…optimise your IT investments
Replication for real-time warehousing
Philip HowardResearch Director – Bloor Research
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Agenda
What is data replication?
When would you use it?
What are its requirements?
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
What is data replication?
“The process of copying a portion of a database from one environment to another and keeping the subsequent copies of the data in sync with the original source. Changes made to the original source are propagated to the copies of the data in other environments.”
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
When would you use data replication?
Data warehousing and BILoading real-time data for operational BISupporting real-time query/reportingIntegrating CEP with operational data
Operational synchronisationE.g. Lookers v BookersE.g. synchronising (POS and) central pricing data
High/continuous availabilityData migration (zero downtime)Master data management
To update/broadcast from a hubHigh/continuous availability
…
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Enabling data replication
PerformanceNative interfacesSupport for parallelismCompressionChange data captureImpact minimalism
HeterogeneityTopology supportSynchronisationGraphical development and management/monitoring
In operational/HA environments: transactional integrity
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Performance 1: native interfaces
High level interfaces (O/JDBC) not fast enough
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Performance 2: parallelism
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Performance 3: compression
One size does not
fit all
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Performance 4: CDC
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Performance 5: impact minimalism
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Heterogeneity
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Topology support
1 to 11 to ManyMany to 1
M to M1 to 1 to 1
etc
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Synchronisation
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Development & Monitoring
telling the Information Management storyConfidential © Bloor Research 2009 telling the right storyConfidential © Bloor Research 2011
Conclusion
Replication serves sundry purposesFastest growing adoption for BI
Key requirement is performance but multiple others
Complementary (not competitive) to both data integration and data virtualisation