June 15, 2026
Domain of the Dead
The ghost domain problem in DNS, and what we're doing about it
Your site can look alive after it’s gone — and commenters are fighting over whose fault that is
TLDR: A website name can be removed from the official system yet still look alive to monitoring tools because old internet records hang around. Commenters were split between “spooky but important” and “come on, this is just stale cache,” turning a niche bug into a blame-filled mini-brawl.
A delightfully creepy internet glitch just dropped: a website name can be removed from official records and still look perfectly fine to some monitoring tools for days. That’s the “ghost domain” problem, and the company behind the post says it happens because old address info can linger in the internet’s memory, making a dead site appear healthy when it’s actually gone. In plain English: the lights are on, the house is demolished, and your status page is still smiling.
But the real show is in the comments, where the crowd instantly split into camps. One side basically yelled, “This isn’t some spooky mystery of the internet — you were checking the wrong place!” That was the energy from quuxplusone, who argued the service had simply learned the hard way that stale cached data can hide outages until the old info expires. Another commenter, winstonwinston, went full galaxy-brain with the deadpan line that every domain is a ghost domain until the timer runs out, which is exactly the kind of nerd joke that gets respectful snorts online.
Then came the practical fixers: johnhtodd dropped a research link and suggested blocking names that have already been pulled at the registry level. So the vibe was equal parts “wow, scary bug”, “actually this is old news”, and “please don’t let uptime dashboards gaslight us.” In other words: classic internet drama, with a side of haunted websites.
Key Points
- •The article describes the ghost domain problem, where a domain removed from a parent zone can still appear healthy to uptime monitors because recursive resolvers retain cached delegation data.
- •It cites multiple registry and registrar workflows, including DENIC (.de), EURid (.eu), AFNIC (.fr), and ICANN-related gTLD enforcement, as ways domains can be removed from zone publication.
- •The article says the observed difference is caused by resolver cache state: cold caches may return NXDOMAIN, while warm recursive caches can continue resolving the domain.
- •In the example setup, the publisher traced requests through PHP/curl, glibc nsswitch, systemd-resolved, and an upstream shared recursive resolver, concluding the key issue resides in upstream recursive caching.
- •The mechanism is presented as documented in prior research and active standards work, including the 2012 and 2023 ghost-domain papers, an IETF draft on delegation revalidation, RFC 2181, and RFC 9499.