A large Australian DNS outage prompted the original article. Its durable point was not that one provider or control panel is always wrong. It was that a business should know who controls the records that connect customers to its website.
Confirm ownership and access
Record the legal registrant, registrar account, DNS provider and hosting provider. Check renewal details and recovery access. Public registration data can help an investigation, but it does not replace the registrar’s authoritative account records.
Keep an approved record of the important DNS entries and why they exist. Unknown records are a reason to investigate with the people responsible, not to delete them during a tidy-up.
Plan the change
Lowering DNS time-to-live values before a planned move can help, but caches and resolver behavior mean a change is not instant everywhere. Preserve mail, verification and service records as well as the website address. Make the rollback criteria explicit.
Measure the actual route
When a site is slow or unavailable, separate DNS lookup time, connection time, server response and page rendering. A waterfall can help locate the delay. One result from one location does not establish that DNS is the cause.
After a change, test the important hostnames, certificates, redirects and customer journeys from more than one network. Monitor the business result as well as crawler reports.
See our domain migration guide, hosting and SEO checks and SEO services.