Merge pull request #100636 from openshift-cherrypick-robot/cherry-pick-99808-to-enterprise-4.20

[enterprise-4.20] OSDOCS#15472: Configuring TNA with agent-based installer

installing/installing_with_agent_based_installer/installing-with-agent-based-installer.adoc

@@ -78,6 +78,9 @@ include::modules/installing-ocp-agent-boot.adoc[leveloffset=+2]
 // Adding {ibm-z-name} agents with {op-system-base} KVM
 include::modules/installing-ocp-agent-ibm-z-kvm.adoc[leveloffset=+2]
 
+// Configuring a local arbiter node
+include::modules/installing-ocp-agent-local-arbiter-node.adoc[leveloffset=+2]
+
 // Verifying that the current installation host can pull release images
 include::modules/installing-ocp-agent-tui.adoc[leveloffset=+2]
 

modules/installation-configuration-parameters.adoc

@@ -707,6 +707,15 @@ ifdef::agent[]
 Supported values are `3`, `4`, `5`, or `1` when deploying {sno}.
 endif::agent[]
 
+|arbiter:
+    name: arbiter
+|The {product-title} cluster requires a name for arbiter nodes. For example, `arbiter`.
+
+
+|arbiter:
+    replicas: 1
+|The `replicas` parameter sets the number of arbiter nodes for the {product-title} cluster. You cannot set this field to a value that is greater than 1.
+
 |credentialsMode:
 |The Cloud Credential Operator (CCO) mode. If no mode is specified, the CCO dynamically tries to determine the capabilities of the provided credentials, with a preference for mint mode on the platforms where multiple modes are supported.
 

modules/installing-ocp-agent-local-arbiter-node.adoc

@@ -0,0 +1,93 @@
+// Module included in the following assemblies:
+//
+// * installing/installing_with_agent_based_installer/installing-with-agent-basic.adoc
+
+:_mod-docs-content-type: PROCEDURE
+[id="installing-ocp-agent-local-arbiter-node_{context}"]
+= Configuring a local arbiter node
+
+You can configure an {product-title} cluster with two control plane nodes and one local arbiter node so as to retain high availability (HA) while reducing infrastructure costs for your cluster.
+
+A local arbiter node is a lower-cost, co-located machine that participates in control plane quorum decisions. Unlike a standard control plane node, the arbiter node does not run the full set of control plane services. You can use this configuration to maintain HA in your cluster with only two fully provisioned control plane nodes instead of three.
+
+[IMPORTANT]
+====
+You can configure a local arbiter node only. Remote arbiter nodes are not supported.
+====
+
+To deploy a cluster with two control plane nodes and one local arbiter node, you must define the following nodes in the `install-config.yaml` file:
+
+* 2 control plane nodes
+* 1 arbiter node
+
+The arbiter node must meet the following minimum system requirements:
+
+* 2 vCPUs
+* 8 GB of RAM
+* 50 GB of SSD or equivalent storage
+
+The arbiter node must be located in a network environment with an end-to-end latency of less than 500 milliseconds, including disk I/O. In high-latency environments, you might need to apply the `etcd` slow profile.
+
+The control plane nodes must meet the following minimum system requirements:
+
+* 4 vCPUs
+* 16 GB of RAM
+* 120 GB of SSD or equivalent storage
+
+Additionally, the control plane nodes must also have enough storage for the workload.
+
+.Prerequisites
+
+* You have downloaded {oc-first} and the installation program.
+* You have logged into the {oc-first}.
+
+.Procedure
+
+. Edit the `install-config.yaml` file to define the arbiter node alongside control plane nodes.
++
+.Example `install-config.yaml` configuration for deploying an arbiter node
+[source,yaml]
+----
+apiVersion: v1
+baseDomain: devcluster.openshift.com
+compute:
+  - architecture: amd64
+    hyperthreading: Enabled
+    name: worker
+    platform: {}
+    replicas: 0
+arbiter: <1>
+  architecture: amd64
+  hyperthreading: Enabled
+  replicas: 1 <2>
+  name: arbiter <3>
+  platform:
+    baremetal: {}
+controlPlane: <4>
+  architecture: amd64
+  hyperthreading: Enabled
+  name: master
+  platform:
+    baremetal: {}
+  replicas: 2 <5>
+platform:
+  baremetal:
+# ...
+    hosts:
+      - name: cluster-master-0
+        role: master
+# ...
+      - name: cluster-master-1
+        role: master
+        ...
+      - name: cluster-arbiter-0
+        role: arbiter
+# ...
+----
+<1> Defines the arbiter machine pool. You must configure this field to deploy a cluster with an arbiter node.
+<2> Set the `replicas` field to `1` for the arbiter pool. You cannot set this field to a value that is greater than 1.
+<3> Specifies a name for the arbiter machine pool.
+<4> Defines the control plane machine pool.
+<5> When an arbiter pool is defined, two control plane replicas are valid.
+
+. Save the modified `install-config.yaml` file.