SERDE_JAR_SYNC stored procedure

The SERDE_JAR_SYNC stored procedure copies a SerDe or UDF JAR file from a Hadoop DFS file system directory into the appropriate installation directories on your Hadoop cluster and adds it to your class path. This action makes the objects available for use within queries without stopping and restarting the Db2 Big SQL server.

Syntax

Read syntax diagramSkip visual syntax diagram SERDE_JAR_SYNC ( )

Description

Copies JAR files that are available in the Hadoop distributed file system (DFS) /bigsql folder in either the /serde/ folder or the /udf/ folder to the appropriate installation directories on the Hadoop cluster. The class path is also updated.

It is assumed that you already uploaded your JAR file to the Hadoop DFS before you run the stored procedure.

After the procedure call, you can use the JAR files within Db2 Big SQL without stopping and restarting the Db2 Big SQL server.

Authorization

Only the bigsql user or a user with Db2 Big SQL administrative privileges, can run this Hadoop procedure. However, the bigsql user can grant execute permission on the SERDE_JAR_SYNC procedure to any user, group, or role. For more information see GRANT (routine privileges) statement.

Examples

  1. Download a SerDe JAR file from a reliable web site. You then upload that JAR file to the Hadoop DFS in folder /bigsql/serde. The Db2 Big SQL administrator has granted you the proper write access to this folder. Then, you register the SerDe JAR file in DFS and copy the file into the correct installation directory on your cluster. 
    
CALL SYSHADOOP.SERDE_JAR_SYNC();