You are browsing unreleased documentation. See the latest documentation here.
Concepts
In this page we will introduce concepts that are core to understanding Kong Mesh.
Control Plane
The control plane is the central management layer of Kong Mesh. It is responsible for configuring and managing the behavior of the data plane, which handles the actual traffic between services.
Data Plane
The data plane handles traffic between services. In practice these are the apps that you build and that you want to put inside you service-mesh.
Data Plane Proxy / Sidecar
The data plane proxy or sidecar is the instance of Envoy running alongside the application which will send and receive traffic from the rest of the service mesh. It connects to the control plane which computes a configuration specific to it.
flowchart LR clients servers subgraph data plane app subgraph data plane proxy/Envoy inbounds outbounds end end inbounds -.local traffic.-> app app -.local traffic.-> outbounds clients --> inbounds outbounds --> servers
Inbound
An inbound is the part of the data plane proxy which receives traffic for a specific port. Inbounds are usually grouped between different data planes and form a service.
Outbound
An outbound is the part of the data plane proxy which sends traffic for a specific service. Outbounds group multiple remote inbounds as endpoints.
Resource
A resource is an object or entity that can be created, managed, and interacted with in Kong Mesh. Resources are the building blocks that define the behavior and state of your service mesh. Each resource is defined as a type of API object that has a specific purpose and is represented by its state and configuration.
A resource is most often expressed as yaml and can have 2 formats:
-
Kubernetes
when the backing control plane runs on Kubernetes. In this case Kong Mesh resources are defined as Kubernetes Custom Resource Definitions. -
Universal
in other cases or when accessing resources through Kong Mesh’s REST api.
Resources are most commonly represented in yaml format.
Policy
Policies are a specific type of resources that controls the behaviour and communication of applications running inside your service mesh. They can enable traffic management, security, observability and traffic reliability.
Policies always have a clear specific area of impact and goal. To learn more about policies checkout the in depth introduction.