Error messages name a symptom, not a cause, and the same wording covers several unrelated faults. The first question worth answering is always whether the problem is on your machine or on the server, because the fixes have nothing in common. The guides in this section take one error message each and answer that question first.
Which guide you need follows from what the browser told you. If the address will not resolve at all, the lookup returned no such name and you want the NXDOMAIN guide. If the browser sits there and eventually gives up, the resolver never answered and that is a different fault with different fixes. If pages load but stall for a second first, the lookup is slow rather than broken, which is a third diagnosis again. If the server answered and refused, you are looking at an access rule, not a network problem. And if a scheduled job ran at the wrong hour, the five fields of the cron expression are usually where the mistake is.
Each guide is built the same way: what the message means, the named causes, the fix as numbered steps with the exact command for Windows, macOS, Linux, the router or the browser, and then a section the rest of the internet leaves out. Running the check from your own machine tells you what one network sees at one moment. Running it from several countries tells you whether anyone else has the problem, and that is usually the difference between an afternoon of flushing caches and a five minute fix in a DNS panel.
HostTracker has been monitoring websites since 2004 and now watches over 500,000 websites from 300+ checkpoints in 158 cities, with alerts by email, SMS, voice call, Slack, Telegram and more. To see what a server is returning right now, from locations that are not yours, run the free HTTP check.