The potential uses of Kubernetes seem as limitless as its scalability in the cloud.
As the movement to the cloud has grown, so has the use of containers as an effective way to package, distribute, and deploy applications. As surveys show, Kubernetes is the most widely used orchestration engine for managing cloud-native containers. Kubernetes automates deployment, auto-scaling, resource optimization, backup and recovery, and enables containers to run across different environments, eliminating the need to develop separate versions for each operating environment. Kubernetes can run on a single cloud, on multiple clouds, on-premises, at the edge, or in hybrid environments.
Most business and government organizations are seeking to modernize their infrastructures and gain agility by deploying Kubernetes. However the use cases extend to a wide variety of areas, including some of the most innovative and exciting applications being developed today. Let’s take a look at some of these kool Kubernetes use cases.
According to the World Health Organization (WHO), one-half the world’s population lacks access to healthcare, which has led companies to innovate to fill the gap. One example is Babylon Health, whose goal is to put accessible and affordable healthcare in the hands of every person on earth using a combination of artificial intelligence (AI) and human expertise. The company’s suite of tools has been designed to let patients quickly access information through their phones and enable clinicians to make better decisions based on information entered by the user.
Moving to Kubernetes gave Babylon Health an effective way to orchestrate complex hyperparameter research, overcome the compute deficit of their on-premises installation, and enable researchers to run experiments and receive answers faster. For example, by using Kubernetes and Kubeflow, they created a medical AI solution that could reduce the response time from their clinical validation checker from 10 hours to 20 minutes.
Expanding the Boundaries of Medical Image Processing
Another healthcare area in which AI and Kubernetes are making a difference is medical image processing. An ever-increasing number of images from the growing number of scans being used in medical diagnoses is rapidly outstripping the ability of human radiology technicians to analyze the data. AI systems with deep-learning capabilities can streamline processes, analyze complex images, and alert technicians to abnormalities.
Among the companies focused on improving medical imaging through AI and Kubernetes is Dell, which has developed a large-scale deep learning solution for medical imaging using NVIDIA’s Clara AI healthcare application framework, Clara Deploy SDK, and Dell EMC PowerEdge NVIDIA GPU Cloud (NGC) accelerated systems. The system enables NGC containers to work together to provide an end-to-end medical image processing workflow in Kubernetes. Computational workflows for CT, MRI, and ultrasound data leverage Docker and Kubernetes to orchestrate medical image workflows and connect to picture archiving and communications systems (PACs) or scale medical instrument applications.
BMD Software has created a medical imaging repository based on Kubernetes that combines scalability, distribution, and intelligent management. The solution uses state-of-the-art cloud technologies, the Kubernetes ecosystem that enables automation of operational tasks based on containerization of applications, and the open-source vender-neutral Dicoogle PACs application.
GE Healthcare has designed a platform called Edison Health Link that enables customers to deploy GE's next-generation imaging machines, applications, and analytics. Running on the D2iQ Kubernetes Platform, the system provides connectivity back to GE Healthcare's centrally managed private cloud.
Another project entitled “DeepCell Kiosk: scaling deep learning–enabled cellular image analysis with Kubernetes” takes in configuration details and creates a cluster on Google cloud that runs predefined deep-learning-enabled image analysis pipelines managed by Kubernetes. Kubernetes enables the DeepCell Kiosk to treat hardware as a resource that can be dynamically allocated to scale cluster sizes to meet data analysis demand.
Monitoring and Predicting Droughts and Floods
Developed by the Hydrological Sciences Laboratory at NASA’s Goddard Spaceflight Center, the Land Information System (LIS) is a Kubernetes-based solution for terrestrial hydrology monitoring and data assimilation. The land-surface model enables the monitoring and predicting of water hazards such as flood and drought in North America and around the globe.
Helping Make Energy More Efficient and Renewable
Because Kubernetes enables highly precise resource utilization, it is being touted as a sustainable energy tool that can minimize wasted energy in cloud data centers. A New Stack report describes how Kubernetes can significantly improve server utilization to help lower data center carbon emissions. A similar report describes how Kubernetes can help reduce an organization's carbon footprint. UK researchers describe a low carbon Kubernetes scheduler that can contribute to moving demand for more carbon intense electricity to less carbon intense electricity.
