Bootstrapping the target cluster

Another critical component of a major PostgreSQL upgrade using pglogical is that we need somewhere to send the data. Ideally, this is a cluster comprised of the same number of nodes, with identical hardware or better, and all the same supporting software, abstraction layers, and automation.

The goal of this recipe is to produce a target cluster using the newer PostgreSQL version so we can subscribe to the old one.

Getting ready

Though we'll technically be starting the new cluster from scratch, be sure to have completed the Creating a publication set recipe so we know pglogical is ready to transmit data to the new cluster we build.

Additionally, we will need to know the path to the initdb...