Kubernetes has emerged as an answer to the hybrid-cloud portability problem in the world of computing. But the open-source platform for orchestrating containers (a virtualized method for running distributed applications) doesn’t always blend perfectly with older software. This friction can leave users with a lopsided Kubernetes experience: great for newer stateless apps, but just OK for stateful ones.

Kubernetes was built to be stateless, which means that it’s not necessarily aware of its underlying data. Many enterprises want their data to remain stateful. They want to marry stateful data storage with Kubernetes and get the same performance they get with stateless applications. These businesses have many data applications, workloads and platforms that predate Kubernetes.

“If you don’t really bring them into the fold, you really are not solving the real business challenges that people have today,” said Partha Seetala (pictured, left), chief technology officer of Robin.io. Marrying these to a microservices platform such as Kubernetes requires a special storage subsystem and data services, he added. To address this, Robin has partnered with Google to bring storage and advanced data management capabilities to the Google Cloud Services Platform.

To discuss the news, Seetala and Radhesh Menon (pictured, right), chief marketing officer of Robin.io, spoke with Peter Burris, host of theCUBE, SiliconANGLE Media’s mobile livestreaming studio, at theCUBE’s studio in Palo Alto, California. The discussion covers Robin’s partnership with Google and how the company’s storage and data services take stateful apps on Kubernetes to a new level.

Ambitious APIs

At the heart of Robin’s partnership with Google is standardized application programming interfaces powerful enough to support the running of data-hungry workloads. It’s the result of an engineer-to-engineer collaboration between Robin and Google, designed to detangle the knots tying up large-scale data management.

“Because we’ve had field deployments, like Oracle RAC, where people are deploying multiple petabytes of storage in this single Kubernetes Robin cluster — all that learning and experience has contributed towards this joined … effort to create the standard data management API,” Seetala said.

From bare metal to the multicloud, Robin’s APIs have brought good fortune to the Kubernetes startup. Since landing $17 million in a Series B funding round in 2018, Robin has taken on some of the stickiest problems facing the Kubernetes ecosystem. Last year saw the launch of Robin’s hyperconverged Kubernetes platform, aimed at artificial intelligence and machine learning workloads with automated compute, storage and networking.

Sophisticated APIs have also granted Robin the elite status of Google’s preferred storage solution for Kubernetes, which proves the startup’s staying power, according to the co-founders.

“There are about 25 or 30 different storage windows providing storage for Kubernetes — so what is so special? I think there’s something special that led us to this thing, this point, right? We took a very fundamentally different approach when we solved this problem for GKE, or for Kubernetes,” Seetala said.

Uplevel with app-level snapshots

Robin’s new services for Google Cloud and the Google Kubernetes Engine go beyond providing persistent storage to Kubernetes pods. For example, data services such as snapshots have been around for decades. But instead of mere storage-level snapshots, Robin snapshots entire complex, stateful workloads.

The benefit of an app-level snapshot is that a developer does not need to ask a storage admin for clones of storage volumes. They just need a clone of a single application, and they can run queries on it right away. They can take a snapshot of a MongoDB cluster, for instance, and in minutes, they will have a functional cluster with live data on Kubernetes.

“We have customers in production where they have Oracle RAC [database] as a service offered on Robin,” Menon explained.

This snapshot feature comes packaged in Robin Storage, the preferred storage solution for GKE. It extends GKE to run any stateful enterprise application, including big data, database, artificial intelligence and machine learning. 

Watch the complete video interview below, and be sure to check out more of SiliconANGLE’s and theCUBE’s CUBE Conversations.

Photo: SiliconANGLE