OSDOCS-13287: adds RHOAI to MicroShift

Author: ShaunaDiaz

_attributes/attributes-microshift.adoc

@@ -6,6 +6,8 @@
 :OCP: OpenShift Container Platform
 :ocp-version: 4.18
 :oc-first: pass:quotes[OpenShift CLI (`oc`)]
+:rhoai-full: Red{nbsp}Hat OpenShift AI
+:rhoai: RHOAI
 //OpenShift Kubernetes Engine
 :oke: OpenShift Kubernetes Engine
 :product-title-first: Red Hat build of MicroShift (MicroShift)

_topic_maps/_topic_map_ms.yml

@@ -99,19 +99,6 @@ Topics:
 - Name: Listing update package contents
   File: microshift-list-update-contents
 ---
-Name: Support
-Dir: microshift_support
-Distros: microshift
-Topics:
-- Name: The etcd service
-  File: microshift-etcd
-- Name: The sos report tool
-  File: microshift-sos-report
-- Name: Getting your cluster ID
-  File: microshift-getting-cluster-id
-- Name: Getting support
-  File: microshift-getting-support
----
 Name: Configuring
 Dir: microshift_configuring
 Distros: microshift
@@ -198,6 +185,13 @@ Topics:
 - Name: Understanding storage migration
   File: microshift-storage-migration
 ---
+Name: Using AI models
+Dir: microshift_ai
+Distros: microshift
+Topics:
+- Name: Using artificial intelligence with MicroShift
+  File: microshift-rhoai
+---
 Name: Running applications
 Dir: microshift_running_apps
 Distros: microshift
@@ -239,6 +233,19 @@ Topics:
 - Name: Automated recovery from manual backups
   File: microshift-auto-recover-manual-backup
 ---
+Name: Support
+Dir: microshift_support
+Distros: microshift
+Topics:
+- Name: The etcd service
+  File: microshift-etcd
+- Name: The sos report tool
+  File: microshift-sos-report
+- Name: Getting your cluster ID
+  File: microshift-getting-cluster-id
+- Name: Getting support
+  File: microshift-getting-support
+---
 Name: Troubleshooting
 Dir: microshift_troubleshooting
 Distros: microshift

microshift_ai/_attributes

@@ -0,0 +1 @@
+../_attributes/

microshift_ai/images

@@ -0,0 +1 @@
+../images/

microshift_ai/microshift-rhoai.adoc

@@ -0,0 +1,16 @@
+:_mod-docs-content-type: ASSEMBLY
+[id="microshift-rh-openshift-ai"]
+include::_attributes/attributes-microshift.adoc[]
+= Using Red Hat OpenShift AI with {microshift-short}
+:context: microshift-rh-openshift-ai
+
+toc::[]
+
+Learn how to serve artificial intelligence (AI) models on your {microshift-short} edge deployments in a lightweight manner.
+
+include::modules/microshift-rhoai-con.adoc[leveloffset=+1]
+
+include::modules/microshift-rhoai-install.adoc[leveloffset=+1]
+//hmm, do we want to include again in the install books?
+
+

microshift_ai/modules

@@ -0,0 +1 @@
+../modules/

microshift_ai/snippets

@@ -0,0 +1 @@
+../snippets/

modules/microshift-rhoai-con.adoc

@@ -0,0 +1,12 @@
+// Module included in the following assemblies:
+//
+// * microshift_ai/microshift-rhoai.adoc
+
+:_mod-docs-content-type: CONCEPT
+[id="microshift-rhoai-con_{context}"]
+= How {rhoai-full} works in {microshift-short}
+
+You can train an artificial intelligence (AI) model in the cloud or datacenter on {OCP}, then run your model in your edge deployments on {microshift-short}.
+
+{rhoai-full} ({rhoia}) is a platform for data scientists and developers of AI and machine learning (AI/ML) applications. {rhoia} provides an environment to develop, train, serve, test, and monitor AI/ML models and applications on-premise or in the cloud.
+

modules/microshift-rhoai-install.adoc

@@ -0,0 +1,54 @@
+// Module included in the following assemblies:
+//
+// * microshift_ai/microshift-rhoai.adoc
+
+:_mod-docs-content-type: CONCEPT
+[id="microshift-rhoai-install_{context}"]
+= Installing the {rhoai-full} RPM
+
+To use AI models in {microshift-short} deployments, use this procedure to install the {rhoai-full} {{rhoai}} RPM a new {microshift-short} installation. You can install the RPM on an existing {microshift} instance as long as you restart the system.
+
+.Prerequisites
+
+* The system requirements for installing {microshift-short} have been met.
+* You have root user access to your machine.
+* You configured your LVM VG with the capacity needed for the PVs of your workload.
+//Do we need to config the CSI?
+* You added memory as required for your AI model.
+//I assume there is some kind of memory requirement for using an AI model?
+
+
+. Install the {microshift-short} AI-model-serving RPM package by running the following command:
++
+[source,terminal]
+----
+$ sudo dnf install microshift-microshift-ai-model-serving
+----
++
+[TIP]
+====
+If you create your manifests and custom resources (CRs) while you are completing your installation of {microshift-short}, you can avoid restarting the service a second time to apply them.
+====
+
+. As a root user, restart the {microshift-short} service by entering the following command:
++
+[source,terminal]
+----
+$ sudo systemctl restart microshift
+----
+
+//Q: Can we check the namespace to verify?
+//Kserve manifest and ServingRuntimes CRs are deployed in a `redhat-ods-applications` namespace.
+
+.Next steps
+
+. Configure your hardware, the operating system, and additional components to make use of their accelerators.
+
+. Create manifests that include
+* `ServingRuntime` in your namespace
+* `InferenceService` referencing `ServingRuntime` and a model
+* `Route` CR
+
+// Kserve creates Deployment and other resources.
+
+. Resources from previous step become ready and user can make HTTP/GRPC calls to the model server.