrastructure and operations teams by using tools for monitoring and alerts and cognitive automation, thus ensuring continuous improvement.

2: To manage scale and be proactively secure:

For IoT projects, it is necessary to secure information coming in from sensors and appliances, to prevent it from getting compromised. All IoT systems must patch up software vulnerability in few minutes and be proactively secure. In addition, they should quickly scale from 10 devices (pilot stage) to 100,000 devices (production stage) in a matter of weeks; so systems should be designed to manage this load.

IoT DevOps will help developers, testers and QA, and the IoT operation teams by automating release, vulnerability, and patch management in real-time and install patches and updates without disrupting user experience.

3: To manage rigorous interoperability testing and ensure availability:

IoT applications can have multiple software and hardware configurations within solutions.

Testing the security, performance, connectivity, and a number of parameters for millions of devices is not an easy task. These different configurations and attributes need to be tested before deploying to production. Automated testing methodologies adopted as part of DevOps implementations will ensure that QA processes have considerable test coverage across these different configurations and attributes. Any service virtualization and emulation capabilities developed as part of the automation of testing could benefit change management. Moreover, some of the same toolsets can be used during production for monitoring solutions. Thus automation methodologies developed as part of the IoT DevOps process help handle scale as well as improve availability and security.

4: To build continuously and add seamless intelligence to devices:

In the IoT project lifecycle, implementing continuous delivery and constantly adding intelligence to devices is of paramount importance. For IoT businesses and innovators, DevOps ensures that they can build continuous and add seamless intelligence to devices by having the right framework, tools, processes and people in place.

IoT solutions need to be delivered and operated in a manner where they are able to deliver business value continually. Adopting DevOps methodologies for the entire lifecycle - taking into account the characteristics of IoT solutions - will be critical for Service Delivery teams to deliver on the promises of value creation.

Ravindran is a Functional Manager within the IoT Engineering team in TCS who is currently working on Infrastructure Automation, DevOps toolchains, Security toolsets and how they can be effectively used in deploying and Operating IoT Solutions. He has spent more than 25 years in Infrastructure Management/Service Management covering Servers, Networking and Security in Service Provider as well as End User Organizations.