Data Source Access and Processing

The LIBNAME engine for DataFlux Federation Server supports the following processing:
  • The engine supports input (read), output (create), and update (modify, add, and delete) processing. You can submit most SAS language procedures, including the SQL procedure, perform WHERE and BY-Group processing, and submit the DATA step with the DataFlux Federation Server data services to read, create, update, and delete data.
  • You can list a data source's tables with the DATASETS procedure, and you can list the metadata attributes of a table with the CONTENTS procedure. However, the engine does not support utility processing to modify metadata attributes such as renaming or adding a column or changing a label. The MODIFY statement for PROC DATASETS is not supported because it requires the file to be opened in utility mode.
  • The engine is a sequential engine with limited random access. If you request processing that requires random access that is not supported, a message in the SAS log notifies you that the processing is not valid for sequential access.
  • The engine does not support threaded application processing. Therefore, SAS procedures such as the SORT and SUMMARY procedures cannot perform threaded processing.
  • The engine does not support DATA step views and PROC SQL views.
  • The engine cannot access a damaged file and therefore provides no means for repairing a damaged file.
Some SAS functionality is restricted by the data source itself. For example, when you use the engine to access a SAS data set, some SAS features such as audit trail, referential integrity constraints, generation data sets, and Cross-Environment Data Access (CEDA) are not supported. For data source functionality details, see the data source reference in DataFlux Federation Server Administrator’s Guide.