How to properly organize geographically distributed asynchronous replication between PostgreSQL-XL clusters?

Question

PostgreSQL

PostgreSQL-XL does not support geo-distributed asynchronous replication out of the box. Only synchronous on fast networks. There is such task:
Two geographically separated autonomous Postgresql-XL clusters. There is a need between them to make asynchronous replication. At the same time, clients only read and each client goes to his own XL. The agent that writes is only one and it writes data to both clusters, taking into account asynchronous data replication. Tell me, please, is it advisable to use PostgreSQL BDR for replication purposes in this case? Is it even possible? What tools for this can be better suited than BDR?

Answer 1

So you want a weakly-coupled multimaster? Yes, BDR does it. However, it is clear that your applications will need to learn how to deal with conflicts and ensure that everything is eventual consistent.
Well, you can’t stop repeating that in the case of a multimaster, you need to understand very well what you are doing - a huge number of problems appear that simply did not exist with the master-slave.