The iconic green and yellow John Deere tractors are a familiar sight in fields around the world. Through the company’s commitment to its smart industrial vision, the factories where the machines are made are being transformed.
Vast data stores fuel this effort—along with a well-stocked technology portfolio that spans cloud platforms, on-premises datacenters, and edge devices at the factories. John Deere’s modernization strategy makes the most of its assets while cultivating a path for the future.
An Azure hybrid cloud solution helps connect the dots across all these environments and puts the power of the cloud to work in the company’s existing infrastructure. The result is a unified view of operations across platforms that pivots on Azure Arc, a set of technologies that’s helping John Deere to optimize manufacturing operations. Together with Azure Arc–enabled SQL Managed Instance, the hybrid solution helps accelerate innovation and drive down operational costs.
Cloud-native patterns and practices help us deliver the factory solutions that are a key to our future success. Azure Arc–enabled data services help make this possible for us.
From the invention of the first commercially successful steel plow in 1837 to the current-day use of connected robots, John Deere puts innovation—and terabytes of precision data—to use in making its customers successful. According to Jason Wallin, Principal Architect for Infrastructure and Operations at John Deere, “In agriculture, it’s really only John Deere that seamlessly connects machines, people, technology, and insights to give our customers the advantage.”
That advantage is based in research that shows the value of big data in giving farmers a competitive edge. Farming faces huge challenges from volatile weather, weeds, pests, and diminishing resources. For the last decade, John Deere has helped customers collect and use data related to many aspects of their work—from weather and soil conditions to the equipment life cycle. Farmers use this information to manage their fleets, improve productivity, and create a sustainable living for today’s farming generation and for those of the future.
With so much data distributed across locations worldwide, it’s no longer business as usual at the 185-year-old company. “We’re always looking for ways to optimize our use of technology and help our customers with the challenges they face,” Wallin says.
Cloud modernization inspires on-premises aspirations
The company began exploring new strategies to support its digital goals and started an internal initiative to modernize IT.
“To create these digital solutions, we have to manage and process terabytes and terabytes of data,” notes Pinkal Patel, a Solution Architect in John Deere’s Infrastructure and Architecture organization. “Putting so much data into a datacenter is a challenge to scale and manage since those databases are dispersed across our worldwide locations. As we looked ahead, we were thinking, ‘How are we going to scale this?’”
Big data has a big footprint at John Deere, spanning multiple datacenters, cloud platforms, and factories around the world. The company runs thousands of Microsoft SQL Server Enterprise Edition databases in all sizes. They support mission-critical business and financial systems, homegrown apps, and everything in between. A specialized team manages it all.
Patel led a cross-functional team of John Deere subject matter experts. “We were looking at using the cloud to provide modernization, but we also wanted to do that with our on-premises tech, as well,” he explains.
The team included Group Engineering Manager Murali Subramanian, who manages mainframe infrastructure services and on-premises databases for John Deere. The extensive SQL Server environment is his team’s domain. As he recalls, “We wanted a solution that would allow us to manage resources that are on-premises and also in the cloud through a single control plane.”
Patel, Subramanian, and their colleagues found the solution with these capabilities when Microsoft announced Azure Arc, a set of technologies that unify on-premises, hybrid, and cross-cloud infrastructures, pushing the benefits of cloud platform as a service (PaaS)—such as data services—to local datacenters. Azure Arc-enabled data services provide elastic scale, built-in business continuity features like high availability and disaster recovery, and unified management for data workloads with or without direct Azure connection.
According to Patel, “We were looking for a solution to provide a cloud-like experience for on-prem resources when the private preview of Azure Arc–enabled data services was announced. We realized that it could help support our modernization journey. We were one of the early adopters.”
Azure Arc–enabled data services is a tool we’re using to help us achieve our goals in modernizing our IT environment to support the John Deere tech stack.
A move to modernize using Azure Arc–enabled data services
Azure Arc–enabled SQL Managed Instance and Azure Arc–enabled PostgreSQL Hyperscale are designed for hybrid work environments. They provide the benefits of a fully managed data service and cloud-like elasticity across environments, whether on-premises, in the cloud, or at the edge. Data services are also evergreen—automatic updates are applied according to the company’s policies, so John Deere no longer faces end-of-support disruption.
“That feature alone really simplified the patch and update model for us,” recalls Wallin.
