Hub and Spoke: A Better Way to Architect Your Tech Stack
Picture a bicycle wheel, with its central hub and metal spokes. The spokes transfer the load between the hub and the rim of the wheel. As the hub unites all the individual spokes, energy is transferred evenly from pedaling and braking motions to the wheel, to generate and control movement. It’s a brilliant, yet simple design — one that can be applied to many use cases.
In networking, the hub-and-spoke model simplifies the process of hosting services that are used by multiple workloads. In retail fulfillment, it can be used to shorten the time in transit between fulfillment centers and customers. And in the modern enterprise, the hub-and-spoke model can be applied to design your data architecture, removing the limitations of legacy, monolithic software platforms while providing more flexibility and coverage than microservices.
In this article, I’ll explain how organizations can use the hub-and-spoke model to build out a highly efficient software stack, improving agility while simplifying operations and reducing costs.
The Drawbacks of Traditional Approaches
In traditional monolithic software architectures, there are large, complex software platforms, such as CMS, ERP or CRM systems. Together, these systems may power the entire company — or most of it — and because so much time and money is invested in maintaining them, they’re nearly impossible to replace. Essentially, you’re stuck with it, so when a business unit or team needs new or different functionality, they are bound by the capabilities of the platform. They have to use what’s there, rather than adopting a more specialized tool that provides the exact capabilities needed.
Such vendor lock-in can drive up costs and make migrating data impossible, but that’s not the only disadvantage. Upgrades are time-consuming and expensive because all of the code is tightly woven together. You become dependent on the vendor and the functionality they offer, and you have to adapt your workflows to what’s possible in the platform.
To combat vendor lock-in and the costs and limitations of the monolithic approach, many organizations have adopted a microservices model, in which applications are developed as a collection of loosely connected services, each with a distinct focus. Microservices increase agility, because you can develop and deploy each service independently of the others, without incurring downtime or needing to refactor other parts of the application. Plus, you get the right tool with the right capabilities for the job at hand.
However, there are drawbacks to the microservices model: If one of the services advances and another stagnates or is no longer supported, the integrations and dependencies between them may break. One dependency breaking can have a domino effect, bringing operations to a halt. What’s more, services are directly connected, often through one-click integrations that can’t be customized, which means it’s often impossible to insert custom logic between services to fill any gaps in capabilities.
It comes down to complexity — you’re moving from a simple monolithic model to an overly complex model with hundreds or even thousands of interdependencies.
Finding Your Balance with Hub-and-Spoke
The hub-and-spoke approach to architecting the technology stack combines the stability of a monolithic platform with the agility of microservices so that you can benefit from new capabilities and software innovation, without the burden of complex integrations and dependencies slowing you down. In hub-and-spoke software architecture, the “hub” is the foundational data layer — a single point of data access that can provide about 80% of the functionality you need. The key is that the hub “plays well with others.”
Unlike a bloated behemoth application, the hub is open and designed to connect with other tools and applications through flexible APIs, SDKs, and webhooks. In this way, you can bring specialized tools together to best meet the unique needs of your business, while automating data workflows across your ecosystem.
The hub serves as the baseline of common and/or critical functionality, to which you can easily and cost-effectively connect other business-critical systems such as Stripe, Hubspot, or Salesforce, as well as hyper-specialized tools, as needed. While removing the complexity of managing multiple base services, it eliminates gaps in functionality, since it’s easy to connect a new tool or capability without altering the stack or interfering with operations.
For example, Directus is an open data platform that applies a hub-and-spoke model to help organizations deploy the capabilities of numerous disparate solutions with a single platform. Whereas custom APIs and CMS require months of development work and ongoing maintenance, Directus installs in minutes, connecting instantly to your existing database and other business-critical systems via its dynamically generated APIs. Plus, it enables technical and non-technical users to access and manipulate data safely via a no-code Data Studio app.
Because Directus leverages the hub-and-spoke model, it can be used to replace multiple disparate systems and address multiple use cases for the data. For example, when you consider a microservice architecture, you may have one app for inventory, another for invoicing, and yet another for commerce. Each of those individual services requires a separate database and license, and you have to interconnect them using their disparate APIs. Using the hub-and-spoke model, Directus consolidates all of that data and makes it accessible through a single API, powering all three apps. The services — the spokes — are powered by Directus — the hub — minimizing licensing costs, complexity and administration overhead.
A real-world example can be found in the travel industry. A large airline carrier recently implemented Directus to consolidate the capabilities of multiple systems, including reservations, inventory, departure control, a mobile app, airport kiosks, and internal administrative tools. By powering these systems with a single source of data, they cut costs while eliminating the potential for data discrepancies between systems.
Eliminate Roadblocks with a Common Data Backbone
Organizations that adopt a hub-and-spoke software architecture will realize numerous benefits:
- A complete data ecosystem without any gaps in capabilities
- Reduced complexity and lower operational overhead
- Reduced licensing and training costs from un-needed base services
- The ability to add or remove tools as needed, without interrupting business operations
- Maximum flexibility and agility as the needs of the business evolve
The hub-and-spoke model eliminates roadblocks to software development and deployment, reducing the burden on engineers and developers. It simplifies the process of adding purpose-built applications to your stack by leveraging a centralized, common data backbone. This enables you to reallocate time and resources from low-value but necessary activities to higher-value, revenue-generating activities that drive the business forward.
Join the discussion in our Discord community to see how developers are using Directus to build internal tools to connect, manage and visualize data.