METHOD AND A SYSTEM FOR REPLICATING DATABASES

A method for replicating databases, in which method at least two databases are replicated, each database comprising of one or more tables, and each table having data and a unique tree index comprising index keys and hierarchical sums of cyclic redundancy check values calculated from the data on each tree index level. In the method logical consistency of the data between corresponding tables in the databases is maintained by comparing the sums of cyclic redundancy check values of the database tables; and if a difference between the sums of the cyclic redundancy check values is found, the tree indexes are logically divided into two sub tree indexes, and the sums of the cyclic redundancy check values of the sub tree indexes are compared to each other and the comparison and division is continued until the data causing the difference is found, and the inconsistent data is replicated between the databases.