For John Deere, Azure Arc–enabled SQL Managed Instance bridges the gap between cloud and on-premises operations. Designed to run in containers on Kubernetes, Azure Arc–enabled SQL Managed Instance is a service that can be created on John Deere’s existing infrastructure. It comes from the same code base that powers the core Microsoft SQL engine. The common features make it straightforward for the John Deere teams to lift their existing SQL Server workloads with minimal application and database changes and shift them to Azure Arc–enabled SQL Managed Instance running in their existing Kubernetes distribution. The teams install a set of code in their development environment, including an Azure Arc data controller that operates as the local control plane for any data management operations.
“In our on-premises Kubernetes cluster, we quickly started provisioning our first data controller node and created the first instance,” Patel reports. The teams also tested performance, comparing a workload running on SQL Server to Azure Arc–enabled SQL Managed Instance. Performance was comparable or better with the new approach.
Initially, the engineers looked to Microsoft for help. “That engagement was very helpful when we started, and we soon got the hang of it,” Patel notes. Diving deeper into the capabilities, the team began exploring how to add scale and connect to the existing continuous integration and continuous deployment (CI/CD) pipelines based on Azure DevOps and Azure Pipelines.
Wallin notes that they can run modern, cloud-native patterns within Kubernetes using the company’s existing hardware and facilities. “Azure Arc gives us modularity. We can deploy everything via infrastructure as code, much as we can in the cloud.”
That means developers at John Deere can meet the needs of manufacturing even faster. Using Azure Resource Manager templates, they can create databases on demand. “Now the product groups don’t need to go through a central IT team to provision resources,” Patel explains. “With Azure Arc, it’s truly self-service.”
Azure Arc SQL Managed Instance has an impact not only on the on-premises datacenter but also on the future use cases it enables as we deploy more and more technologies into our factories.
Unified management across environments
The modernization project gave John Deere a unified view of all the relational databases and data assets deployed with Azure Arc, even those in factories. Now the company can gradually phase out many of the monitoring and management tools it was using. Fewer software licenses translate into significant cost savings over time. The company can allocate its resources to more valuable tasks and make better use of staff time. Through Azure Arc, IT administrators also get logs and telemetry from Kubernetes, helping them to analyze the underlying infrastructure capacity and health.
“When you’re physically sending support people out to all the locations where we run SQL Server, it just doesn’t scale,” explains Patel. Azure Arc provides a single pane of glass—that is, one control plane that enables the company to manage its global environments efficiently.
Azure Arc also helps John Deere meet its high governance and compliance standards for apps, infrastructure, and data. Patel observes that it’s one thing to deploy policies consistently in a single datacenter—and another thing entirely in John Deere’s hybrid environment.
“With Azure Arc, we’ll be able to successfully deploy and manage our environments very effectively,” he adds.
Meanwhile, the IT administrators responsible for day-to-day management of the company’s SQL Server environments are getting a reprieve. “They’ll have more interesting things to do,” Wallin suggests, “and can move up the stack to work on automation, container management, and more exciting technologies.”
With Azure Arc, we’ll be able to successfully deploy and manage our environments effectively.
The smart industrial journey continues
The journey continues as John Deere teams continue to roll out Azure Arc–enabled data services. “Every factory is part of the company’s smart industrial focus by deploying operation technologies,” Patel offers. “We foresee the need to deploy local SQL Server instances, the compute, and the applications in these factories to support this initiative.”
As early adopters, John Deere provided valuable input to the Microsoft product team. “We gave feedback to Microsoft about the features that were important for us,” Patel explains.
The team’s advice for other companies? Shift to cloud-first thinking and infrastructure as code. “There was a learning curve for our infrastructure teams in moving to the mindset of the cloud,” Patel remembers, “but we overcame that. The teams now know the value of infrastructure as code and managing the environment effectively.”
Effective management includes the benefit of exchanging upfront IT capital expenditures (CapEx) for the more predictable operational expenditure (OpEx) model that comes with cloud computing services. “With a pay-as-you-go model, we expect to realize cost savings,” Patel reports.
Wallin concludes, “It certainly is a very, very bright future for John Deere. Given our goals—implementing our smart industrial redesign and modernizing our factory operations—we can’t keep doing business the way we’ve always done it. We’re using technology to successfully deploy and manage our environments very effectively.”
Now the product groups don’t need to go through a central IT team to provision resources…. With Azure Arc, it’s truly self-service.
Also Read
John Deere Partners With Microsoft To Bring New Value To Dealers
Farm Machinery Manufacturer Kubota Invests In France-Based AgTech Company