Optimize OWB & ODI Performance

Speed Oracle Unloads, Loads and Transforms



The Oracle DataPump and export utilities can unload data from Oracle, but not much else. The speed of data acquisition from other sources may be insufficient.


Transformations in a single database are inefficient. Large jobs consume database resources, and can degrade query performance. OWB and ODI data mappings can be complex, and may require SQL procedures to modify.


Oracle recognizes that "flat files are frequently used for simple and high performance data movement in ETL applications." However, Oracle is not a flat-file data movement or sorting specialist and, as such, cannot accelerate bulk loads alone.


Beyond performance and functional concerns, ODI pricing may be prohibitive, especially relative to more modern, nimble vendor data warehousing platforms which support more data sources and include more data management capabilities.


Big data bottlenecks are large sorts, joins, aggregations, loads, and sometimes unloads. Parallelization or optimization in other layers or tools can be unwieldy, if not expensive, and may adversely affect performance for other users.


From a security standpoint, IBM's data masking solutions may be expensive or cumbersome for some, or may not provide all of the PII discovery or privacy functions for others.


  • Bulk Extract to Flat Files

    IRI FACT (Fast Extract) software speeds Oracle and other very large database (VLDB) unloads. This is important in both ETL and ELT environments. FACT is faster and less complicated than the "parallel table function" in OWB. See:

    Blog > VLDB > Unload Big Data From Oracle et al

  • Speed ETL and SQL Transforms

    IRI CoSort software performs the same transformations as SQL, only 10x faster. The CoSort SortCL program runs in the file system, bypassing one or more transformation stages in OWB and ODI, including sorting, joining, and aggregation.


    Remove the overhead of bulk data transformation from Oracle by specifying and combining these operations in a single SortCL script and I/O pass. For more information, see:

    Blog > Structured Data Transformations


    In addition to efficiency, the data and mapping definitions in SortCL are in simple 4GL text files anyone can understand and easily modify. The language of SortCL is also modeled in the IRI Workbench GUI, built on Eclipse. Metadata and job script creation are also automated and supported in this rich, familiar IDE.

  • Speed Bulk Loads into Oracle and Other RDBs

    Any of the sort utilities included in the IRI CoSort package will speed bulk loads into most relational DBs, including Oracle, DB2, Sybase, and SQL*Server. For more information on speeding Oracle loads, refer to:

    Blog > VLDB Operations > High Speed DB Loading

  • Re-Platform ODI ETL Operations

    You can now migrate ETL jobs in ODI to Voracity automatically through AnalytiX DS Code-Automation Frameworks (CATfx) technology. Both Informatica and Voracity metadata are modeled in AnalytiX DS Mapping Manager so you can move projects between the two platforms with ease. Voracity is both several times faster and less expensive than ODI, and with proven AnalytiX DS CATfx technology and Lite Speed Conversion services, it is finally possible (and cost-effective) to switch.