Interviews and Reviews

Interview with Edward Viaene, Managing Director, IN4IT LLC

Edward Viaene interview

This is an exclusive interview with Edward Viaene, Managing Director, IN4IT LLC

Edward, can you tell us a bit about your journey in the tech industry? What piqued your interest initially, and how did that lead you to where you are today?

My tech journey is, for the most part, in the IT infrastructure world. In the late ’90s, I read about the Linux operating system in a computer magazine. I immediately became interested in it because it is a free platform that can also be used as an alternative to Unix. It opened the door to learning all sorts of free and open-source tools and programming languages, such as programming with Bash (a shell), PHP, MySQL, and Apache (a web server).

I studied all aspects of operating systems and programming languages to become the typical Linux System Administrator. Some of my time would be spent on screwing servers into racks in data centers. This was before Amazon and others came out with their cloud products. Every time you would need to spin up an instance, you would have to design and buy the actual hardware, install the operating system, the tools, the programming languages, put it in a data center, configure the network, and only then you’d be able to use it.

After a stint as an Information Risk Manager at a bank, I started working with Amazon Web Services (AWS) at a Big Data startup in London. This allowed me to make the transformation from a system administrator into the Cloud & DevOps world. With the startup experience under my belt, I started a Cloud & DevOps consultancy company with a friend from college to help companies run their applications on the cloud.

What are some of the biggest challenges you’ve faced while navigating your career in tech, and what key lessons did you learn from those experiences?

Early in my career, I was surprised that the major challenges were not technical. Programming, building infrastructure, or developing apps is only a small part of delivering the solution. There are many discussions, agreements, non-technical constraints, and sometimes politics that need to be considered during projects.

Growing your career and your company is often about how much you know but also about interpersonal relationships and getting things done. This is especially true in the consultancy business we are in now. Proposing the best technical solution might not win you a contract. There are many more considerations before a customer might engage with your company.

You mentioned creating online courses in Cloud & DevOps. How has the rapid evolution of cloud technology impacted the skills you teach and the way you teach them?

The rapid evolution of new technologies means that many people must learn new things. DevOps introduced new concepts for delivering software. Cloud services disrupted how we run and manage apps. New tooling like Kubernetes meant that everyone in the cloud space would have to learn how to use it.

The new DevOps methods and cloud tools were a golden opportunity for me to start creating online courses. Working earlier at a startup that used all these different technologies gave me a significant advantage. I could select the technologies that would make it and which wouldn’t. One of the clear winners was Kubernetes, which I incorporated into a DevOps course when they released their first stable version in 2015. Being one of the early ones to create content about it helped me release more courses about the same technology later.

Working with technologies that have a high release velocity has its challenges. You can make lectures about specific features that will eventually disappear or be replaced without knowing that upfront. Keeping yourself up-to-date with the latest features is necessary to adapt the course material.

I initially thought an online training course would be obsolete after 1-2 years. That didn’t prove to be true. Technology adoption will go in different waves all over the world. Technologies used for years in the United States might not have been adopted in companies in other parts of the world yet. I have courses launched in 2016 that still see enrollments today. You still have to adapt the material when software changes, but the core of the products often never changes.

For those looking to break into the Cloud & DevOps fields today, what advice would you give them based on your experience both as an instructor and a professional?

With so much online free content available, the cost of entry is lower than ever. The increased complexity doesn’t make being in the profession easier, but having many tools and materials to learn from has been a game-changer in the last few years. Enablers like Generative AI should be embraced to quickly get answers and code examples. When you don’t know the right keywords to find the answers, it can significantly speed up your search.

Make no mistake, though. To succeed, fundamental knowledge is still key. Using training material and GenAI is only part of it. Understanding exactly what is going on under the hood is still what distinguishes good from great. Once you understand the Cloud & DevOps concepts, it’s time to dig deeper into everything from networking to operating systems. There is a lot to learn. It’s not only about getting the Cloud Provider’s certificate and calling it a day. The landscape is, and will always be, huge.

