“I see Ops becoming the cool kid again” – CloudBolt’s CEO predicts the future

Brian Kelly, CEO of the hybrid cloud management platform CloudBolt believes that this year will hold some great things for IT. He even goes so far as to say that IT professionals will become the “cool kids’ again. The CEO spoke to us about what else he sees happening this year and why he has made these predictions.

Where do you see the cloud heading in 2020?

 Well, we’ve already seen the hybrid cloud really start to take off in 2019 with the Big Three investing big here. Google launched its own hybrid cloud platform – Anthos – powered by Kubernetes. Microsoft’s Azure Arc was recently announced at Microsoft Ignite. Amazon ended the year with AWS Outposts. And the latter here, Amazon, has always been a public cloud champion, so it is telling that they have fully embraced the reality of the hybrid cloud and that their public services may need to be deployed as part of a hybrid architecture. It means we’re really going to see 2020 be the year that hybrid cloud architectures reign supreme in the enterprise with this groundwork already laid. In 2019, we saw the hybrid cloud take off, but it’s going to be this year where we really see it start to soar.

How do you predict the cloud industry will change in general?

That’s an even bigger question. One thing I believe we will see is less hype and more concrete conversations around containers (Kubernetes, for example), serverless and mesh, and most importantly, choice. With the investments from the Big Three that we mentioned, the choice is going to be what is the biggest change – more platforms exist than ever before with easy integration of whichever tools that are already in place in an organization. IT needs to build the cloud their way. The cloud industry is changing for the better for IT from a flexibility standpoint and is only going to continue to improve.

You have already made several predictions for the future. What do you mean when you say, “self-service will become the standard?”

As IT focuses on governance and cost control, they still need to support the agility that the business continues to demand. Self-service is the answer because it guarantees both control and agility while making developers more productive.

Take for instance the world of private cloud. Many enterprises today still require developers to submit an IT ticket to request a virtual machine (VM). IT then typically needs to go to a dozen places to provision that machine–that could be the VMware, networking, and storage teams–all while needing to firefight critical issues in production. The bottom line is that this lengthy, legacy provisioning process, which can take days or weeks, slows things down. Instead, what if developers could order that VM through a self-service catalog and get their resource in minutes?

Similarly, take the world of public cloud. Developers can now get the speed they need but are they experts in configuring an AWS instance? And do they really need 2 TB of storage for their projects? Bottom line, unless some kind of guardrails are in place, costs can skyrocket and misconfigured workloads can result. Again, that’s where having IT-driven self-service can help. By creating blueprints of resources that developers can access–with cost and governance guardrails in place–IT can enable developers access to self-service while enforcing visibility and control.

 What problems are you seeing in DevOps and what advice do you give to overcome these?

There are two problems I’m seeing in the DevOps world. First, many IT operations groups have trouble supporting resource demands from developers and business teams. When Devs have to wait for resources to be provisioned, it leads to the second problem, “shadow IT.” Developers go directly to the cloud providers to get what they need and the next thing you know, there are parts of your IT infrastructure about which you don’t know anything. That leads to runaway costs and security risks.

The main advice I would give is, move to a self-service IT model that allows developers to access the resources they need–storage, compute, even complex multi-tier application stacks–when they need them. By thoughtfully creating and configuring these offerings with cost and governance guardrails in place, you maintain the agility that developers want with the control and oversight that IT needs. In a world that’s rapidly moving to hybrid cloud and multi-cloud, one of the key ways to do that is to invest in a cloud management platform that is extensible enough to work with any solutions or tools that DevOps requires.

What do you think will happen with the Dev and Ops relationship?

There will be a rebalancing of the relationship between Dev and Ops. For a while now, the “Dev” side of “DevOps” has been seen as the driver of innovation and the “Ops” side has been seen as more of an executor of this innovation (and at times, even an inhibitor). The devs get all the glory as the creative brains, while Ops is quietly in the background.

But this relationship is going to shift because of the world of the hybrid cloud, where self-service, controlling costs, and maintaining proper standards for governance and compliance are critical. Once IT demonstrates that they are focused on managing the infrastructure in a way that serves DevOps – namely, this simplifying access to cloud resources with self-service I mentioned – the tables will turn on the relationship. They will be seen more as a partner in infrastructure development and less as a mere “execution” arm.

Additionally, I think we will see Ops continue to become more “Devvy” as the drive towards greater and more intelligent automation demands that IT operations become more adept at leveraging scripting tools.

So, you think IT is becoming the ‘in’ thing again?

For sure – I see Ops becoming the cool kid again. As we go to a hybrid cloud world, I think we’ll see more chaos. We’ll see more users in the organization accessing public cloud directly. We’ll see a greater influx of tools and processes for workload provisioning and configuration. That could be home-grown scripts, infrastructure-as-code, and cloud management platforms. What does that result in? Enterprises will need to invest more in areas like cross-cloud visibility, governance (making sure the right people get access to what they need), and cost control (managing workloads to their target locations). This is the traditionally non-sexy Ops stuff, but in a world of hybrid cloud, these capabilities are critical. Ops will be king again and it will certainly be justified with the innovation needed to get to this self-service (automation), all the while controlling spend and resource usage to satisfy the business.