All solutions
For SRE and platform teams

External monitoring that feeds your on-call, SLOs and change review

Synthetic checks from 12 regions with per-region results, incidents that open and resolve into PagerDuty or Opsgenie with dedup, SLA/SLO reports with error budgets, and a change timeline that records what shifted — DNS, TLS, headers, redirects, routes — before the pager went off.

incidents · live
checkout-api unreachable
downtime · 3 regions · opened 2m ago
Critical
TLS chain changed on cdn.acme.io
intelligence · acknowledged
Warning
db-replica latency recovered
anomaly · resolved automatically
Resolved

Sounds familiar?

  • Internal observability says everything is green while customers in one region can't reach you.
  • Post-incident review needs 'what changed' and the answer is spread across three consoles and a Slack thread.
  • SLO reporting is a spreadsheet someone rebuilds each month.

Which plan

Pro ($85) is the fit: 200 monitors at 15-second intervals from every region, SLA reporting, environments, role-based access and the credential vault. Enterprise adds 10-second checks, SAML and unlimited monitors.

Compare plans →

The setup

  1. 1

    Check from where users are

    Every monitor runs from the regions you pick and reports each one separately, so a regional failure is a regional alert instead of a flap.

    Multi-region checks
  2. 2

    Route incidents through your paging tool

    PagerDuty and Opsgenie destinations dedupe on the monitor and auto-resolve when it recovers; incident.io and Rootly get an incident opened with severity.

    Incident management
  3. 3

    Put error budgets on a report

    SLA/SLO reports compute availability against a target per monitor or group, track the budget and can be scheduled for delivery on Pro.

    SLA reporting
  4. 4

    Keep the change timeline

    Every check fingerprints DNS, TLS, headers, redirects and content; changes become events with before/after that a post-mortem can link to.

    Change detection
  5. 5

    Monitor the paths, not only the endpoints

    Traceroute and ICMP monitors record hop-by-hop latency and loss per region, so transit changes are visible alongside HTTP results.

    Traceroute monitor

Questions from sre & platform teams

How does PagerDuty deduplication work?
Each monitor's alerts share a dedup key, so repeated failures update one PagerDuty alert instead of paging repeatedly, and recovery sends a resolve for that key.
Can I run probes inside our VPC?
The probe agent is the same monitor-worker image as the hosted fleet: run it as a container inside your network and it registers with a one-time token and reports its health to the fleet console. Routing specific monitors to a private agent is the next step on the roadmap — see the docs page on the probe agent for the current state.
Can we manage monitors as code?
Every object is available through the REST API with scoped keys; most teams drive it from CI with a service-account key. The OpenAPI import creates HTTP monitors straight from a spec.