Distributed Server: Using SASHDAT Libraries

When your deployment uses Hadoop as a co-located HDFS or NFS-mounted MapR, the SAS Deployment Wizard registers a predefined library for it. This library is available for use in the SAS Folders tree, and it is located in /Shared Data/SAS Visual Analytics/Public/Visual Analytics Public HDFS.

You can specify a SASHDAT library as a staging library. This is a common use because the rows for the output table are distributed among the machines in the cluster. The server can then read the data in parallel when it loads the table to memory.

You must specify a SAS LASR Analytic Server library for the output library when you use a SASHDAT library for staging.

You can specify a SASHDAT library as an output library. The engine distributes the rows for the table to the machines in the cluster. Afterward, you can select the table from the SAS Folders tree, right-click, and select Load a Table. This menu option loads the table to memory on a SAS LASR Analytic Server.

You can partition SASHDAT tables when they are used in an output library. You can select a column to use from the Partition by menu. Partitioning the table ensures that all of the rows with the same formatted value as the selected column are distributed to one machine in the cluster. The rows are also placed in the same block. When you load a partitioned table to memory, the partitioning information is retained, and the result is a partitioned in-memory table.

The following restrictions apply to using SASHDAT libraries with SAS Visual Data Builder: