DRDA character type parameters in Unicode

Remote Db2 applications can send and receive DRDA command and reply message parameters that contain character type data encoded in Unicode CCSID 1208 (UTF-8). Using Unicode instead of EBCDIC for these DRDA parameters can improve performance and avoid potential character conversion errors.

Prior to DB2® 10, remote applications passed DRDA command and reply message parameters that contain character type data in EBCDIC. These applications might incur additional CPU costs and character conversion errors for the following reasons:

• Db2 for z/OS® stores metadata and catalog data in Unicode (UTF-8). Therefore, Db2 converts incoming DRDA EBCDIC data to Unicode (UTF-8).
• The IBM® Data Server driver or client must convert DRDA character type data to EBCDIC before sending it to Db2 for z/OS. The driver or client must also convert the data that is received from Db2 for z/OS, in EBCDIC, before returning it to the application.
• Other remote applications might need to convert the DRDA parameters to and from EBCDIC.

Passing these character type parameters in Unicode removes this extra conversion step.

From an application programming perspective, you do not need to perform any extra action to send DRDA character type parameters in Unicode. Db2 for z/OS automatically negotiates use of Unicode data with remote client systems that support the exchanging of DRDA character type data parameters in Unicode (UTF-8).

Because of this new ability to pass DRDA character type parameters in Unicode, potential problems might exist with certain package names and collection IDs that contain special characters. To prevent these problems, run the premigration queries.