Data Migration is moving data from place to place; more precisely, between computer systems, formats, and storage types. It is also recognized as simply translating data from one format to another. In order to have a successful data migration, computers will use data extraction and data loading in which the old data will be formatted to the new data system. Data extraction is when the data is read from the old system; data loading is where the data is being loaded onto the new system. Companies are usually forced to migrate data when they decide to switch computer systems or database management systems.
§ Businesses will sometimes will migrate date in order to take advantage of more efficient storage technologies. By using virtualization, one will move physical blocks of data from one tape to another. The process of data migration will almost always leave the data unchanged.
§ One of the main reasons that businesses perform data migration is to update software or move from one database vendor to another. In the physical transformation process there is a chance that the data be affected negatively, unless it is a modern database system. Modern database systems are written to be agnostic – meaning that the database only requires a testing cycle in order to be confident that the performance will be adversely affected.
§ Application migration is a very substantial transformation. The reason is because every application or suite operates on its own specific data model. To keep the integrity of data being migrated, one would use application programming interfaces (API’s).
§ Business process migration is the movement of databases, applications and data from one organization to another. Primary reasons for business process migration are driven by mergers and acquisitions, reorganization, and enter new markets in order to gain a competitive advantage.
§ Storage and database migration will be taken care of by the IT department without involvement of the entire business. As far as application and business process migration, this is a different aspect. When interacting with these last two categories in the process of data migration, significant downtime, synchronization, clear visibility for stakeholders, and much more need to be taken into account.
Data Migration has many critical steps and numerous considerations that need to be taken into account before performing this daunting task. Here are five key characteristics when considering data migration:
1) Performance-How quickly can data be copied from the source to the target and balanced against system overhead?
2) Primary volume/ source data protection-If something goes wrong, the migration can be terminated and application processing restarted or continued on the source data/device.
3) Tiered storage-Many data centers use hardware from several vendors, which can result in source and target hardware being from different vendors.
4) Application downtime-Applications have different levels of business criticality and therefore have varying degrees of acceptable downtime.
5) Multivendor environments-Many data centers use hardware from several vendors, which can result in source and target hardware being from different vendors.
When performing Data Migration, of any kind, you have to be careful that the carrier and other technologies may be lost in the transfer, leaving the migration useless. A migration can also be very time consuming: the reason is because its a continual process, which also introduces the, "costly" aspect of the equation.