GlobalConfiguration resource
This page explains how to use the GlobalConfiguration resource to define the global configuration parameters of F5 NGINX Ingress Controller.
The resource supports configuring listeners for TCP and UDP load balancing, and is implemented as a Custom resource.
Listeners are required by TransportServer resources and can be used to configure custom listeners for VirtualServers.
Prerequisites
When installing NGINX Ingress Controller using Manifests, you need to reference a GlobalConfiguration resource in the -global-configuration
command-line argument. NGINX Ingress Controller only needs one GlobalConfiguration resource.
GlobalConfiguration specification
The GlobalConfiguration resource defines the global configuration parameters of the Ingress Controller. Below is an example:
apiVersion: k8s.nginx.org/v1
kind: GlobalConfiguration
metadata:
name: nginx-configuration
namespace: nginx-ingress
spec:
listeners:
- name: dns-udp
port: 5353
protocol: UDP
- name: dns-tcp
port: 5353
protocol: TCP
- name: http-8083
port: 8083
protocol: HTTP
- name: https-8443
port: 8443
protocol: HTTP
ssl: true
Field | Description | Type | Required |
---|---|---|---|
listeners | A list of listeners. | listener | No |
Listener
The listeners:
key defines a listener (a combination of a protocol and a port) that NGINX will use to accept traffic for a TransportServer and a VirtualServer:
- name: dns-tcp
port: 5353
protocol: TCP
- name: http-8083
port: 8083
protocol: HTTP
Field | Description | Type | Required |
---|---|---|---|
name | The name of the listener. Must be a valid DNS label as defined in RFC 1035. For example, hello and listener-123 are valid. The name must be unique among all listeners. The name tls-passthrough is reserved for the built-in TLS Passthrough listener and cannot be used. |
string | Yes |
port | The port of the listener. The port must fall into the range 1..65535 with the following exceptions: 80 , 443 , the status port, the Prometheus metrics port. Among all listeners, only a single combination of a port-protocol is allowed. |
int | Yes |
protocol | The protocol of the listener. Supported values: TCP , UDP and HTTP . |
string | Yes |
ssl | Configures the listener with SSL. This is currently only supported for HTTP listeners. Default value is false |
bool | No |
ipv4 | Specifies the IPv4 address to listen on. | string | No |
ipv6 | Specifies the IPv6 address to listen on. | string | No |
Using GlobalConfiguration
You can use the usual kubectl
commands to work with a GlobalConfiguration resource.
For example, the following command creates a GlobalConfiguration resource defined in global-configuration.yaml
with the name nginx-configuration
:
kubectl apply -f global-configuration.yaml
globalconfiguration.k8s.nginx.org/nginx-configuration created
Assuming the namespace of the resource is nginx-ingress
, you can get the resource by running:
kubectl get globalconfiguration nginx-configuration -n nginx-ingress
NAME AGE
nginx-configuration 13s
With kubectl get
and similar commands, you can use the short name gc
instead of globalconfiguration
.
Validation
Two types of validation are available for the GlobalConfiguration resource:
- Structural validation by
kubectl
and Kubernetes API server. - Comprehensive validation by NGINX Ingress Controller.
Structural validation
The custom resource definition for the GlobalConfiguration includes structural OpenAPI schema which describes the type of every field of the resource.
If you try to create (or update) a resource that violates the structural schema (for example, you use a string value for the port field of a listener), kubectl
and Kubernetes API server will reject such a resource:
-
Example of
kubectl
validation:kubectl apply -f global-configuration.yaml
error: error validating "global-configuration.yaml": error validating data: ValidationError(GlobalConfiguration.spec.listeners[0].port): invalid type for org.nginx.k8s.v1.GlobalConfiguration.spec.listeners.port: got "string", expected "integer"; if you choose to ignore these errors, turn validation off with --validate=false
-
Example of Kubernetes API server validation:
kubectl apply -f global-configuration.yaml --validate=false
The GlobalConfiguration "nginx-configuration" is invalid: []: Invalid value: map[string]interface {}{ ... }: validation failure list: spec.listeners.port in body must be of type integer: "string"
If a resource is not rejected (it doesn’t violate the structural schema), NGINX Ingress Controller will validate it further.
Comprehensive validation
NGINX Ingress Controller validates the fields of a GlobalConfiguration resource. If a GlobalConfiguration resource is partially invalid, NGINX Ingress Controller use the valid listeners and emit events about invalid listeners.
You can check if the Ingress Controller successfully applied the configuration for a GlobalConfiguration. For our nginx-configuration
GlobalConfiguration, we can run:
kubectl describe gc nginx-configuration -n nginx-ingress
...
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal Updated 11s nginx-ingress-controller GlobalConfiguration nginx-ingress/nginx-configuration was updated
The events section includes a Normal event with the Updated reason that informs us that the configuration was successfully applied.
If you create a GlobalConfiguration nginx-configuration
with two or more listeners that have the same protocol UDP and port 53, you will get:
kubectl describe gc nginx-configuration -n nginx-ingress
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal Updated 55s nginx-ingress-controller GlobalConfiguration nginx-ingress/nginx-configuration was updated
Warning AddedOrUpdatedWithError 6s nginx-ingress-controller GlobalConfiguration nginx-ingress/nginx-configuration is invalid and was rejected: spec.listeners: Duplicate value: "Duplicated port/protocol combination 53/UDP"
The events section includes a Warning event with the AddedOrUpdatedWithError reason.
Using IPV4 and IPV6 Addresses with GlobalConfiguration
You can customize the IPv4 and IPv6 Address listeners in the global configuration and apply them to your VirtualServer resources. See the corresponding example here