The Future of Cloud Connectivity in an IoT World
Imagine your company has just completed development of a great application solution that leverages new Internet of Things (IoT) components.
The application has been rigorously tested in a comprehensive manner, meeting and exceeding all functionality, performance, and security tests. The associated connectivity points are based on state-of-the-art software-defined networking and network function virtualization frameworks.
It gets rolled into production and, voila! Multiple outages and unsatisfactory customer experiences. Does this sound familiar?
When new technologies are embraced and operationalized, they usually fail sooner rather than later. The IoT, new architectures, and cloud systems are developing into perfect storms that will take time to develop and move on to maturity, eventually providing calm, consistent conditions.
Based on historical data, we know that as new technology is introduced, learning curves are required, mistakes are made, and defects appear out of nowhere. Now, let’s think about additional security risks and threats that are more often the major targets on the radar. Moving forward, the risks intensify further. Fortunately, technical staff are inherent optimists (something that usually surprises the business), so in their minds, everything is fine until things go wrong.
Does this mean we should never use new technology?
Of course not. However, it does mean that developers and testers must be allocated adequate budgets, schedules, and tools to plan for failure so that risks can be mitigated and benefits realized.
Remember, I said the technical staff are optimists, so the reality is that they will underestimate what is really required. The technical community must clearly communicate the risks (financial, legal, and credibility-wise) and the multiple points of potential failure or security vulnerabilities so that stakeholders can make informed decisions.
The IoT, cloud, and connected world we will live in require tremendous abilities to recognize all the end-to-end solution components and the integration challenges that accompany them. Each component is on the critical path, including, of course, networking and connectivity. Therefore, planning for component failure regardless of root cause has tremendous benefits, especially as several new sets of disruptive technology rolls into production.
Planning for failure is the key to successful complex IoT cloud deployments.
Steven Woodward is presenting the session Future Perspective: Cloud Connectivity in an IoT World at the 2016 IoT Dev + Test conference. Learn more at https://iotdevtest.techwell.com.
