Fit for Purpose Data Replication/Transformation
Just like choosing the right transportation option that is fit for purpose, choosing the right data replication tool from the many available is also a fit-for-purpose exercise to meet requirements.
By Doug Mack ,
IBM Db2 Web Query
Just like choosing the right transportation option that is fi t for purpose, choosing the right data replication tool from the many available is also a fi t-for-purpose exercise to meet requirements.
When it comes to replicating Db2* for i data you have many options, including some new (or new to you) solutions. IBM recently announced Db2 Mirror for IBM i, a continuous availability option for 24-7 database availability. A product that’s actually been out a few years now, IBM’s Db2 Web Query DataMigrator ETL provides extract, transformation and load (ETL) replication capabilities.
Db2 Mirror is a database-level, real-time replication solution designed specifically for mirroring your databases to a second server or LPAR for continuous availability purposes. Db2 Mirror is a good fi t for the purpose of availability, but a side benefi t of it could include horizontal scaling of your applications or even separation of transaction workloads on one system from query workloads on the second one.
DataMigrator ETL provides a low-cost data replication solution that works more on a timed basis (but could be near real time) to replicate but also transform your data. Transformations might be to simplify the data perspective to the users of analytics. Other examples are creation of aggregates or reshaping of the data to optimize it for analytical queries. DataMigrator ETL is best fi t for the purpose of consolidating data from multiple systems (including maybe non Db2 databases like SQLServer or Oracle), or creation of a data warehouse to support your query and analytics workload.
There may actually be times where your requirements might require both Db2 Mirror and DataMigrator ETL. A good example might be the Active-Passive use of Db2 Mirror, where your transactions are running on one node, but your ETL processes use the mirrored image of the database on the second node as the source to feed into a data warehouse.
Planes, trains and tuk-tuks—you might need to combine but choose the one that fits the individual purpose for the best outcome.
Doug Mack currently leads the Db2 for i Analytics team for IBM Systems Lab Services technology enablement organization.
About IBM Db2 Web Query
Db2 Web Query for i provides a complete business intelligence and data warehouse solution set designed for and integrated with IBM i and Db2 for i. Learn more about our solutions →