I'm stumped.Using this command to set up the slave and replication on PG v 15:
You must, by some means:
Have PostgreSQL binaries installed on the machine that is to become the replica.
Have the output of pg_basebackup present on the replica; removing the cluster that is initially installed and replacing it with the backup tends to be easiest.
The means by which you run pg_basebackup aren’t all that interesting since there is this thing called copy file.
The output of pg_basebackup is the data directory for the cluster so of course the cluster cannot be running until you’ve put the contents into place and ensured the proper configuration and added the standby.signal file. Admittedly, if you choose to destroy an existing cluster in order to install the new one in its location of course you’ll need to shut down that existing cluster before destroying it. But this later point isn’t going to be documented because it isn’t an inherent task but rather one you choose as a convenience due to using a packager that handles most of cluster configuration for you and a lack of desire to configure the replica cluster from scratch.
David J.
