Update Install docs to show how to install EG where the infrastructure/cluster provider is installing Gateway API CRDs
I looked into this. The gateway api maintainers [do not provide an upstream helm chart](https://github.com/kubernetes-sigs/gateway-api/issues/1590) that we can source as a dependency.
Personally, I think that the envoy-gateway helm chart should leave the installation of these CRDs to the users: Just list it as a requirement, and link to their installation docs.
Due to this bundling of Gateway API CRDs and envoy-gateway CRDs, using the helm chart gets a bit fiddly for us when installing into our Google Kubernetes Engine clusters:
apiVersion: kustomize.config.k8s.io/v1beta1
kind: Kustomization
namespace: envoy-gateway-system
helmCharts:
- name: gateway-helm
releaseName: envoy-gateway
# The helm chart includes Gateway API, which we already install through a GKE Addon
includeCRDs: false
# GKE uses an older version of Gateway API than the latest helm chart
# https://gateway.envoyproxy.io/v1.0.0/install/matrix/
version: v0.5.0
repo: oci://registry-1.docker.io/envoyproxy
valuesFile: ./values.yaml
resources:
- https://raw.githubusercontent.com/envoyproxy/gateway/v0.5.0/charts/gateway-helm/crds/generated/config.gateway.envoyproxy.io_envoyproxies.yaml
- https://raw.githubusercontent.com/envoyproxy/gateway/v0.5.0/charts/gateway-helm/crds/generated/gateway.envoyproxy.io_authenticationfilters.yaml
- https://raw.githubusercontent.com/envoyproxy/gateway/v0.5.0/charts/gateway-helm/crds/generated/gateway.envoyproxy.io_envoypatchpolicies.yaml
- https://raw.githubusercontent.com/envoyproxy/gateway/v0.5.0/charts/gateway-helm/crds/generated/gateway.envoyproxy.io_ratelimitfilters.yaml
Originally posted by @djfinnoy in https://github.com/envoyproxy/gateway/issues/1721#issuecomment-2035558087
When i run the helm, I don't get the CRDS at all. Is that normal?
This issue has been automatically marked as stale because it has not had activity in the last 30 days.
I also noticed this, because I intent to run more than one gateway controller in my cluster.
This issue has been automatically marked as stale because it has not had activity in the last 30 days.
Still relevant.
This issue has been automatically marked as stale because it has not had activity in the last 30 days.
Can this get labeled with no stalebot?