Shell has launched a Renewable and Energy Solutions initiative that aims to provide a reliable supply of electricity to 100 million people in developing countries by 2030. The platform is composed of Arrikto’s MLOps and Kubeflow running on Kubernetes to make machine learning workflows portable and scalable. The infrastructure runs on Amazon EKS and uses a GitOps approach to enable the DevOps team to build solutions on different Kubernetes clusters in a reproducible way. Shell says the solution enables its teams to build thousands of machine learning models in 2 hours instead of 4 weeks, and to reduce the time required to write the underlying code from 2 weeks to 4 hours.
Stockholm-based Rebase Energy is using the IBM Cloud Kubernetes Service as the infrastructure for its business aimed at enabling organizations of all sizes to optimize and share renewable energy through the company’s analytics tools. Rebase Energy also is using IBM ILOG CPLEX Optimization Studio, an AI program designed to help automate decision-making.
In “Building the solar plant of the future with cloud technologies,” Daniele Polencic of learnk8s.io describes how Kubernetes can be used to optimize the efficiency of large solar panel installations. Kubernetes in a solar plant, he explains, lets you benefit from a centralized scheduler to issue deployments, secure and encrypted updates delivered as containers, and a proven technology able to scale to thousands of devices. When a device fails, Kubernetes will reschedule all of the applications deployed on that computer to another.
Amazon Web Services offers cloud sustainability guidance, including design principles for sustainability in the cloud. The Cloud Native Computing Foundation (CNCF), which governs the open-source Kubernetes platform, has formed a working group for environmental sustainability whose mission is to promote sustainability awareness and develop a culture within the CNCF landscape to establish sustainability best practices and standards.
The U.S. military has made modernization an imperative. Among the initiatives employing Kubernetes is the U.S. Air Force’s Next Generation Air Dominance (NGAD) program, which seeks to establish air dominance through a variety of approaches, including manned, unmanned, optionally manned, cyber, and electronic formations that would not resemble the traditional “fighter.”
A tech demonstrator flew in 2020 with a Kubernetes delivery system for the weapons system, and NGAD fighters are expected to achieve significant numbers in service in the early 2030s. Key to the success of the new fighter is the ability to separate the flight control system from the rest of the aircraft’s computer system, which enables software to be installed on the fly.
Orchestrating Automotive Intelligence
As more and more intelligence is built into automobiles, Kubernetes is helping manufacturers like Mercedes Benz, Ford, Tesla, and BMW deliver the software and updates necessary for today’s connected cars. Although autonomous vehicles are still being perfected, the research and development using AI and Kubernetes is advancing the art of verification and validation. GM’s Cruise Automation division has extended the security capabilities of Kubernetes.
Automating Animation Rendering
When the team at Luma Pictures was creating visual effects for the 2019 film “Spiderman: Far From Home,” it wrestled with the problem of storing and manipulating massive amounts of simulation data. The solution involved moving its render pipeline to Google Cloud to take advantage of greater computing power. The success of this project has prompted the company to migrate more of its tools to Google Kubernetes Engine (GKE) in the cloud.
CoreWeave provides a Kubernetes-based rendering solution for animation studios that utilizes powerful NVIDIA GPUS in the cloud. The D2iQ Kubernetes Platform also enables animation studios to containerize the animation rendering process while easing collaboration among developers in dispersed locations.
Driving the Blockchain Platform
The openness, scalability, security, and upgradeability of Kubernetes has made it a natural fit for developing and managing blockchain applications. Best known as the underlying ledger system of cryptocurrency, blockchain offers a wide variety of benefits to businesses and government agencies. Blockchain can be used to provide real-time tracking of goods as they move through the supply chain, increase privacy for health records without exposing identifiable patient information, expedite home sales, streamline royalty and micropayments, support fraud and piracy prevention, and provide secure digital identity management for voting and government services. Use cases in the financial sector range from capital markets and regulatory auditing and compliance to peer-to-peer payments.
