Data Migration is the process of transferring data among repositories, formats and computer systems. Data migration is usually performed programmatically for automated migration, freeing up human resources from tedious tasks. This is necessary when an organization needs to change computer systems or move to new, also merging systems.
For effective migration data from previous system displays on new providing design for extracting and loading data. Programmatic data migration may involve many steps, but it contains minimum data retrieval when data is read from the old system and data load when data is written into the new system.
After loading into new system, results are checked to determine whether the translation was accurate, everything is moved, and whether processes are supported in the new system. During check load of both systems parallell may be necessary for identifying areas of non-compliance and to prevent erroneous data loss.
Automatic and manual data cleaning is usually done to improve the quality of data, eliminate redundant or outdated information, as well as to verify compliance with the requirements of the new system.
Stages of data migration (design, extraction, cleaning, loading, testing) for applications, from moderate to high complexity, usually are repeated several times before deploying a new system.