SQL/MX 2.x Reference Manual (H06.10+, J06.03+)

SQL/MX Utilities
HP NonStop SQL/MX Reference Manual—544517-008
5-58
Examples of import
•
The number of rows imported successfully.
•
A line saying "Import Completed Successfully" or "Import Completed with non-fatal
errors" if import has reached the end of the input data without encountering any
non-ignorable errors.
•
The name for the parsing error log file (if any).
•
The name for the parsing error message log file (if any).
•
The name for the execution error log file (if any).
•
The name for the execution error message log file (if any).
If import terminates due to any non-ignorable error, the following information is also
appended:
•
The value, X, to specify with the -F <first-row> option on a subsequent execution
of import to have the subsequent import pick up where the first import left.
•
A warning message if the last Y number of rows in the parsing error output file
were detected on or after row X (see the previous bulleted point) in the input file (if
Y is greater than 0).
•
A warning message if the last Z number of rows in the execution error output file
were detected on or after row X (see previous bullet) in the input file (if Z is greater
than 0).
When transactions are being used, X is the number of rows read by import as of the
last successful transaction.
When transactions are not being used (while using the Fast Loading technique --
known as Side-Tree inserts by project personnel) and termination is due to exceeding
a user-specified error threshold, X is the number of rows read by import before the
threshold was exceeded.
When transactions are not being used and termination is due to a non-ignorable error
other than exceeding a user-specified error threshold, X is zero.
The two warning messages in the above list are needed for you to understand that if
you fix the problem in the error output file(s) and import those rows before restarting
import with the original input file using the -F <first-row> option, the restarted
import will attempt to import those rows for the second time (unless you delete those
rows from the input file).
The rows with errors reported after the row, indicated for <first-row> must be dealt with
carefully. If you fix such rows in the error log file(s), import the fixed rows from the
error log file(s), but do not delete the rows in the original input file before restarting
Note. If the associated error messages are interspersed in the same file as the rows in error,
the values reported for Y and Z will count each associated error message as if it were a row
(even if the input file is of type FIXED and the associated error messages are variable in
length.)