Endpoint Monitoring

Tracevium continuously probes your HTTP endpoints and records response time, status code, and availability.

How checks work

When you add an endpoint, Tracevium enqueues a recurring job that sends an HTTP GET request to the target URL at the configured interval. Each request is called a check. The result — status code, response time in milliseconds, and whether it passed or failed — is stored and used to calculate uptime.

A check is passing when the response status is 2xx and arrives within the timeout. Any other outcome (4xx, 5xx, network error, timeout) is a failure.

Check intervals

Uptime calculation

Uptime is expressed as a percentage over a rolling window (24 h, 7 d, 30 d):

uptime % = (passing checks / total checks) × 100

Checks that occur during a scheduled maintenance window are excluded from uptime.

Response time tracking

Every check records round-trip response time in milliseconds. The endpoint detail page shows a time-series chart so you can spot latency regressions before they cause failures.

Failure sensitivity

A single failed check does not immediately open an incident. Tracevium waits for consecutive failures before triggering an incident, preventing false positives from transient network blips. When the endpoint recovers, the incident resolves automatically and a recovery notification is sent.

What gets monitored

• Status code — must be 2xx for the check to pass.
• Response time — recorded in milliseconds for every check.
• Availability — DNS and connection errors count as failures.
• SSL — expired or invalid certificates cause the check to fail.

Adding an endpoint

Go to your project → Add endpoint. Enter the full URL (must include https://), choose a check interval, and save. Monitoring starts within one interval period.

Pausing and deleting

Paused endpoints do not run checks and are excluded from uptime stats. Deleting an endpoint removes it and all its check history permanently.