Integration · Developers

Your monitors in Prometheus

Scrape one endpoint with an API key and get every monitor’s state, response time and 24-hour uptime as Prometheus metrics, next to the rest of your infrastructure.

14-day free trial · Then free on 20 monitors · No credit card

What you get

  • Every monitor, one scrape

    Up or down, paused, the last response time, when it was last checked and the share of the last 24 hours it was up, for every monitor in the workspace.

  • Labels to filter on

    Each series carries monitor_id, name, type and project, so you can group by client project or monitor type in PromQL.

  • Rules in your own stack

    Write alert rules on uptimecheck_monitor_up or uptimecheck_open_incidents next to the ones you already run. Uptimecheck’s own alerts keep working as before.

  • Charts in Grafana

    With Prometheus as a data source, chart response times and uptime in Grafana next to your own metrics.

How to connect Prometheus

  1. Create an API key under Settings › API. Read only is enough.

  2. Add a scrape job to prometheus.yml: target app.uptimecheck.com, scheme https, metrics path /api/v1/metrics and the key as the bearer token.

  3. Set scrape_interval to 60s or slower, reload Prometheus and check that the new target is up.

The endpoint is described in the API reference.

On Professional and Business, with an API key from Settings › API. A read-only key is enough.

What it measures

uptimecheck_monitor_up
1 when the last check passed, 0 when it failed
uptimecheck_monitor_paused
1 while the monitor is paused
uptimecheck_monitor_response_ms
The last check’s response time in milliseconds
uptimecheck_monitor_last_check_timestamp_seconds
When the monitor was last checked, as a Unix time
uptimecheck_monitor_uptime_ratio
Share of the last 24 hours it was up, from 0 to 1
uptimecheck_open_incidents
Incidents open right now in the workspace

prometheus.yml

scrape_configs:
  - job_name: uptimecheck
    scrape_interval: 60s
    scheme: https
    metrics_path: /api/v1/metrics
    authorization:
      type: Bearer
      credentials: uc_your_api_key
    static_configs:
      - targets: ["app.uptimecheck.com"]

GET /api/v1/metrics

# HELP uptimecheck_monitor_up Whether the monitor's last check passed (1) or failed (0).
# TYPE uptimecheck_monitor_up gauge
uptimecheck_monitor_up{monitor_id="mon_9xk2m1q8w4",name="Checkout",type="http",project="Northwind"} 0
uptimecheck_monitor_up{monitor_id="mon_4hd8s2k7p1",name="Storefront",type="http",project="Northwind"} 1
# HELP uptimecheck_monitor_response_ms Response time of the last check, in milliseconds.
# TYPE uptimecheck_monitor_response_ms gauge
uptimecheck_monitor_response_ms{monitor_id="mon_4hd8s2k7p1",name="Storefront",type="http",project="Northwind"} 184
# HELP uptimecheck_monitor_uptime_ratio Share of the last 24 hours the monitor was up (0 to 1), over the time it was monitored.
# TYPE uptimecheck_monitor_uptime_ratio gauge
uptimecheck_monitor_uptime_ratio{monitor_id="mon_9xk2m1q8w4",name="Checkout",type="http",project="Northwind"} 0.995139
uptimecheck_monitor_uptime_ratio{monitor_id="mon_4hd8s2k7p1",name="Storefront",type="http",project="Northwind"} 1
# HELP uptimecheck_open_incidents Incidents open right now in the workspace.
# TYPE uptimecheck_open_incidents gauge
uptimecheck_open_incidents 1
A scrape job and part of what it gets back, in the Prometheus text format. Every monitor in the workspace has its own series.

Questions

Which metrics are there?

uptimecheck_monitor_up (1 or 0), uptimecheck_monitor_paused, uptimecheck_monitor_response_ms, uptimecheck_monitor_last_check_timestamp_seconds, uptimecheck_monitor_uptime_ratio over the last 24 hours, and uptimecheck_open_incidents for the workspace. Monitor metrics are labelled monitor_id, name, type and project.

How often should Prometheus scrape?

Every 60 seconds or slower. That stays inside the API’s rate limit of 120 requests a minute per key, which the key’s other uses share.

Why is a monitor missing from uptimecheck_monitor_up?

Paused monitors and monitors that haven’t been checked yet are left out of uptimecheck_monitor_up and the response time. A paused monitor shows 1 in uptimecheck_monitor_paused.

Which plans include it?

Professional and Business, like the REST API. Any API key works, a read-only one included.

Can I chart it in Grafana?

Yes, through Prometheus: add your Prometheus as a data source in Grafana and chart the uptimecheck_ metrics like any other.

Hear it from us, not your customers.

Click around the live demo without an account, or start your trial and add your first monitor.

14-day free trial · Then free on 20 monitors · No credit card