Once repmgr is in charge of a fully online PostgreSQL cluster consisting of multiple nodes, we need to account for server maintenance. Whether we need to apply a security patch to PostgreSQL or replace server hardware, it is inevitable that we will perform at least one switchover over the lifetime of the cluster.
What else must we consider when performing maintenance on a PostgreSQL cluster managed by repmgr? What about upgrading repmgr itself? This recipe will answer those questions to round off the chapter.
