SQData Hadoop Replication Feature

SQData enables the replication of changed data into Hadoop and/or Hbase from traditional sources such as IMS, DB2, Oracle, SQL Server, VSAM, etc.  SQData supports both point-in-time (insert only) and synchronized (incremental updates) to Hadoop targets, both of which are described below in more detail.

Highlights:

  • Replication directly into Hadoop and/or Hbase from IMS, DB2, VSAM and non-z/OS databases
  • Parallel apply for high throughput and reduced latency
  • Tunable update cycles based on time and/or the number of CDC records pending
  • Supports both synchronized and point-in-time targets
  • Provides bulk data load capabilities
  • Extensive built-in data cleansing and transformation capabilities
  • Real-time performance metrics and alerts
  • Includes all SQData toolset features, including event publishing, active-active replication, etc.

Point-in-Time Replication Replication → How it Works

All source changes are appended to an existing target HDFS file(s).  Use cases typically include slowly changing dimensions, tracking trends, and/or historical auditing of the source data.  Analytics leverage timestamps and source change operation (insert, update, delete) for point-in-time targets.

hadoopnew