Mastering OpenStack - Episode 01 - Simple Architectures

Presentation By:
Roozbeh Shafiee
Spring 2015
IRAN OpenStack Users Group
MASTERING
OPENSTACK
(Episode 01)
Simple Architectures

IRAN Community| OpenStack.ir
OpenStack Simple Architectures
Agenda:
●
Requirements Before Beginning
●
OpenStack Components
●
OpenStack Usage Porposes
●
Advises Before Starting A Project
●
Components Relations
●
Lagcy Networking
●
Iran OpenStack Community

|
Requirements Before Beginning
IRAN Community OpenStack.ir

|
Requirements Before Beginning:
●
Dominance on Virtulization and HA Topics (RHCVA or LPIC-3)
●
Linux Administration Skills ( Ubuntu/Debian or RHEL/CentOS )
●
Database Concepts ( SQL and NoSQL )
●
Basic Network Knowledge ( Cisco CCNA )
●
Hardware Maintenance Experience ( Servers and Storages )

|
OpenStack Components

|
OpenStack Components (Kilo Release):
●
Compute - Nova
●
Networking - Neutron
●
Object Storage - Swift
●
Block Storage – Cinder
●
Image Service - Glance
●
Identity - Keystone
●
Dashboard - Horizon
●
Telemetry - Ceilometer
●
Orchestration - Heat
●
Database - Trove
●
Data Processing - Sahara
●
Bare-Metal Provisioning – Ironic
●
Queue Service - Zaqar
●
Shared file system - Manila
●
DNS Service - Designate
●
Key Management - Barbican

|
OpenStack Components Schema:

|
OpenStack Simple View Schema:

|
OpenStack Compute – Nova:
●
Conceptually similar to Amazon EC2
●
Can leverage multiple hypervisors (Xen, KVM, VMware ESXi, Hyper-V)
●
Made up of several sub-components (nova-api, nova-compute,
●
nova-schedule)

|
OpenStack Networking – Neutron:
●
Provides “network as a service” functionality
●
Allows for the programmatic creation of per-tenant networks
●
Uses a plug-in architecture; plugins exist for a number of different
back-end implementations, including Nicira NVP, Open vSwitch (OVS),
Linux bridge, Cisco, and others
●
Has essentially replaced nova-network

|
OpenStack Object Storage – Swift:
●
Think of it as similar to Amazon S3
●
Provides distributed object storage
●
Supports the OpenStack Object API as well as raw HTTP
●
Authentication is handled via OpenStack Identity (typically)

|
OpenStack Block Storage – Cinder:
●
Can be compared in concept to Amazon EBS
●
Provides block storage functionality to instances running on Compute
●
Recall that instances are stateless
●
Block Storage provides persistent storage for instances
●
Similar to Compute, it has several sub-components (cinder-api,
cinder-volume, cinder-scheduler)
●
Leverages storage drivers to interact with underlying storage platforms

|
OpenStack Image Service – Glance:
●
This can be compared to Amazon’s AMI catalog
●
Provides image storage, image retrieval, and image discovery services
●
These images are leveraged to launch instances (VMs)
●
Can leverage the Object Store (“Swift”) or other storage platforms,including
Amazon S3, filesystems, and HTTP

|
OpenStack Identity Service – Keystone:
●
Provides policy and authentication services for other OpenStack services
●
Leverages a pluggable architecture that can support LDAP, SQL, and other
backend services

|
OpenStack Dashboard – Horizon:
●
A web application that runs on Apache
●
Provides a graphical user interface (GUI) for other OpenStack services
●
Leverages the APIs of the other OpenStack services

|
OpenStack Telemetry – Ceilometer:
●
It is easier to roll back a failed deployment and ensure thorough cleanup
●
Administrators can delegate resource creation privileges to non-administrative
users
●
Improvements included implementation of new resource types and improved
scalability

|
OpenStack Orchestration – Heat:
●
It is easier to roll back a failed deployment and ensure thorough cleanup
●
Administrators can delegate resource creation privileges to non-administrative
users
●
Improvements included implementation of new resource types and improved
scalability

|
OpenStack Database Service – Trove:
●
The database service went through its second release cycle in Juno
●
MySQL replication
●
Mongo clustering, Postgres, and Couchbase.

|
OpenStack Data Processing – Sahara:
●
Automates provisioning and management of big data clusters using Hadoop
and Spark
●
Big data analytics are a priority for many organizations and a popular use case
for OpenStack
●
This service lets OpenStack users provision needed resources more quickly

|
OpenStack Bare Metal Provisioning – Ironic:
●
An integrated OpenStack program which aims to provision bare metal machines
instead of virtual machines
●
Forked from the Nova baremetal driver
●
Best thought of as a bare metal hypervisor API and a set of plugins which interact
with the bare metal hypervisors
●
Use PXE and IPMI in concert to provision and turn on/off machines

|
OpenStack Usage Purposes

|
OpenStack Some Usage Purposes:
●
High Performance Computing (HPC)
●
Content Delivery Network (CDN)
●
Multi Site - Geo-location Sensitive Data
●
Big Data Analytics
●
Hybrid Cloud
●
Network Function Virtualization (NFV)
●
Media Streaming
●
Virtual Desktop Infrastructure (VDI)
●
Voice over IP (VoIP)
●
And many more...

