How to Optimize Migration of Large Tables

There are several ways to reduce migration time and speed up the migration process.

1. If there have been no problems during the export you can turn off trace mode. Thus there will be no trace generation which will help to reduce migration time.
   
2. You can split the data migration into parts and run them simultaneously in our tool. This can be done in two ways:
   
   1. Using Data Options
   2. Using the Queries

When you select “Data” tab at “Conversion Options” page for Tables level in the object tree you see Edit boxes for the start row and the number of rows to be exported (like below). By varying these numbers in separate instances of Wizard (like 1-5000, 5001-10000, etc.) you can run the migration of different parts of data simultaneously.



Please, also note that you should uncheck “Generate DDL” option at the tab “DDL” for processes other than the first one to avoid error of table existence. In this case only data will be migrated.

You can export data according to unique index or primary key or any other query condition. To perform this in Wizard please select “Use Query” in combo-box at the Object Selection“ page like below. Using this option you can write your own query to select data. DDL statement will be generated according to the selected columns. Table name will be the same as in FROM clause. But we advise you to create empty tables first and then to do the migration of data this way. You’ll get errors of existing object when trying to create a table, but data will be loaded fine.



For both these solutions please make sure that the “Generate the DROP TABLE statement” check-box is unchecked at the options page. Please also use separate project directories for each migration to prevent project files recreation.

If you have any other questions, please contact us: support@ispirer.com