Please share a few words about yourself for people that don’t know you yet (if they exist 😊 )
I’ve been in IT since 1998 and have worked for some of the largest technology companies in the world including Microsoft and Amazon Web Services (AWS). I’m the founder and Chief Technologist at CloudSkills.io, a cloud training and advisory services firm. I’m a 5x published author, 3x college drop out, international conference speaker, Microsoft Azure MVP, and host of the CloudSkills.fm podcast.
What barriers do you think companies face when it comes to transitioning to the cloud?
I think the biggest challenge is starting with a solid foundation. By far the biggest issue I’ve seen in the field is new customers struggling with cloud governance. It’s not uncommon to walk into a customer environment — when a company is new to cloud — and see enormous overspending, overly permissive security controls, and absolutely zero automation, which contributes to the element of human error.
What types of companies do you think have not yet moved to the cloud and why?
Legacy applications, lack of skills on the team, and strict security requirements are the common blockers I’ve seen preventing enterprises from getting into the cloud. You have to remember that most enterprises are used to a rhythm where things move on an annual or semi annual basis. When you have tens of thousands of users throughout your IT organization, shifting to new processes and implementing a culture of continuous learning is a long term investment that will not be achieved overnight.
I’d love to hear what you think around the topic of Multi Cloud: Do you think that companies are actually moving to it? I understand that companies do not want to be limited to a provider locked, but I see how it can be especially complex to manage multiple clouds in parallel.
I believe the vendor lock-in concern has merit under certain circumstances. However, it’s still too early for most organizations to go all in on multi-cloud. If you can’t get your arms around one cloud, what makes you think you can do two?
We need to do a better job of recognizing that the average IT organization does not have the same skillset as the teams at Netflix, Capital One, or Uber. The maturity of multi-cloud management tools is not even close to being ready for primetime, unless you’re strictly using Kubernetes for 100% of everything that you do, which is not the case for enterprises.
Do you think Kubernetes is the future? Do you have any tips for controlling the cost of the Kubernetes clusters?
I don’t think Kubernetes is the entire future, but I believe it will be a big part of it, either directly, or as some offshoot innovation that is born from it.
I think there are many teams out there right now using managed Kubernetes services that could do a better job of controlling costs by not leaving so many resources on the table. For example, I see a lot of teams building multiple managed k8s clusters with low pod density. This typically happens in an effort to implement isolation between teams, but logical isolation using namespaces on fewer managed clusters is often a better approach for increasing pod density in the cluster and using more of the available resources.
After moving to the cloud, how do you think a company can be under budgetary control?
Resource tagging and performance monitoring are a couple of great examples that often get ignored. Way too many companies are spending money on idle compute resources, detached storage volumes, and “proof of concept” environments that get forgotten about and left running for weeks on end.
I often encourage customers to consider things like reserved virtual machine instances, which reduces hourly run rate when making a long-term commitment with the provider. But, that assumes virtual machines are even the right choice.
Businesses should explore new ways of running their applications and offload time consuming processes to managed services. Moving to a serverless model for web apps and event-driven automation solutions are two interesting ways to pay only for what you use. Businesses with curiosity and willingness to innovate will be able to pull this off easier than those who resist the concept of trying something new.
Do you think that “Cost Optimization” is something that is a one-time action or something that continues daily?
I believe that cost optimization is a continuous practice that will have to evolve as time goes on. Cloud platforms will continue to be a moving target, and they should be, because innovation requires new ways of thinking.
What do you think will be the next big thing in cloud services?
I don’t know, but I can promise you this; if your company doesn’t have a culture of learning, experimentation, and being OK with failure, then you’re fighting an uphill battle.
What do you think that companies in the cloud services space need to improve?
Educating customers has to be put on a pedestal. You can’t get people to adopt your services if they don’t know how to use it. We’ve already seen some good efforts from the likes of Microsoft, Amazon, and Google when it comes to stepping up in terms of documentation. But cloud service providers need to do more to lower the barrier to entry. The way to do that is by educating customers and making things simple to use and understand.
The most significant tip that you can give for a new customer that migrates from on-premises to the cloud?
Recognize that the value of the cloud comes from solving problems in new ways. It’s going to take a lot more than lifting and shifting what is already on-premises to the cloud.
The blog post was written by the CEO of Pileus