5) Single-Site Overcloud Deployment Concepts

What is Overcloud?

In RHOSP, OpenStack is deployed using a two-layer architecture:

 Undercloud – Deployment and management environment

 Overcloud – The actual production OpenStack cloud

Undercloud

 Runs on a single machine

 Uses OpenStack TripleO tools

 Deploys and manages Overcloud nodes

Overcloud

 Production environment

 Multiple node types (Controller, Compute, Storage)

 High Availability capable

Deployment Workflow (Single-Site)

1. Install Undercloud

2. Register nodes (bare metal)

3. Assign roles (Controller/Compute)

4. Deploy Overcloud

5. Validate deployment

Key Concepts

 Bare metal provisioning

 Heat templates

 Node roles

 High Availability

 Pacemaker clustering

6) Launching Instances in OpenStack

What is an Instance?

An instance is a virtual machine created and managed by OpenStack Nova.

1. Image (from Glance)

2. Flavor (defines CPU, RAM, Disk)

3. Network

4. Key pair (for SSH access)

5. Security group

Steps to Launch an Instance (Horizon Dashboard)

1. Log in to Horizon

2. Go to Project → Compute → Instances

3. Click Launch Instance

4. Select:

o Source (Image/Volume)

o Flavor

o Network

o Security group

5. Launch

Basic Resource Overview

1. Flavors

Define:

 vCPUs

 RAM

 Root disk size

Example:

 [Link] → 1 vCPU, 2GB RAM

2. Images

Prebuilt OS templates (e.g., RHEL, Ubuntu, CirrOS)

3. Networks
  Provider networks

 Self-service (tenant) networks

 Subnets

 Routers

 Floating IPs

4. Volumes

Persistent block storage attached to instances.

1. Cloud Computing Fundamentals

1.1 Definition of Cloud Computing

Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of
configurable computing resources (servers, storage, networks, applications, services) that can be
rapidly provisioned and released with minimal management effort.

Essential Characteristics

1. On-demand self-service
Users provision resources automatically without human interaction.

2. Broad network access

3. Resource pooling (Multi-tenancy)

4. Rapid elasticity
Resources scale up/down automatically.

5. Measured service
Usage is monitored and billed based on consumption.

1.2 Cloud Service Models

IaaS (Infrastructure as a Service)

 Virtual machines

 Virtual networking

 Storage

 Example: OpenStack, Amazon EC2

PaaS (Platform as a Service)

 Runtime environment

 Middleware

 Developer tools

User deploys applications without managing infrastructure.

SaaS (Software as a Service)

 Complete application

 No infrastructure management

 Example: Microsoft 365

1.3 Cloud Deployment Models

 Public Cloud – Shared infrastructure

 Private Cloud – Dedicated to one organization

 Hybrid Cloud – Combination of public and private

 Community Cloud – Shared among specific organizations

2. Introduction to OpenStack Platform

2.1 What is OpenStack?

OpenStack is an open-source Infrastructure-as-a-Service (IaaS) platform that manages compute,

It provides:

 REST APIs

 Web dashboard (Horizon)

 CLI tools

It is modular, meaning different services handle different responsibilities.

2.2 OpenStack Architecture (Conceptual View)

OpenStack consists of:

 Controller Services
  Compute Services

 Networking Services

 Storage Services

 Database & Message Queue backend

Communication between services happens using:

 REST APIs

 Message Queue (RabbitMQ)

 Database (MariaDB)

3. OpenStack Personas

3.1 Cloud Administrator

 Deploys and upgrades cloud

 Configures networking and storage

 Manages hypervisors

 Sets global quotas

3.2 Cloud Operator

 Monitors health of services

 Troubleshoots VM issues

 Manages capacity

3.3 Project (Tenant) Administrator

 Manages users inside project

 Assigns roles

 Controls project-level quotas

3.4 End User

 Launches instances

 Creates networks

 Attaches volumes

 Manages SSH keys

4.1 What is RHOSP?

Red Hat OpenStack Platform is an enterprise distribution of OpenStack provided by Red Hat.

It includes:

 Tested packages

 Enterprise lifecycle support

 Security hardening

 Certified hardware support

4.2 RHOSP Architecture

RHOSP uses a two-layer deployment model:

1. Undercloud

 Deployment management environment

 Runs TripleO tools

 Provisions bare metal nodes

2. Overcloud

 Production OpenStack cloud

 Hosts workloads

 Includes Controller and Compute nodes

4.3 Node Types in Overcloud

Controller Node

Runs:

 API services

 Scheduler

 Database

 Message queue

 Identity service

High availability often implemented.

Compute Node
Runs:

 Hypervisor (KVM)

 Nova compute service

 Hosts virtual machines

Storage Node

Runs:

 Cinder backend

 Swift or Ceph (optional)

5. Core OpenStack Services (Detailed)

5.1 Keystone (Identity Service)

Function:

 Authentication

 Authorization

 Service catalog

Key Concepts:

 User

 Role

 Project

 Token

5.2 Nova (Compute Service)

Responsible for:

 VM lifecycle

 Scheduling

 Hypervisor management

Nova components:

 nova-api

 nova-scheduler
  nova-compute

 nova-conductor

5.3 Neutron (Networking Service)

Provides:

 L2 and L3 networking

 DHCP

 Routing

 Floating IP

 Security groups

Supports:

 VLAN

 VXLAN

 GRE

5.4 Glance (Image Service)

Stores and retrieves:

 VM disk images

 Snapshots

Supports backends:

 File system

 Object storage

 Ceph

5.5 Cinder (Block Storage)

Provides:

 Persistent volumes

 Attach/detach volumes

 Snapshots

Used for databases and stateful workloads.

5.6 Horizon (Dashboard)

 Instances

 Networks

 Volumes

 Users

5.7 Heat (Orchestration)

Template-based deployments using:

 HOT (Heat Orchestration Template)

Allows automated infrastructure creation.

6. Single-Site Overcloud Deployment Concepts

6.1 Deployment Phases

1. Install Undercloud

2. Register bare metal nodes

3. Introspect hardware

4. Assign roles

5. Deploy Overcloud

6. Validate deployment

6.2 High Availability Concepts

 Multiple controllers

 Pacemaker clustering

 HAProxy for load balancing

 Database replication

6.3 Networking in Deployment

Networks used:

 Provisioning network

 Internal API network

 Storage network
  Tenant network

 External network

Network isolation improves:

 Performance

 Security

 Fault tolerance