There are many different ways to deploy Kong Gateway. In this post, Viktor Gamov (Principal Developer Advocate at Kong) walks through the four most popular ways.
Depending on your particular use case, you may find that one or more of these is a good fit. Let's get started!
Using the DB-less and declarative configuration mode is one of the quickest and easiest ways to get started with Kong Gateway. For this mode, start with a Kong.yml file (which includes declarative configuration for Kong objects like services, plugins, and routes).
We can start Kong by providing this declarative configuration. One note here is that the Kong Admin API will only be available in Read Only mode. But if you want to make any changes to the configuration, you’ll need to submit the file to the /config endpoint with the new configuration.
Because this configuration doesn't rely on pulling data from anywhere or automatic updates, we call this DB-less mode. It's extremely simple and lightweight. If you're running the Kong Ingress Controller, this is a great option because KIC deploys and ships the declarative configuration through this config API.
If you're looking to scale your deployment and make incremental changes to your configuration, you'll need to use the Kong Admin API.
This requires introducing a database. The Kong Admin API will be available with configurations for services, plugins, and routes. We call this traditional mode.
This can be used with just a single cluster node or with multiple clusters — but all of them will derive their configuration from the database and the configuration established there. You’ll still need a load balancer in front of the database to assist with routing (this can also be Kong), but the configuration will be consistent.
With this deployment mode, you can also use Kong Manager and the Developer Portal with the database. But one caution here: you may run into a challenge where you’re spending a lot of effort maintaining the database, instead of on your applications. In order to optimize performance, Kong supports a cluster of Postgres, integration with read-only replicas, and doesn't call from the database each time.
If you’re still looking to solve the challenge of having too many calls from the database, you can deploy Kong in hybrid mode.
With this mode, different Kong nodes take on different Kong roles. This is done by creating a Kong instance that is explicitly serving the role of data plane (there can be multiples) with another instance that manages centralized configuration called the control plane. The data plane (or data planes) serves user traffic and is separated from the control plane.
In this mode, only the control plane talks to the database, and it isn't directly serving user traffic.
Kong Manager and Developer Portal are available in this mode as well and are part of the control plane, working directly with the database. So this deployment mode solves a lot of the scaling challenges you might face in traditional mode.
For organizations with many APIs looking to segregate deployment based on environment (for example, different environments for development, QA, regions, etc), the best deployment option is Kong Konnect.
In this deployment mode, we still have our Kong data planes, but the control plane is a cloud-managed control plane. This allows you to segregate and separate different environments using Runtime Manager. Kong Konnect also includes a Service Hub, where you can discover and publish your services — along with Developer Portal as a managed solution.
Kong Konnect also has built-in analytics for observing and monitoring your Kong instances. This allows you to have different Runtime Groups running in different environments (Docker, Kubernetes clusters, VMs, etc). With the managed cloud control plane in Kong Konnect, configuration is centralized and available to submit into different environments. This allows you to support GitOps or APIOps configurations. And with a SaaS control plane, the Kong Konnect team is able to quickly and frequently release new features, updates, and changes.
Get started quickly and easily by downloading Kong Gateway to launch in DB-less, traditional, or hybrid mode — or start a free trial of Kong Konnect to explore the cloud-managed control plane option.
Don’t forget to join the Kong Community to stay up to date on all things Kong — with access to Meetups, user calls, Tech Talks, and more!
Continued Learning & Related Content
Meet Emily. She’s an API product manager at ACME, Inc., an ecommerce company that runs on dozens of APIs. One morning, her team lead asks a simple question: “Who’s our top API consumer, and which of your APIs are causing the most issues right now?”
Ever feel like you're fighting an uphill battle with your API strategy? You're building APIs faster than ever, but somehow everything feels harder. Wasn’t API-first supposed to make all this easier? Well, you're not alone. And now industry analys
It’s been almost a year since we released our Konnect Terraform provider . In that time we’ve seen over 300,000 installs, have 1.7 times as many resources available, and have expanded the provider to include data sources to enable federated managem
In the last two parts of this series, we discussed How to Strengthen a ReAct AI Agent with Kong AI Gateway and How to Build a Single-LLM AI Agent with Kong AI Gateway and LangGraph . In this third and final part, we're going to evolve the AI Agen
In my previous post, we discussed how we can implement a basic AI Agent with Kong AI Gateway. In part two of this series, we're going to review LangGraph fundamentals, rewrite the AI Agent and explore how Kong AI Gateway can be used to protect an LLM
This is part one of a series exploring how Kong AI Gateway can be used in an AI Agent development with LangGraph. The series comprises three parts: Basic ReAct AI Agent with Kong AI Gateway Single LLM ReAct AI Agent with Kong AI Gateway and LangGr
What Is RAG, and Why Should You Use It? RAG (Retrieval-Augmented Generation) is not a new concept in AI, and unsurprisingly, when talking to companies, everyone seems to have their own interpretation of how to implement it. So, let’s start with a r