Running Pg 9.2 on CentOS. Currently have a standard primary/standby setup, with the standby in a different DC across the country from the primary. These are all on VMs.
We want to add another standby in the same standby DC. What I'd like to do is have the current standby become a cascading standby, and restrict user queries to the new hot standby (and/or any others we hang off of it). But the soon-to-be-cascading standby would be strictly for DR and cascading WALs.
The docs and articles I've read seem to focus heavily on the streaming replication aspect of things. But I imagine we want to also have WAL files getting copied downstream as well. I'm writing to ask how the pros from Dover handle this. Custom script in place of archive_cleanup_command on the cascading standby to push the archived files downstream before deleting them? Or some rsync job on the cascading standby running every minute to push changes downstream?