Nowadays we have a project of a high-performance parallel encrypted database on a cluster platform that will work with large volumes of data. For the smooth operation of the system, a reliable and high-speed data replication mechanism is required which should satisfy certain criteria of PostgreSQL DBMS and the mentioned project. In this article we consider the very essence of data replication, replication strategy, the basic methods and techniques of replication and also made compliance analyses with the replication method for high-performance parallel DBMS on the cluster platform. We also made an experiment to study the reliability and speed of the method for data replication between nodes in the cluster where the work was done with different amounts of the replica and a different nodes number. The aim was to obtain information on the speed of the Slony-I replication method at different volumes of the replicated data and the different number of nodes in the cluster on which we make a conclusion about the reliability of this method in relation to the project mentioned above.