What HTTP failures actually mean for monitoring
How to read status codes, timeouts and TLS errors in monitoring results, and what each one does and does not tell you.
By the UptimeMonitor360 team · updated
Status classes
2xx means the server produced a response it considers successful. 3xx is a redirect; monitors follow up to five and validate each destination. 4xx means the request was rejected (401 and 403 often mean credentials or IP allow-lists; 404 means the path is wrong or the deployment changed). 5xx means the server failed.
Timeouts
A connect timeout means nothing accepted the TCP connection in time: firewalls, DNS pointing at the wrong address, or an overloaded load balancer. A total deadline exceeded after connecting means the application was slow to respond or to finish sending.
TLS errors
Expired certificates, hostname mismatches and untrusted chains are failures because browsers treat them as failures. A missing intermediate certificate is the most common cause of a chain error.
What a 200 does not prove
A 200 proves the server answered. It does not prove a checkout works, a database is healthy or a page renders. Add keyword or JSON assertions on a health endpoint that checks dependencies.
HEAD versus GET
Some servers answer HEAD differently from GET or not at all. Monitor the method your users actually use.
Monitor the behaviour described here continuously: start free with 5 monitors or try the free tools.