Update README.md - add section about two or more replicas with traefik
Pull Request
Description of the change
It was explained how to create a traefik ingressRoute Resource that uses sticky sessions
Benefits
Users can login when more nextcloud replicas are used
Possible drawbacks
I don't see any
Applicable issues
when more replicas
awesome, thanks for this! I've made the tiniest suggestion and then this can merged :)
Maybe we should use annotations and write docu for multiple ingress-controller:
- here for traefik: https://doc.traefik.io/traefik/routing/providers/kubernetes-ingress/#on-service
- here for nginx: https://kubernetes.github.io/ingress-nginx/user-guide/nginx-configuration/annotations/#session-affinity
- here for haproxy: https://haproxy-ingress.github.io/docs/configuration/keys/#affinity
Your docu is very traefik specifiy and would be maybe replaced with gateway-api and standard over multiple ingress-controller: https://gateway-api.sigs.k8s.io/implementations/ Current state for session policy: https://gateway-api.sigs.k8s.io/geps/gep-1619/
@wrenix I feel like having a full ingress section of the docs would be nice, and we can certainly approach that later if you'd like, but I think getting this merged is still a good idea in the meantime. In particular, I've seen a lot of hype around the new gateway api (I'm also hyped!), but haven't actually played with it yet. I'd be down for some docs on it as well.
Sorry, i believe you have miss understand me. The Idea is, that we should minimize the documentation of not using CustomResources from ingress-controller (ingress-operator in this case) if the ingress-controller could work also with annotations on the ingress-manifest.
The Idea/Reason: An annotation is just a value which has to be set in that helm-chart that is much easier then to deploy an additional manifest (or to put another/not-necessary manifest in that helm-chart. But when the Gateway API is more distributed (as a new standard) then i like the idea to put a "not-necessary Manifest" in a helm-chart. <- this small unimportant sentence was to much present i my last comment)