Informatica Data Replication has triggered a lot of interest and also a lot of questions about why replication is seeing such a resurgence in the market today. The answer is simple: the same conditions that caused its creation for mission-critical operational systems back in the early 1990’s are now happening with data warehousing as well.
Back in the late 1980’s and early 1990’s, many organizations were struggling with distributing their RDBMS systems to scale them up and to support operational needs across geographic and/or departmental boundaries. Relational database systems could support distributed transaction activity across multiple instances, but at the cost of performance and throughput as the mechanisms to maintain transaction consistency (synchronous with two-phase commit) slowed everything down. Wall Street approached Sybase to help develop an alternative, and Sybase Replication Server, the first flagship product, was born.
Today we’re seeing the same thing happen with centralized enterprise data warehouses and operational data stores: they simply cannot meet the needs of many organizations today. Loads are increasing on every level: volumes of data, of users, and of computations to support analytics and data discovery all continue to grow exponentially. So organizations are moving to distribute their warehousing and analytics platforms, just as they had distributed their operational systems. Analytics and operational BI are now as mission-critical as many operational business applications, and require the data to be available and up-to-date 24/7. Batch windows are shrinking and for many are now virtually non-existent, so ETL/ELT alone cannot meet the need. Real-time update in place is now essential – and Data Replication enables exactly that.
So, although it’s never disappeared – it’s always been the #2 data integration technology in use after ETL – like a phoenix rising from the ashes, Data Replication is back!