From 550dd08e5241fd4f7834147ca2f4c3480ca38d9d Mon Sep 17 00:00:00 2001 From: Nitya Dhanushkodi Date: Fri, 12 Aug 2022 11:22:41 -0700 Subject: [PATCH] peering: update docs with exposing servers (#14191) --- website/content/docs/connect/cluster-peering/k8s.mdx | 12 ++++++++---- website/content/docs/k8s/helm.mdx | 4 +++- 2 files changed, 11 insertions(+), 5 deletions(-) diff --git a/website/content/docs/connect/cluster-peering/k8s.mdx b/website/content/docs/connect/cluster-peering/k8s.mdx index 4986a61b13..7471efed86 100644 --- a/website/content/docs/connect/cluster-peering/k8s.mdx +++ b/website/content/docs/connect/cluster-peering/k8s.mdx @@ -23,8 +23,7 @@ The following CRDs are used to create and manage a peering connection: You must implement the following requirements to create and use cluster peering connections with Kubernetes: - Consul version 1.13.1 or later - At least two Kubernetes clusters -- The Kubernetes clusters must be running in a flat network -- The network must be running on Consul on Kubernetes version 0.47.0 or later +- The installation must be running on Consul on Kubernetes version 0.47.1 or later ### Helm chart configuration @@ -48,14 +47,19 @@ To establish cluster peering through Kubernetes, deploy clusters with the follow -Install Consul on Kubernetes on each Kubernetes cluster by applying `values.yaml` using the Helm CLI. +Install Consul on Kubernetes on each Kubernetes cluster by applying `values.yaml` using the Helm CLI. With these values, +the servers in each cluster will be exposed over a Kubernetes Load balancer service. This service can be customized +using [`server.exposeService`](/docs/k8s/helm#v-server-exposeservice). When generating a peering token from one of the +clusters, the address(es) of the load balancer will be used in the peering token, so the peering stream will go through +the load balancer in front of the servers. For customizing the addresses used in the peering token, see +[`global.peering.tokenGeneration`](/docs/k8s/helm#v-global-peering-tokengeneration). ```shell-session $ export HELM_RELEASE_NAME=cluster-name ``` ```shell-session -$ helm install ${HELM_RELEASE_NAME} hashicorp/consul --version "0.47.0" --values values.yaml +$ helm install ${HELM_RELEASE_NAME} hashicorp/consul --version "0.47.1" --values values.yaml ``` ## Create a peering token diff --git a/website/content/docs/k8s/helm.mdx b/website/content/docs/k8s/helm.mdx index 0576bdb8c9..837a03f562 100644 --- a/website/content/docs/k8s/helm.mdx +++ b/website/content/docs/k8s/helm.mdx @@ -73,7 +73,9 @@ Use these links to navigate to a particular top-level stanza. - `source` ((#v-global-peering-tokengeneration-serveraddresses-source)) (`string: ""`) - Source can be set to "","consul" or "static". "" is the default source. If servers are enabled, it will check if `server.exposeService` is enabled, and read - the addresses from that service to use as the peering token server addresses. + the addresses from that service to use as the peering token server addresses. If using admin partitions and + only Consul client agents are enabled, the addresses in `externalServers.hosts` and `externalServers.grpcPort` + will be used. "consul" will use the Consul advertise addresses in the peering token.