Using the FEEDBACK command

You can retrieve diagnostic data about a previously issued API command by issuing the FEEDBACK command.

The type of command for which you want diagnostic data affects how you specify the FEEDBACK command and where the data is placed:

A command that processed a result set

Use the RESULT option of the FEEDBACK command to retrieve data about the last command that processed a specific result set.

If the command that processed the result set returned a RESPONSE other than OK, a FEEDBACK resource table record is appended to the end of each resource table record in the result set that had an error associated with it. You can use the FIRST, NEXT, and COUNT options of the FEEDBACK command to retrieve multiple FEEDBACK records.

The diagnostic data in a result set is available to the FEEDBACK command until another command processes the same result set. At that point, the data is replaced with FEEDBACK records for the subsequent command.

Note: No FEEDBACK records are produced if the command that processed the result set returned a RESPONSE of OK.

A command that did not process a result set

Use the FEEDBACK command without the RESULT option to retrieve data about the command issued immediately before FEEDBACK.

The FEEDBACK resource table records are returned in a separate feedback area. The records in that feedback area are cleared and refreshed for each command that is not result set-oriented. So for commands that place their diagnostic data in the feedback area rather than in a result set, FEEDBACK can retrieve data only for the most recently issued command.

Once you have issued the FEEDBACK command to retrieve diagnostic data for a command, the feedback record or area is cleared. You cannot request the same FEEDBACK resource table records more than once.