The aim of the paper is to design and implement an approach that provides high availability and load balancing to PostgreSQL databases. Shared nothing architecture and replicated centralized middleware architecture were selected in order to achieve data replication and load balancing among database servers. Pgpool-II was used to implementing these architectures. Besides, taking advantage of pgpool-II as a framework, several scripts were developed in Bash for restoration of corrupted databases. In order to avoid single point of failure Linux HA (Heartbeat and Pacemaker) was used, whose responsibility is monitoring all processes involved in the whole solution. As a result applications can operate with a more reliable PostgreSQL database server, the most suitable situation is for applications with more load of reading statement (typically select) over writing statement (typically update). Also approach presented is only intended for Linux based Operating System