By | November 10, 2021

No additional tools are required to determine whether a given data has been migrated or not. A virtual data room will enable your data from any source.

The Process of Data Transferring and Migration

Data migration is the process of transferring data from a source database to a target database, and their schemas may differ. There are many reasons why organizations are embarking on data migration projects, from application upgrades and new enterprise systems to full-scale restructuring through a merger. Analyzing the experience of data migration of a large number of information systems, a typical data migration procedure can be distinguished, which includes:

  • Analysis of data formats of the structure of the source base and target, preparation of a plan for migration and data transformation:
  • Defining relationships between tables (object hierarchies).
  • Determining the sequence of data transfer according to the dependency hierarchy. Sometimes you can ignore the relationships, but simply disable all foreign keys before migration and recreate them after all data manipulations are completed. 
  • Executing a script to change objects in a new version of the database.
  • Direct data transfer with the necessary transformations “on the fly”.
  • Executing a script to restore disabled indexes, additional conversions, etc. after completing the data migration procedure.

The simplest option for data room providers is to create a middleware. It must communicate with the source and target databases and perform the necessary transformations. When migrating, it is important to know in which order to migrate data. Therefore, within the framework of this block, there is a dialogue with the user and the choice of the type of formation of the sequence of traversing the scheme of the source database: automatic or user.

Data Migration from Any Source with Virtual Data Room

During data migration with a virtual data room, errors may occur, such as the impossibility of transfer due to mismatch of data types, mismatch of integrity constraints, etc. To avoid this, it is recommended to:

  1. Stop migration when the first error occurs.
  2. User input of correct values.
  3. Skipping records with errors and all dependent records.
  4. Padding by default.

The data transfer engine in VDR uses two other types of scripts in addition to conversion scripts. The first type of scripts is used to disable triggers and some restrictions during data migration, and the second – to enable them after data migration. With automatic sequencing, the final result is selected as the first from a list of possible alternatives. When a user-defined sequence task is selected, the system generates all possible alternative sequences, one of which is selected by the user as the final one.

The migration mechanism with VDR must ensure that all records from the source database are viewed and transferred, ensure data integrity, and ensure that records are transferred with all dependencies in mind. A record can be migrated only if it is independent, or the records that it depends on have already been migrated to the target database. Therefore, the main rule that determines the order of display and transfer is the dependencies between the records. Entry dependencies are defined by foreign keys.

The data migration mechanism with virtual data rooms is implemented in the form of a procedure that is first called for independent records, and then recursively for those dependent on this record. Therefore, the traversal sequence must start with a table containing independent records. Then the basic order given on the set of tables of the database schema is the dependency between the tables.