|
Advises Before Starting

|
Advises Before Starting:
●
Using enterprise-ready operating systems such as Ubuntu (LTS) and RHEL
●
Using standard physical servers and storages by reputable Hardware vendors
●
Team working with enthusiastic members
●
Document all states of the project
●
Report probably bugs to OpenStack development team
●
Follow OpenStack distribution to find best solution for projects

|
Components Relations

|
How Do OpenStack Components Work?
●
Daemon
●
Script ( Mostly Python and Shell )
●
Command-Line Interface ( CLI )
All services authenticate through a common identity service and individual services
Interact with each other through public APIs except where privileged administrator
Commands are necessary.

|
Legacy Networking
(nova-network)

|
Simple Architecture – Legacy Networking (Nova)

|
Tools and Dependencies:
●
OpenStack Release: Kilo
●
Host Operating System: Ubuntu 14.04 / RHEL 7.1
●
OpenStack Package Repository: Ubuntu Cloud Archive / RDO
●
Hypervisor: KVM
●
Database: MySQL
●
Message Queue: RabbitMQ (Ubuntu) / Qpid (RHEL)
●
Networking Service: Nova-Network
●
Network-Manager: FlatDHCP
●
Image Service (Glance) Backend: File

|
●
Identity Service (Keystone) Driver: SQL
●
Block Storage Service (Cinder) Backend: LVM/iSCSI
●
Live Migration Backend: Shared Storage using NFS
●
Object Storage: OpenStack Swift

|
Don't Use OpenStack Packages From Main Repositories
We recommend that you do not use the default Ubuntu OpenStack install packages
and instead use the “Ubuntu Cloud Archive” PPA from Launchpad. The Cloud Archive
is a package repository supported by Canonical that allows you to upgrade to future
OpenStack releases while remaining on Ubuntu 14.04.

|
Why not use the OpenStack Network Service (neutron) ?
This simple architecture does not use the OpenStack Network Service (neutron),
because it does not yet support multi-host networking and our organizations (univer‐
sity, government) have access to a large range of publicly-accessible IPv4 addresses.

|
Why Multi-Host Networking ?
Multi-host is a high-availability option for the network configuration, where the
nova-network service is run on every compute node instead of running on only a
single node.

|
What Does OpenStack Controller Node Contain ?
●
Horizon Dashboard
●
API Services
●
Database (MySQL)
●
Message Queue Server (RabbitMQ)
●
Scheduler For Choosing Compute Resources ( nova-scheduler )
●
Identity Services (Keystone, nova-consoleauth )
●
Image Services ( glance-api , glance-registry )
●
Services For Console Access of Guests, and Block Storage Services, including
the scheduler for storage resources ( cinder-api and cinder-scheduler )

|
What Does OpenStack Compute Node Contain ?
●
Hypervisor (KVM)
●
Libvirt
●
nova-compute
●
Nova-api-metadata (generally only used when running in multi-host mode)
●
nova-vncproxy
●
nova-network

|
What's The Scenario in Network Section ?
●
The network consists of two switches (One for the management/private and the
other one for public traffic)
●
The cloud controller and the compute nodes have two network cards
●
The OpenStack Block Storage and NFS storage servers only need to access the
private network (Only need one network card)
●
Multiple cards run in a bonded configuration are recommended if possible.
●
Floating IP access is direct to the Internet, whereas Flat IP access goes
through a NAT.

| IRAN Community OpenStack.ir

|
Iran OpenStack Community

|
Stay in Touch and Join Us:
●
Home Page: OpenStack.ir
●
Meetup Page: Meetup.com/Iran-OpenStack
●
Mailing List: OpenStack-ir@Lists.OpenStack.org
●
Twitter: @OpenStackIR , #OpenStackIRAN
●
IRC Channel on FreeNode#OpenStack-ir

Roozbeh Shafiee
Iran OpenStack Community Manager
Roozbeh@OpenStack.ir
OpenStack.ir
Thank You
We need to work together to build a better community

Mastering OpenStack - Episode 01 - Simple Architectures

More Related Content

What's hot (20)

Viewers also liked (6)

Similar to Mastering OpenStack - Episode 01 - Simple Architectures (20)

More from Roozbeh Shafiee (17)

Recently uploaded (20)

Mastering OpenStack - Episode 01 - Simple Architectures