Back to Insights
Honest Uptime Monitoring: Real Percentages, Incident Timelines & Instant Alerts

Honest Uptime Monitoring: Real Percentages, Incident Timelines & Instant Alerts

Product Updates

There is a dirty little habit in uptime monitoring: rounding up. A tool runs a handful of checks, sees nothing obviously broken, and proudly prints 100%. It feels great until the day a customer emails to say your site was down for six minutes — minutes your dashboard swears never happened. We just shipped a thorough rebuild of ViewCel Uptime Monitoring, and it is built on the opposite instinct: show the real number, even when it is not pretty.

What changed

The new in-depth Uptime section on the Features page lays it out in full, but here is the short version of what now backs every uptime figure in ViewCel:

  • Continuous health checks on intervals as tight as every minute, around the clock.
  • Distributed monitoring from independent worker infrastructure, so a single bad route never fakes an outage.
  • Response-time tracking on every check, so you watch slowdowns build long before they become downtime.
  • An incident timeline with start, end, and duration for every outage — a history, not a single blurry number.
  • Honest uptime math: real percentages from real checks. We show 99.24% when it is 99.24%.
  • Instant down alerts by email and webhook the moment a check fails, and again when service recovers.
  • SSL & response validation that catches certificate problems and unexpected status codes, not just hard failures.
  • Confirmation thresholds so a single flaky request never wakes you at 3 a.m.

Why we stopped rounding up

Recently we dug into a target that proudly reported a flat 100% uptime — and discovered the number was a lie of omission. The underlying query had quietly degraded, and a stale calculation was papering over real failures. When we fixed it, the honest figure came back: 99.24% over 30 days. That is a perfectly good number. It is also a true one, and a true 99.24% is worth far more than a comfortable, fictional 100%.

incident logged Every check recorded — including the bad one
Response time is captured on every check. The dip is logged as a real incident, not smoothed away.

Reliability you can actually act on

Honest data is only useful if it reaches you in time. The moment a check fails its confirmation threshold, ViewCel fires an alert by email and webhook — and follows up when the service recovers, with the incident duration attached. You get the full story: what went down, for how long, and how slow things were getting before they did.

See the complete capability list in the new Uptime Monitoring section on the Features page.

Back to Insights