Watch any number that matters — against limits you set.
CPU, temperature, stock price, queue depth, a custom metric you push yourself — Threshold monitoring watches the value against bounds you define, with the same 3-check confirmation as every other check.
Not every problem is an HTTP request.
Sometimes the warning sign is a CPU pegged at 100%, a server room running hot, a queue backing up to thousands of jobs, or an external data feed drifting out of range. Threshold Monitoring gives you the same alerting discipline — interval, 3-check confirmation, channel-of-your-choice — for any value you can name a number on.
A number, a line, an alert.
Any numeric value
Server environmentals like CPU, memory, or temperature. External data like a stock price. A custom metric you push from your own pipeline. Threshold doesn’t care where the number comes from.
Above, below, or outside a range
Define a single ceiling, a single floor, or a window. Anything crossing the line triggers the check.
Configurable interval
Every 1, 5, 10, 15, or 30 minutes. 1-minute on paid plans; the free Specialist tier is 5-minute minimum.
3-check confirmation + alerts
Three consecutive crossings open an incident, three clean readings close it. Email, SMS, Slack, Teams, or webhook.
Roadmap note: baseline-relative thresholds (alert on deviation from a learned baseline) are a planned future addition — fixed bounds ship today.
The evidence is captured at the moment of failure.
When your monitor catches a problem, our global probe network fires automatically on the third failed check — so when your engineer gets the alert, the forensic data is already attached.
Your service
Starts returning errors.
Monitor detects
3 consecutive failed checks.
Global probes fire
40+ locations check at once.
Evidence delivered
Situation report attached to the incident.
The verification runs the moment the issue is happening — not 5 minutes later when you’re trying to reproduce it. A situation report with the regions, routes, and response codes is attached to the incident in your dashboard before your engineer even opens the alert. The first question — “where is this happening?” — is already answered.
Threshold alerts in three steps.
Point us at the value
Connect the numeric source you want tracked — a server environmental, an external data feed, or a custom metric.
Set your bounds
Define the threshold — above, below, or outside a range — that counts as a problem.
Choose alerts & save
Get notified when the value crosses the line, confirmed across 3 checks.
Set the line.
We’ll tell you when it’s crossed.
Free forever tier. Setup in 5 minutes. No credit card.