Mastering OpenStack电子书

Mastering OpenStack

Table of Contents

Mastering OpenStack

Credits

About the Author

About the Reviewers

www.PacktPub.com

Support files, eBooks, discount offers, and more

Why subscribe?

Free access for Packt account holders

Preface

What this book covers

What you need for this book

Who this book is for

Conventions

Reader feedback

Customer support

Downloading the example code

Downloading the color images of this book

Errata

Piracy

Questions

1. Designing OpenStack Cloud Architecture

OpenStack – think again

Introducing the OpenStack logical architecture

Keystone

Swift

Glance

Cinder

Nova

nova-api

nova-compute

nova-volume

nova-network

nova-scheduler

Queue

Database

Neutron

The Neutron architecture

Horizon

Gathering the pieces and building a picture

Provisioning a flow under the hood

Expanding the picture

A sample architecture setup

Deployment

The conceptual model design

The logical model design

What about storage?

Networking

The logical networking design

Network layout

The external network

The storage network

The management network

The internal VM traffic

The physical model design

Estimating your hardware capabilities


CPU calculations

Memory calculations

The network calculations

Storage calculations

Best practices

Summary

2. Deploying OpenStack – DevOps and OpenStack Dual Deal

DevOps in a nutshell

DevOps and cloud – everyone is coding

DevOpsing OpenStack

Breaking down the OpenStack pieces

Making the infrastructure deployment professional

Bringing OpenStack to the chain

Continuous integration and delivery

Eat the elephant

Preparing the infrastructure code environment

The Chef environment

Prerequisites for settings

The Chef server installation

Workstation installation

Time to cook OpenStack

Where is my kitchen?

OpenStack cookbooks

Resolving OpenStack cookbook dependencies

Uploading cookbooks using Berkshelf

Defining roles

Configuring the environment

Summary

3. Learning OpenStack Clustering – Cloud Controllers and Compute Nodes

Understanding the art of clustering

Asymmetric clustering

Symmetric clustering

Divide and conquer

The cloud controller

nova-conductor

nova-scheduler

X-api

Image management

Network outfit

The Horizon decision

Planning for the message queue

Consolidating the database

Cloud controller clustering

Cooking the cloud controller

The compute node

Overcommitment considerations

The CPU allocation ratio

The RAM allocation ratio

Deciding on the hypervisor

Changing the color of the hypervisor

Storing instances' alternatives

External shared file storage

Internal nonshared file storage

Cooking the compute node

Preparing for plan B

Back up with backup-manager

Simple recovery steps

Summary

4. Learning OpenStack Storage – Deploying the Hybrid Storage Model

Understanding the storage types

Ephemeral storage

Persistent storage

Object storage is not NAS/SAN

A spotlight on Swift

The Swift architecture

Fire and forget

Indexing the data

A rich API access

Physical design considerations

Where is my data?

Swift hardware

Where to place what

The Swift network

Cooking Swift

Joining Cinder

Choosing the storage

CAP under scope

Stirring up the storage

Cinder can do more

The Cinder use case

Beyond Cinder – Ceph

Ceph in OpenStack

Cooking Ceph

Storing images in Ceph

Summary

5. Implementing OpenStack Networking and Security

The story of an API

Security groups

Managing the security groups using Horizon

Managing the security groups using the Neutron CLI

Managing the security groups using the Nova CLI

An example of a web server DMZ

Firewall as a Service

Coupling a firewall with Neutron

The Neutron plugin

There can be more than one plugin

Empowering the traffic isolation

VPN as a Service – a case study

General settings

VPNaaS configuration

Creating the Internet Key Exchange policy

Creating an IPSec policy

Creating a VPN service

Creating an IPSec site connection

Summary

6. OpenStack HA and Failover

HA under the scope

Do not mix them

HA levels in OpenStack

A strict service-level agreement

Measuring HA

The HA dictionary

Hands on HA

Understanding HAProxy

Services should not fail

The load balancer should not fail

OpenStack HA under the hood

HA the database

HA in the queue

Keep calm and use HA

Implementing HA on MySQL

Implementing HA on RabbitMQ

Implementing HA on OpenStack cloud controllers

Implementing HA on network nodes

Summary

7. OpenStack Multinode Deployment – Bringing in Production

Confirming the multinode setup

Assigning physical nodes

Preparing the OpenStack Initiator

The network topology

The OpenStack network mode

The physical network topology

The OpenStack deployment

The MIN installation

Chef server preinstallation

Discover and cook

Cooking time

Testing the cloud

Arming the deployment

Bringing HA into action

Adapting the deployment

Running first tenant

Summary

8. Extending OpenStack – Advanced Networking Features and Deploying Multi-tier Applications

Navigating through Neutron

Neutron plugins

Virtual switching infrastructure

The Linux Bridge plugin

The Open vSwitch plugin

Load Balancer as a Service

Work around LBaaS

Integrate LBaaS in the cloud

Stack in OpenStack

HOT explained

Installing heat

Heating things up

Summary

9. Monitoring OpenStack – Ceilometer and Zabbix

Telemetry in OpenStack – Ceilometer

Ceilometer definition

Ceilometer glossary

The Ceilometer architecture

The Ceilometer installation

Ceilometer and heat

Autoscaling

Extending HOT for alarming

Arming OpenStack monitoring

Zabbix in action

Placing Zabbix

Installing the Zabbix server

Configuring the Zabbix agent on OpenStack nodes

Watching OpenStack

Summary

10. Keeping Track for Logs – Centralizing Logs with Logstash

Tackling logging

Demystifying logs in OpenStack

The log's location

Adjusting logs in OpenStack

Two eyes are better than one eye

Logstash under the hood

The Logstash workflow

Placing the Logstash server

Installing the Logstash server

Configuring Logstash

Logstash at your beck and call

Filtering the OpenStack logs

Visualizing the OpenStack logs

Summary

11. Tuning OpenStack Performance – Advanced Configuration

Pushing the limits of the database

Deciding the resources outfit

Caching for OpenStack

Memcached in OpenStack

Integrating memcached

Stressing RabbitMQ

Benchmarking OpenStack at scale

Rally in a nutshell

Meeting OpenStack SLA

Installing Rally

Rally in action

Scenario example 1

Scenario example 2

Summary

Index