The monitor page
Reading the charts, KPIs, and event timeline on a monitor's own screen.
Every monitor has a page of its own, and it is built to answer the three questions you arrive with: is it up, how has it been, and what exactly happened. This page reads top to bottom in that order.
#The toolbar
The top row holds the target (clickable for HTTP family monitors), the range picker, refresh, the pause toggle, and settings. The range picker drives everything below it: 1 hour shows per-check resolution, longer ranges aggregate by the hour. How far back you can go is your plan's metric retention, 7 days on free, 30 on Pro, 90 on Team.
#The KPI row
Four numbers summarize the range:
- Uptime, as a percentage to two decimals, with a sparkline of how it moved.
- Avg, Latest, and Peak latency. Latest is always the newest check, whatever the range; the other two follow the range.
The word behind the number changes with the type, because the measured time means something different: response time for the HTTP family and schedulers, connect time for ports, resolve time for DNS.
#Charts and the recent-checks bar
The latency chart plots every check in the range as a line, so a slow hour looks like a slow hour and not a smoothed-over average. Below it, the recent-checks bar shows each check as a cell, up, slow, down, and records gaps in place: a maintenance window or a pause reads as a visible hole, not as fake uptime.
#Connection breakdown and expiry checks
HTTP family monitors get two extra panels:
- Connection breakdown splits the last check into reaching the server (DNS, TCP, TLS) versus the server thinking (time to first byte). That split is the diagnosis: connect time up means network or infrastructure, TTFB up means your backend.
- Expiry checks shows the TLS certificate and domain registration rows from SSL and domain expiry; on the free plan they appear locked.
#What each type adds
| Type | Its own section on the page |
|---|---|
| Keyword, API | The content rules and the last check's verdict, rule by rule. |
| Heartbeat | The ping URL card up top, a cadence card (interval, grace, last ping), and On-time / Missed KPIs instead of latency. |
| Port | The TCP endpoint and its connect-only timing. |
| DNS | The query and the answers the last resolve returned. |
| Scheduler | The outbound trigger and its last run. |
A heartbeat has no request latency, so it also skips the latency chart; its cadence card and recent-checks bar carry the story instead.
#The activity timeline
The bottom of the page is the event log: every status transition, when it happened, and why, including which content rule failed and what the body actually contained. When an incident retro asks "what did the monitor see", this is the section that answers.