As the Harvard Business Review explains, quantum computing has been developed to perform tasks that traditional computing devices struggle with, even when the task is dispersed across millions of machines. Use cases include accelerating materials discovery and drug development in chemical and biochemical engineering, simulating market movements, speeding calculations in financial services, and reducing costly failures in complex manufacturing.
Kubernetes, says IBM’s Holly Cummins, will be a fundamental part of quantum computing orchestration. The cloud is a natural fit for quantum computation, she explains, because the cloud excels at pooling resources, enabling hardware access, and accelerating innovation.
In the “State of Cloud Native Development Q1 2021” survey, 65% of respondents said they were using containers for quantum development and 59% said they were using Kubernetes.
Scaling Pokémon Go to Support a Vast Global Audience
When Niantic launched the Pokémon Go mobile gaming application, player traffic surged to 50 times the initial target, ten times more than the worst-case scenario.
In response, Google Customer Reliability Engineering (CRE), a new service at the time, was able to seamlessly provision extra capacity on behalf of Niantic to stay ahead of the surge.
One of the more daring technical feats accomplished by Niantic and the Google CRE team was to upgrade to a newer version of Google Kubernetes Engine (GKE) that would enable more than a thousand additional nodes to be added to its container cluster.
Because the application ran on the GKE, Niantic’s DevOPs team was free to focus on deploying live changes for the players. Today, every player has an individual Kubernetes cluster. The Kubernetes infrastructure has enabled Niantic to turn Pokémon Go into a service for millions of players, continuously adapting and improving.
Going forward, Niantic is extending the Pokémon Go Kubernetes platform for additional services such as augmented reality and geolocation service, with the goal of enabling the world to build applications on top of it. This means extending the platform to third-party developers and making it easy for them to create and publish games.
KubeCraftAdmin. An immersive sandbox 3-D user interface that populates the world with different animals that correspond to the resources in a cluster, in different pens. Pods are pigs, ReplicaSets are cows, services are chickens, and horses are deployments. Animals will spawn to reflect the cluster state and die when resources get deleted.
Kube DOOM. A video game for visualizing and killing pods inside a Kubernetes cluster. It’s a modified version of the Doom video game.
Kube Invaders. Enables users to address the uncertainty of distributed systems at scale by putting into practice the principles of chaos engineering while playing Space Invaders. Kube Invaders lets players stress a Kubernetes cluster to see how resilient it is.
Whack-a-pod. A demo that turns a Kubernetes cluster into a Whack a Mole game in which the pods are moles. The goal is to try to knock down the service by killing the Kubernetes pods that run it.
HTC Vive. Lets you visualize Kubernetes cluster pods in virtual reality and enjoy an interactive experience while leveraging Kubernetes API service.
Creating a Personal Music Library
When Google Music shut down in 2020, Andy Smith decided it was time to host his music collection himself. Navidrome, an open-source web-based streaming server met his requirements for browser-based music streaming and mobile support as well as his nice-to-have user support. He built his music application on a Kubernetes cluster that he uses for tinkering and created a Helm chart for it.
Meeting the Demands of NFT Mania
Non-fungible tokens (NFTs), which are a way of proving ownership over digital goods, has become a $2 billion market almost overnight. NFTs can sell from a few dollars to a few million dollars, depending on the content of the NFT. Once a work is on a blockchain, it is unique, much like an original piece of art. Since it’s unique, people buy NFTs in the same way you might buy original artwork from your favorite artist. The record for a single NFT is $69 million.
Dapper Labs was one of the first ventures to provide services to meet the global demand for crypto collectibles. In 2019, the company implemented an observability stack that includes Grafana Cloud, PagerDuty, Prometheus, Kubernetes, and Google Cloud Platform as its core monitoring tools.
Limitless Kubernetes
While the above sampling shows the wide variety of innovative Kubernetes applications in use and under development, Kubernetes is still a relatively young technology. As Kubernetes continues to mature, there will be many more innovative uses, including AI-driven applications at the edge. The potential uses of Kubernetes seem as limitless as its scalability in the cloud. Stay tuned.
