May 2024 Summaries

ngrok's API gateway provides a flexible and programmable way for developers to manage their APIs' security, reliability, and performance. Unlike traditional deployed or cloud-based API gateways, ngrok's developer-defined option is feature-rich, works everywhere, and allows self-serving production without operational headaches or costs. The API gateway can be configured using two methods: at the Edge level with a web-based editor or directly with the ngrok agent as an ngrok endpoint. This flexibility enables developers to apply API policies in real-time, making it possible to test new drop-in rate limiting policies quickly and easily. Various examples of API policy management are provided, including blocking traffic from specific countries, maintaining and deprecating API versions, manipulating headers on inbound requests, adding compression to responses, enforcing the TLS version of requests, logging unsuccessful events to observability platforms, and limiting request sizes.

The Kubernetes Gateway API is a standardized way to configure and manage network traffic routing within Kubernetes clusters, offering a unified configuration for various protocols under one roof, eliminating the need for separate tools for each protocol. It provides advanced routing capabilities, an extensible architecture, role-based access control, and simplifies traffic management, empowering developers to focus on application logic while reducing operational complexity. The Gateway API is designed to be extensible, allowing integration with future functionalities and third-party plugins, and it's a community-driven standard within the Kubernetes ecosystem, promoting standardization and future-proofing. However, its adoption rate is still growing, and some challenges exist regarding its relative newness and learning curve for those accustomed to ingress controllers.

Secure site-to-site connectivity with ngrok enables organizations to access third-party APIs, databases, and IoT service API endpoints running at external sites securely, without requiring customers to open inbound ports on their firewalls. This unlocks a wide range of use cases and empowers companies to offer unique capabilities to their customers. Site-to-site connectivity can be achieved through ngrok's comprehensive guides, which provide step-by-step instructions with code examples for implementing secure access to APIs, databases, device APIs, and other resources at external sites. The guides cover various scenarios, including mTLS encryption, and empower users to whitelist the address used by the ngrok agent to connect to the ngrok service, as well as connect to the agent running in the customer's network using the agent API.

The text discusses a project aimed at simplifying the deployment and security of APIs, particularly those using large language models (LLMs), by leveraging ngrok's new developer-defined API Gateway. The author explores the potential of using AI to facilitate gardening tasks, such as planning a vegetable garden, and outlines a process for deploying an open-source LLM on Azure using SkyPilot, a framework for managing AI workloads across cloud platforms. The guide emphasizes the importance of implementing robust authentication to protect APIs from unauthorized access and highlights the challenges of traditional authentication methods. By utilizing ngrok's API Gateway and JSON Web Tokens (JWTs), developers can secure their APIs efficiently, ensuring data protection and reducing the complexity and costs associated with traditional solutions. This streamlined approach allows rapid innovation and integration of secure APIs into applications, enabling developers to focus on building advanced features like AI-powered gardening insights.

The text compares the developer experience (DX) of using Kong Gateway and ngrok's API gateway for deploying JWT-secured APIs, emphasizing the importance of DX in API design and deployment. Both gateways aim to simplify the process of setting up API authentication and routing, but they differ in approach. Kong Gateway, based on Nginx, requires involvement from DevOps or IT teams for infrastructure deployment and involves using curl commands and the Admin API to configure services and routes. In contrast, ngrok's API gateway offers a more streamlined process with built-in JWT validation, simplified provisioning, and configuration using YAML and CEL expressions, eliminating the need for complex networking setups and allowing developers more control over operations. ngrok's approach abstracts infrastructure complexity, providing DDoS protection and global load balancing at the edge, making it easier and quicker for developers to deploy and manage APIs across different environments while maintaining security through integration with Auth0 for token management.

The new API introduced by ngrok allows customers to automate the creation and management of Bot Users in their account. Bot Users are useful for organizing credentials that stick around after a user leaves an organization, and can be used to create onboarding scripts and bootstrapping processes without manually creating them beforehand. The API provides endpoints for creating, updating, and deleting Bot Users, as well as leveraging scripting languages to automate these tasks. This update aims to help customers manage credentials more efficiently, making it easier to create unique credentials for each 3rd party integration or customer using their ngrok account.