AZR211: Windows Azure: An overview - Ecomtim

spongereasonInternet and Web Development

Nov 12, 2013 (3 years and 11 months ago)

84 views

Compute/Storage/Network

Users

On
-
Premises Data Center

A

VM

A

VM

A

VM

A

A

A

Application

VM

Virtual Machine

TIME

IT CAPACITY

Actual

Load

Allocated

IT
-
capacities

“Waste“ of
capacities

“Under
-
supply“
of capacities

Fixed cost of
IT
-
capacities

Load
Forecast


Barrier for

innovations

On
-
Premises

Service Provider

A

VM

A

VM

A

VM

A

VM

A

VM

Compute/Storage/Network

Cloud Platform

Compute/Storage/Network

A

A

VM

Users

A

A

Cloud Applications

Compute/Storage/Network

A

The
focus
of
this talk

Actual

Load

Allocated IT
capacities

Reduction

of initial
investments

Reduction of
“over
-
supply“

No “under
-
supply“

Possible
reduction of
IT
-
capacities
in case of
reduced load

TIME

IT CAPACITY

Load
Forecast

Usage

Compute

Time

Average


Inactivity

Period

On & off workloads (e.g. batch job)

Over provisioned capacity is wasted

Time to market can be cumbersome

Compute

Time

Average Usage

Unexpected/unplanned peak in demand

Sudden spike impacts performance

Can’t over provision for extreme cases

Average Usage

Compute

Time

Successful services needs to grow/scale

Keeping up w/ growth is big IT challenge

Cannot provision hardware fast enough

Compute

Time

Average Usage

Services with micro seasonality trends

Peaks due to periodic increased demand

IT complexity and wasted capacity


IIS

Web Role

Instances

Virtual Machine

Worker Role
Instances

Load
Balancer

HTTP/HTTPS, TCP

Application

Database

Virtual

Server

Database

Database

Application

Table

Partition

Partition

Partition

Key A

Key B

Key C

Partition key

Key 1

Key 2

Key 3

Key 1

Key 2

Key 3

Key 1

Key 2

Key 3

Row key

Key 3

Key B

Property

Application

Container

Container

Blob

1000110100110
0111101111101
1011010001101

Blob

1000110100110
0111101111101
1011010001101

Blob

1000110100110
0111101111101
1011010001101

Blob

1000110100110
0111101111101
1011010001101

Blob

1000110100110
0111101111101
1011010001101

1000110

Application

SQL Azure

SQL Azure

Synchronization
with SQL
Data
Sync

SQL Server

Synchronization

w
ith SQL Data
Sync

Southeast
Asia

East


Asia

North
Europe

Western
Europe

North

Central US

South

Central US

Windows Azure
Application

Traffic

Manager

Application

Windows Azure
Datacenter

On
-
Premises

Computer

On
-
Premises
Applications and
Data

Web/Worker
Role

Instances

Windows Azure
Application

Endpoint
Agent

IPsec

Windows Azure Application

Queue

5) Delete
message

2)
Enqueue

message

3)
Dequeue

message

Web Role

Instance

1) Receive
work

4) Do
work

Worker Role

Instance

Service Bus

Application

(On
-
premises or
cloud)

Application

(On
-
premises or
cloud)

Windows Azure

Blobs

10110110
10001101

01110110
10101111

Windows Azure

Windows Azure
Application

In
-
Memory Caching
Service

Cached Data

Cached Data

Cached Data

Client

Access Control

Application

(
On
-
premises

or
cloud
)

Identity
Providers (
IdPs
)

AD FS
2.0

Windows
Live ID

Google

Facebook

Yahoo

OpenID

2
) Authenticate
user, then return
IdP

token

IdP

Token

1
) Access application and get
redirected to
IdP

4
) Validate
IdP

token, then create
Access Control (AC) token according to
rules for this application

Rules Engine

5
) Return
Access Control
token

AC
Token

3)
Send
IdP

token

to
Access
Control

IdP

Token

7
) Validate Access
Control token, then
use its contents

6
) Submit
Access Control
token

AC
Token

Browser

Content Partner

Datasets

Applications

Client

Inbound Bandwidth:

Free

Outbound Bandwidth:

$
0.12/GB
(North
America/Europe)

$
0.19/GB
(Elsewhere
)

Compute

Web/Worker role:
$0.04
to $0.96/instance per
hour
, depending on
instance size

Storage

SQL Azure:

$9.99/GB
per
month

1000110100110
0111101111101
1011010001101

K

K

K

Tables and blobs:
$
0.14/GB
per month,
$0.01/10,000
operations