Four challenges of using multiple API gateways: Webinar takeaways

IT architectures are becoming more complex with multiple API gateways. And without central visibility and control over your APIs, multiple gateways can mean various problems.

Axway conducted customer and market research, to understand why IT leaders are adopting multiple API gateways, and what the most common challenges and risks are.

(We summarized our findings in the white paper called API gateways: The more, the merrier, and of course, shared what we learned with our Product Design team.)

In the latest webinar hosted by Axway’s API Management Platform experts, Erik Wilde and Jana Frejova, we took a deeper dive into the four most common challenges that arise once you’ve adopted more than one gateway.

This slideshow requires JavaScript.

Four challenges of multiple API gateways

Here are the four top challenges you will need to overcome based on our research with companies that have multiple API gateways.

1. Governance and security: Policies and how they are set up, are different across different API gateway vendors (even across diverse deployment models from the same vendor), making it difficult to apply them consistently.
2. Consumption visibility: Due to each gateway being normally tied to a single portal, having multiple API gateways creates a siloed experience from the governance side to the consumption side. Consumers don’t have a single place to discover effortlessly and access all APIs and producers lack centralized visibility into who is consuming their APIs.
3. Economic trade-offs: A decentralized architecture with multiple API gateways can have many monetary implications: it can be plain expensive, lead to duplication of effort, and deter the reuse of your APIs. Just as with modern IT architectures overall, the goal should be to stop building systems and start cultivating ecosystems. In other words, centralize and decentralize as much as possible. The main emphasis should be on delivering more value to the business with APIs.
4. Debugging and monitoring: Having automation in place to watch API consumption and behavior is a regular practice to ensure security and reliability.

This becomes more demanding when tracing API consumption across various gateways that aren’t governed by the same API management solution.

The limits of multiple API gateways without centralized visibility or governance are highlighted when there’s a bug. It becomes more cumbersome to trace the root cause and solve the issue.

 The magic pattern is to design them in a way to play well in a decentralized environment. It goes to API Portals, security, and governance, etc. 

With the “magic pattern,” you can manage the difficulties with better governance… and a better outcome.

The main takeaway is centralized visibility is the way to go. You have to anticipate and build for it.