The preceding solutions may be out-of-reach for individuals or small organizations, or even big ones whose core competency lies elsewhere.
Thus, when seeking to maximize uptime on your own domains, it's all about redundancy.
When the DNS specification and associated RFCs were first published, redundancy was cultivated by simply virtue of having "multiple" (that is, two) nameservers, on disparate networks. To wit, nameservers should be as quoted.
"both topologically and geographically dispersed locations on the Internet, to minimise the likelihood of a single failure disabling all of them."
This is no longer enough. In the modern, commercialized internet of today, each DNS provider or DNS operator must be treated as a logical Single-Point-of-Failure unto itself.
The complexity...