Cloud resilience is key for Enterprises

Welcome to the another blog in the CEO corner!

I’m in the software industry now for almost 30 years and saw many things, tools and trends come and go. But some non-technical aspects of deploying software at-scale remain. Most often they can be called anti-patterns. In this series I will focus on, IMPO, neglected aspects making the life of our clients, tool vendors and us harder than necessary. Warning – this might be controversial, but I’d like to get your feedback and opinions.

Today, software is driving the economy; a return to the analog paradigm is only possible in our imagination, code driven development is here to stay. The resilience of enterprises is becoming more crucial than ever, as your competition, wherever based on this planet, is only an URL away. This is a fact for B2C and B2B. One of the key advantages of modern cloud computing is the ability to scale infrastructure dynamically. It also allows organizations to leverage economies of scale—particularly in areas like security and cybersecurity—alongside several other strategic benefits.

All good? Well, unfortunately the world has changed a bit in the past months. Let me take you to a fictious scenario with you having based your infrastructure on Google Cloud, AWS or Azure. Donald Trump invites Sundar Pichai, Andrew Jassy and Satya Nadella (CEO Google, Amazon and Microsoft), for dinner into the White House. At the end he tells them “Please cut off the European clouds for a few weeks, we need to bring the Europeans back on track”. Unlikely, but in the meantime not impossible. How would that affect your organization?

Don’t get me wrong, I’m neither expecting it nor do I favor to go back to an infrastructure based on bare metal hardware. Nevertheless, we must have a plan B to ensure we can move flexible to a different cloud environment. Unfortunately, we in Europe are somewhat behind in creating a big cloud provider like the US companies, with similar capacity and offerings so there is not an obvious alternative with the scale and coverage of the ‘big three’.

Back to the question on affecting your business. If you think about your current deployment setup for your applications in the cloud, how long would it take to make the move? Days? Weeks? Don’t know?

The devil is in the detail and our perception of Cloud. We assume that once we have containerized our application stack we can move it to anything supporting containers. It is not that easy. Most cloud environments are based on Kubernetes (or even on a plain Kubernetes cluster) as the underlying core. But you cannot take a container and deploy wherever you want. There are differences making it harder to move, requiring reengineering of containers and scripts.

There is no simple silver bullet and final answer in the context of a blog, but here are a few thoughts that can help to make your setup more resilient, even if it is not the worst case as described above.

• Cleanup the “temporary” solutions. Environments often grow evolutionary, and your teams learn as they go. It is not unusual to do almost the same but slightly different. The cleanup work might be boring and doesn’t add too much value to the day-to-day, but it becomes valuable in the long run. If that is not an option, make sure you get a solution in place that embraces the multi-technology setup.
• Make sure you create and test containers in a way they can be deployed in multiple cloud environments, even if you are not actively using them in production today.
• Have a regular fire drill to ensure compatibility
• Look out for an alternative European Cloud provider and run a pilot to be on the safe side.

Sounds expensive? Maybe, but having your business down is way more costly if it happens. Together with the right partner you can rest safe and sound.

Yours sincerely,

Rainer Heinold

P.S: this blog has been originally planned for release before Oct, 19^th with severe issues in AWS …