Notes about referencing error information

For a binary-format input file for a non-FIX table that is output by the database reorganization utility, no error data is subject to dump image listing output or error data file output to an error information file.
For a binary-format file, the following types of input data error prevent columns beginning with the affected column from being output. Consider this when referencing an error data file:
- An error in the variable-length character string or BINARY column
- Input data column length less than a defined length for that column
For a binary-format file, the dump image listing from an error information file is not output for a table containing LOB columns, columns of an abstract data type, or repetition columns. If the length section of a variable-length character or BINARY column is invalid, data is output only up to the character string preceding the erroneous variable-length character or BINARY column.
If a binary-format file contains data whose input data length or repetition-column elements count is a negative value, a row of data cannot be identified. Therefore, if the -e option is omitted, output processing is canceled; if the -e option is specified, an error information file is not created. In such a case, see the KFPL24500-E message and take appropriate action.
For a DAT-format file, data with a row length greater than the value of maxreclen operand of the source statement (if the maxreclen operand is omitted in the source, a default value of 32 KB) is not output to an error data file.
Information is output to an error data file not in the order in which data is stored in the input file, but in the order in which error messages are output to an error information file.
If data that has not been output remains in the error data file, only one warning message is issued, irrespective of the number of data items remaining in that file.
If a logical error (key duplication error or error detected by a plug-in function) occurs during communication processing when data is stored, a temporary file for creating an error data file is created on the server where the input data file is located. The created file has the following name:
error-data-file-name+data-storage-server-name
If the path name exceeds 1,023 bytes, data-storage-server-name is not added. The beginning of error-data-file-name is overwritten by data-storage-server-name in such a manner that the path name dos not exceed 1,023 bytes.
This file is created in the same directory as for the error data file and is deleted after communication processing terminates.

If an input data error is detected when the input data file is in the user-created binary format or fixed-size data format, a dump image listing of the input data in which the first error was detected is output to the error information file. If the work area (memory) required for writing the dump image listing cannot be acquired, the KFPL25222-W message is displayed. In such a case, the system resumes data loading without output of the dump image listing.
The table below shows whether a dump image listing is output when an input data error is detected.

Table 5-41 Whether or not a dump image listing is output when an input data error is detected

Input data file format	Output of dump image listing	Work memory allocation
Input data file format	Output of dump image listing	Allocatable	Unallocatable
DAT format	N	--	--
Extended DAT format	N	--	--
Fixed-size data format	Y	Outputs a dump image listing and then resumes data loading.	Resumes data loading without outputting a dump image listing.
Binary format	Y	Outputs a dump image listing and then resumes data loading.
pdrorg-output binary format	N	--	--

Legend:: Y: Output; N: Not output; --: Not applicable

5.6.3 Notes about referencing error information