Cloud-Native Adoption Increasing Rapidly Since Pandemic

When quarantines forced a remote-work model, businesses had an urgent need for better ways to manage and protect their distributed data and applications.

By  Erin Poulson

By Erin Poulson May 26, 2021

The coronavirus pandemic forever changed the way many businesses operate – and this transformation has been the driver of significant adoption of cloud-native IT solutions and capabilities.

So said Vaibhav Kamra, chief technology officer of Kasten by Veeam, to host Alex Williams in a recent New Stack Makers podcast. Kasten is a market leader for Kubernetes backup, disaster recovery, and mobility and in early 2020 the company was strategizing on how to attract customers and scale out its services. When states began to shut down due to the pandemic, the conversation took a sharp turn. Suddenly companies across the nation were scrambling to move business online.

“[The] adoption of cloud-native ecosystems, technologies, and Kubernetes really started to accelerate,” said Kamra. “A lot of folks have taken this opportunity to kind of double down on their container and cloud native adoption. And that has really helped drive a lot of the conversations that we have and the adoption we've seen.”

Related

Clearing the Fog Around Cloud Native

Many of those Kasten conversations revolved around data management and protection, and how to guide customers new to a cloud-native approach through those critical issues. One thing that made that job easier was the fact that cloud-native technologies are getting much more recognition and support from traditional IT vendors.

In fact, according to a ZDNet article, a 2020 survey by the Cloud Native Computing Foundation reported that 78% of the companies that use containers in production are using Kubernetes.

Cloud-Native Approach Requires a Different Mindset

Moving to containers / Kubernetes is often done when companies turn to public cloud, but it requires a shift in thinking. When it comes to building cloud native vs. monolithic apps, the way infrastructure is provisioned and managed to facilitate developers changes, according to Sean Roth, director of product marketing at Nutanix.

“Cloud native implies 'infrastructure as code,'’ said Roth. “Ideally, developers specify the resources applications need in podspecs, like scripts instructing Kubernetes how to orchestrate application workloads. DevOps has emerged as a new function enabling IT Ops and software developers to work in a more coordinated manner.”

In many cases, developers speed ahead with their applications and IT operations struggles to keep up with their resource needs, he said. “DevOps, and the mindset that goes with it, alleviates that roadblock to a degree.”

The differences between traditional and cloud-native computing that Kasten is seeing, Karma explained, are “the changing roles and how applications are built, and [that has] a lot to do with the application architecture.”

As applications are rearchitected into microservices running on containers, “there’s a lot less visibility into what the applications are doing now,” said Kamra. “Applications are changing quite dynamically. … We’re starting to see things like polyglot persistence, where individual services in the application are using different storage system or different data services.”

“What does an application look like in this environment?” said Kamra. “It is a collection of a lot of things. All of these infrastructure components, certificates, secrets, configurations, personas, user accounts, data volumes, database services that might actually live outside of Kubernetes.”

The loss of visibility into an application and all of its myriad components can be a challenge for IT teams when it comes to protecting data or moving it around within those applications. 

“Observability is indeed a top Kubernetes challenge, throughout the stack and from the application container perspective,” said Roth.

Kasten’s K10 solution includes object storage that gives needed visibility and simplifies monitoring of the applications themselves. Protection for microservice-based applications changes from their monolithic application ancestors.

Related

App Dev at the Speed of Business

“You've got these microservice architectures,” said Kamra. “They can be very different every time you do a deploy – your application could be changing. So we wanted to give application developers the control over data management, which traditionally they haven't had. It was always under operations.”

Kasten’s K10 solution allows developers to create a blueprint and specify how they want things like backup and recovery to work. And policies are critical to these blueprints.

“Data management and data protection have moved from being very static to more dynamic in these environments,” said Kamra. “It's just a factor of how people are deploying applications in their environment. … So if [the] definition of an application is changing every time [it runs], then the way you set up your data management has to be policy-based. It cannot be static.”

Current Challenges Drive the Future of Container Capabilities

IT infrastructure (compute, storage, network) self service is an important part of what cloud native technologies enable, said Roth. 

“In the ideal scenario, developers can carve out their own resources without having to deal with the complexity of the underlying infrastructure,” he said. “Resources allocation is neatly abstracted away from that, making it easy for developers”.

He pointed to Nutanix Karbon as an example of this concept. “DevOps pros and app developers can leverage simple workflows to deploy new nodes, scale the cluster, provision persistent storage, etc,” he said.

The dynamic nature of cloud-native applications means container capabilities are evolving. Kamra said one area of current growth is self-service.

“We are seeing focus on enabling capabilities in the platform and allowing … developers to actually leverage them on their own,” he said. 

For instance, a customer will request the ability to deploy data management but wants to enable developers to back up or restore their own applications. Or an IT team wants to enable developers to deploy a new database when needed without having to get IT involved to protect the data.

Related

Cloud-Native Innovations Fuel Growth

A potential problem with self-service is that the overlap between IT and developers could result in critical tasks being left undone, such as backup, if each group assumes the other is handling it.

One of the technologies addressing that issue is Open Policy Agent (OPA), said Kamra. According to the Cloud Native Academy, OPA is “a policy engine that automates and unifies the implementation of policies across IT environments, especially in cloud native applications.”

“The automation will make sure the right things happen,” Kamra explained. “Whether it is platform ops taking over those responsibilities, or developer groups—it changes depending on the environment, the culture, the groups. But people are making sure that these things are in place so that things don't get dropped.”

As an example, platform ops could go into the system and create a policy that says if an application is deployed, a data protection policy must also be defined for it.

Cloud native applications – built as collections of smaller interrelated components – by their nature result in more efficient use of the underlying infrastructure, according to Roth. 

“In taking utilization efficiency even further, a multi-tenancy approach is very effective, but can be challenging to establish,” he said.

Multi-tenancy, or multi-cluster management, is a concern, said Kamra. 

Related

The Disruptive Force of Cloud Native

“Where you went from, maybe last year just seeing customers with a handful of Kubernetes clusters to now having so many of them that managing all of them as a unit is really what is top of mind,” he said. “Multitenancy shows up a lot. Especially not just with service provider environments, but even within the enterprise. You're starting to see that, especially as people want to enable self-service in these environments.”

Kamra said his team at Kasten have been working on strengthening multi-cluster management capabilities in their products, and the Kubernetes community’s work on multi-cluster has recently begun to pick up steam.

In addition to multi-cluster management, Kamra said things like security and ransomware are very high, very top of mind for customers right now. While they’re not challenges specific to Kubernetes or cloud-native technology, strategies to address those concerns will likely find their way into near-future cloud-native solutions.

Erin Poulson is a contributing writer who specializes in IT and business topics.

© 2021 Nutanix, Inc. All rights reserved. For additional legal information, please go here.