Predictive analytics are only as good as the data you use to train and test with and the currency of the data used for analysis. In many organizations, particularly in banking, finance and insurance, that data is in the mainframe. In fact, 97 of the top 100 banks use System z, as do 10 of the top 10 insurance companies.
Key insights lie within the customer and transaction data, but it can be difficult to extract from databases like IBM’s DB2 or CA’s IDMS, or in enduring formats like VSAM or tape with metadata encoded in COBOL or PL/1.
Even without the additional complexity of integrating mainframe sources, in many projects data integration can consume up to 80% of the dev effort for big data projects.
Data Integration and Management Platform
vStorm Enterprise is a complete software solution that will help you:
Access data from relational databases, log files, and mainframe databases and file systems
Solve data conversion challenges of COBOL and PL/1 Copybooks, EBCDIC and other mainframe formats
Deliver data to your analytics platform in near-real-time
Respect mainframe governance concerns
Gain a single platform for JDBC, mainframe, log file and other data integration formats
Integrate data from and Oracle/MySQL, Teradata, Postgres, Netezza, and even other Hadoop platforms
In the Hadoop ecosystem, Apache Sqoop has been a reliable way to access JDBC data from distributed systems. Although it can be stretched to access some JDBC mainframe data, it cannot convert formats like VSAM and it isn’t a cost-efficient solution because of the different cost structure of the mainframe.
Many of the steps required to implement Sqoop to extract mainframe RDMS data generate MIPS charges or incur additional mainframe storage costs for staging. vStorm Enterprise avoids this by extracting data in binary form and performing all operations in-flight, with the conversion processing performed by the software on Linux. The result is no staging, no MIPS (processing) charges.Learn more about Sqoop on the mainframe