You run a Cloud and DevOps boutique consultancy. What are some common misconceptions businesses have about cloud migration, and how do you advise them to approach it?

To benefit from the cloud, you must build the app from the ground up with a cloud-first approach. When you do a migration, many architectural decisions were made years ago. It’s not always possible to make the changes necessary to achieve the ideal cost and performance benefits the cloud can offer.

We think migrations are compromises. It’s often not in the business’s interest to make many changes to achieve the ideal architecture. We often take the approach of taking the first steps to migrate the application to the cloud and then creating a long-term plan to improve it. The long-term plan can include security improvements, performance improvements, or even switching certain services from unmanaged to managed services.

Maintainability is super important for our customers. Once the application is migrated, we optimize the infrastructure so that less maintenance is needed over time. In general, this means using more managed services and fewer generic compute services.

In your experience, what are some effective strategies for balancing cost optimization in the cloud with maintaining high performance and security standards?

Cost optimization and performance often go hand in hand on the cloud. If you can get more performance from your application, you’ll automatically optimize your cost by paying less for compute. In recent years, an interesting approach has been to migrate parts or all of the compute from Intel/AMD to the ARM64 CPU architecture. The ARM64 CPU architecture has a much better price-performance ratio and can help reduce costs.

Another sometimes overlooked benefit is to commit to your cloud compute for one or more years. The cloud uses the pay-as-you-go model, but once you are up and running on the cloud, you’ll have a certain baseload for the years to come. Committing to that baseload will easily reduce your compute costs by 20-30%.

While reducing costs and maintaining high performance, you also must ensure compliance with modern security standards on the cloud. Security benchmarks like CIS (Center for Internet Security) will help you identify the gaps there might be. Cloud providers have tools to monitor your infrastructure for security compliance continuously. Once issues appear, you can alert a team for manual remediation or even automatically delete cloud resources when non-compliant.

You emphasized the importance of choosing a single cloud provider and sticking with their ecosystem. How do you see this approach evolving with the rise of multi-cloud strategies?

A multi-cloud strategy is not for everyone. Cloud providers have very complex pricing, which might complicate a multi-cloud strategy. Inbound transfer (getting data into the cloud provider) is often free, but getting data out is charged. As you will most likely need to transfer data between cloud providers, designing the architecture is more difficult because of this pricing detail.

Luckily, new technologies can help when using multiple clouds simultaneously. Every cloud provider has a product to run containers, so building your application in containers will help with portability between clouds. Databases also have replication technologies to automatically synchronize the data between cloud providers. One limitation will always persist, though: the speed of light is a constant. Replication between regions will have latency you’ll not be able to avoid.

Sustainability is becoming increasingly crucial in tech.  Beyond migrating to ARM64 architecture, what other sustainable practices do you see gaining traction in the cloud computing space?

Today, cloud architects build their cloud architecture based on the customer’s demands and constraints. There’s not yet much visibility into how sustainable a specific cloud architecture is. This might change as cloud providers incorporate sustainability features directly within their platforms. Some data centers might have a lower carbon footprint that architects might consider when making decisions. When cloud providers give more insights into which products and hardware are more sustainable than others, cloud customers can make decisions with insights they previously did not have.

Looking ahead, what emerging trends in the tech industry are you most excited about, and why?

Everyone is looking at generative AI, the evident emerging trend. But do we know about the landscape beyond this trend? When Gen AI gets a real foothold, a lot more computing power will be necessary, which is being built right now. Cloud computing will need more electricity, so more power plants must be built. This all needs to be done efficiently enough. Otherwise, using an AI chat agent would be too expensive. If we have new incentives to make everything from electricity generation to hardware and software more efficient, then general computing will also benefit. Technological advances and more competition will lower everyone’s cost of cloud computing over time. That’s something to be excited about.

Comments
To Top

Pin It on Pinterest

Share This