Or, even if you are only exposing APIs and data to internal consumers, perhaps that data contains sensitive PII and additional security mechanisms are still necessary.

Perhaps you use third-party monitoring or cybersecurity tools and need to stream data from your own systems to that software. Maybe you’re a manufacturer who wants to stream data to regulatory authorities so they can monitor your operations remotely, for example. There are plenty of circumstances where you might want to expose a data stream externally and require the security and reliability protections that next-gen API management tools deliver. The most basic use cases for data streaming platforms like Kafka involve streaming data internally – from one set of applications to another within the same IT environment, for example.īut this is certainly not always the case. Thinking Beyond Internal APIsĪt this point, you may be thinking, “We don’t expose data streams externally, so why would we need a next-gen API management solution to support them?” That means that you could have, for example, a Kafka stream on the backend that is exposed to clients using HTTP – effectively allowing apps that can only make requests via HTTP to connect to Kafka data streams, without having to build support for protocol translation into the applications themselves. In addition, next-gen API management tools offer the ability to translate between different front and backend protocols. They operate as gateways that can not only mediate between traditional REST-based client and API endpoints, but also support event brokers, asynchronous APIs, and even apply policies at the message level to support message-based communication. Next-generation API management solutions fill this gap. Nor do they typically support protocols other than HTTP, which is a problem because event brokers like Kafka use their own protocols, not HTTP, to enable communication between servers and clients. Traditional API gateways aren’t very compatible with event brokers because the gateways are not designed to manage real-time data streams. Or, they might need to stream sales transaction data to a fraud detection engine so it can catch fraudulent transactions in real-time.įor use cases like these, legacy API management doesn’t work well at all. They might use an open source event broker like Kafka to stream monitoring data from across their IT assets to the tools that IT engineers use to observe and analyze systems, for example.

Instead, businesses want to be able to stream data in real-time to multiple types of applications.

Increasingly, however, not all data is shared synchronously, and not all applications can consume data using HTTP requests. That’s why the standard approach to API management is to deploy a gateway, configure some load balancing and security policies and call it a day. Traditionally, these conditions were true for most API use cases at most organizations. All of your clients can consume data from APIs using HTTP requests. All of your APIs are synchronous, which doesn’t always work well with continuous data streams.Ģ. That approach works well if two conditions are true:ġ.

Critical role backblaze offer code