Database replication for disconnected operations with quasi real-time synchronization

Abstract

Database replication is a way to improve system throughput or achieve high availability. In most cases, the use of an active-active replica architecture is efficient and easy to deploy. Such a system has CP properties (from the CAP theorem: consistency, availability, and network-partition tolerance). Creating an AP (available and partition-tolerant) system requires the use of multi-primary replication. Because of the many difficulties in its implementation, this approach is not widely used, however, ALICE’s deployment of CCDB (experiment conditions and calibration database) needs to be an AP system in two locations. This necessity became the inspiration for examining the state-of-the-art methods in this field and testing the available solutions. The tests that were performed evaluated the performance of the chosen replication tools: Bucardo, and EDB Replication Server, these showed that the tested tools could be successfully used for the continuous synchronization of two independent database instances.