To ensure that WAL data is available for recovery or emergency restore, we need to secure it on a tertiary location away from the primary or secondary server. We start this by telling PostgreSQL to store the old WAL files instead of deleting them. The cp command we used to copy the files will not overwrite the existing archives due to the -n setting. This prevents accidentally corrupting the existing transaction logs.
Then, we need to create the directory where the files will reside. The mkdir command does this, and the chown command ensures that the PostgreSQL server can write to that directory. Once the directory is in place, we need to reload the server because we changed archive_command.
Once a WAL file is no longer needed ...
