How to Monitor Website Downtime
Set up meaningful checks, verify notifications and use incident evidence to troubleshoot outages.
Login2Cloud product team · Published 2026-09-27
Start with the customer-facing URL
Use the final HTTPS address visitors should reach. Confirm the expected status range; a redirect may be normal for a homepage, but it can hide a login error on an API. For a page with a stable marker, require meaningful text such as a checkout heading. Avoid timestamps or personalized text that changes on each visit.
Choose interval and timeout together
The interval controls when checks start; the timeout caps how long a check waits. A long timeout is not a substitute for fixing a slow endpoint. Confirmation reduces alerts from brief failures but adds detection delay. In Login2Cloud the default incident policy requires consecutive failures; a single failed measurement is visible before an incident is confirmed.
Prove the alert path
Create an email or signed-webhook channel and choose outage and recovery events. Use Send test notification where the destination is eligible. A queued test is not yet delivered: inspect its delivery status and confirm receipt. Maintain more than one useful recipient when your operating model requires it.
Respond using the evidence
Check whether the failure is DNS, connection, TLS, timeout, HTTP status or a content assertion. Compare total duration with connection and first-byte timings. Review deployment logs in your own environment. An external monitor provides symptoms and timing; it cannot automatically establish the root cause inside your application.
Communicate and close the loop
Schedule planned maintenance before work begins. Monitoring history continues while applicable notifications are suppressed. Publish a concise status update describing impact and the next step, then confirm recovery with successful checks. Manual communication should never rewrite measured availability.
Common questions
Should I alert on every failed check?
A confirmation policy can reduce noise. Choose a delay that fits the service and make that delay clear to the team.
Does maintenance erase downtime?
No. Checks and history remain available; maintenance controls notification behavior.