We've encountered another system that has a situation very much like the one described here:
https://www.postgresql.org/message-id/20140904175036.310c6466%40erg
In particular, a secondary is restoring each WAL file using restore_command, creating a .ready file for it, and then
nevercleaning it up. This is PostgreSQL 9.6.6 on Windows using the EDB installer. We can't seem to see anything
unusualabout the configuration of the server (archive_mode = on, but it's otherwise properly operating as a secondary).
--
-- Christophe Pettus
xof@thebuild.com