A
Service-Oriented Database Architecture (SODA) Approach
Designing Scale-Out Applications with SQL
Server 2005
A
Service-Oriented Database Architecture (SODA) Approach
There are two ways to
manage the growth requirements of database systems: scale-up; and scale-out.
Scale-up comes from improved hardware (more cpu, faster cpu) and from extensive
tuning by application developers and database administrators. Scale-out
requires distributing and partitioning data as well as distributing the data processing.
This involves characterizing the different types of data in your application as
all data is not the same with respect
to the database. In addition, scaling and failover capabilities can be greatly
improved by dividing the process implementation into composable services that
can execute asynchronously or in parallel. In this fast-paced immersion event,
we'll discuss designing an application architecture that implements a data
management methodology that fully reviews the purpose and usage of your data leading
to effective scale-out designs. We will provide you with a comprehensive understanding
of how to make your application scale by leveraging the new capabilities of SQL
Server 2005 that are geared around a Service-Oriented Database Architecture. The
revolutionary enhancements in the new SQL Server release enable new
possibilities for scale-out using a reliable event-driven active database model
that replaces the session-based connected database communication model that
exists today. This leads to new design patterns that deliver increased
throughput, leverage commodity hardware, and increase availability – empowering
you to deliver applications that meet ever-increasing up-time and transactional
demands.
Outline
Day 1
1.
Introduction:
SODA
¨
What does SODA
look like
¨
A Review of
Scale Up
à
SQL Server
2005 enhancements for Scale Up
¨
An
introduction to Scale Out
à
Scale Out in
SQL Server 2000
·
Replication
·
MSMQ
·
Distributed
partitioned views
à
Where
traditional scale out falls short
¨
What are the
essential components of SODA
à
Enterprise Service Infrastructure - Service Broker
à
Programming
Complex Logic - SQLCLR
à
Interoperability
and Networking – XML and Web Services
à
Routing
à
Notifications
and Caching
à
Monitoring the
Application
2.
Scale
Out - A Data Management Methodology
¨
Data
Classification
à
Reference Data
à
Resource Data
à
Activity Data
à
Service
Interaction Data
¨
Services and
Databases
à
Connected
systems vs. Disconnected Systems
·
Services, Queues,
Message Types, and Contracts
·
Conversation
Management (Begin-End Conversation)
·
Sending and
Receiving Messages
à
Loose Coupling
·
Homogeneous
Systems
·
Heterogeneous
Systems
·
XML and Web
Services in SQL Server 2005
à
Event-driven
database systems
·
Query Notification
·
Event
Notifications and WMI
à
Data dependent
routing and partitioning
·
Routes
·
Remote Service
Bindings
·
Data-directed
routing
·
Peer-to-Peer Replication
à
Demand and
asynchronous services
·
Time-based
scaling using activation programs
·
Writing
Activation Programs with SQLCLR
à
Challenges in
Service design
|