If you have a second machine, why not just put a Postgres read replica on it? Letting the WAL deal with replica consistency is much simpler than making the client responsible for keeping an external cache in sync, and you get the benefit of keeping everything in Postgres.

Either I pay a performance penalty waiting for my cache entry to be synced to the replica, or I risk reading stale data from the replica, no?