We can write a Perl or shell script to start multiple datacopy process with TabSplit utility, now you don't need to write such a Perl or shell script, I have combined the code of TabSplit and DataCopy, we can start multiple datacopy thread on Windows or multiple process on Linux/Unix to parallel data copy between different Oracle databases quickly. It's native parallel support now!
Just need to specify the parallel degree with "DEGREE" command line option.
datacopy user=..... table1=sql_summary table2=sql_summary2 degree=2
Table SQL_SUMMARY has ten million rows, let's check out the log output.
0 rows processed at 2010-04-16 13:35:55.
0 rows processed at 2010-04-16 13:35:55.
1000000 rows processed at 2010-04-16 13:36:04.
1000000 rows processed at 2010-04-16 13:36:04.
2000000 rows processed at 2010-04-16 13:36:14.
2000000 rows processed at 2010-04-16 13:36:14.
3000000 rows processed at 2010-04-16 13:36:24.
3000000 rows processed at 2010-04-16 13:36:24.
4000000 rows processed at 2010-04-16 13:36:34.
4000000 rows processed at 2010-04-16 13:36:34.
5000000 rows processed at 2010-04-16 13:36:43.
5000000 rows processed at 2010-04-16 13:36:43.
5207041 rows processed at 2010-04-16 13:36:45.
5378556 rows processed at 2010-04-16 13:36:47.
DataCopy now need to access some dictionary views, we should grant select catalog role to the source user. You can download the new version for performance testing, I hope DataCopy to be a very useful data migration utility for Oracle.