PointWardenMonitoring console

Know when availability changes.

PointWarden checks public HTTP Endpoints, confirms Incidents from evidence, and notifies your Workspace when the observed state changes.

Start monitoringSee how it works
Observed evidenceAvailability claim
Time (UTC)EventStatusDetails
  1. Passing CheckAvailableGET /health · 200 · 112 ms
  2. Passing CheckAvailableGET /health · 200 · 98 ms
  3. Passing CheckAvailableGET /health · 200 · 121 ms
  4. Failing CheckSuspectedGET /health · timeout
  5. Failing CheckSuspected2 failures · confirming
  6. Incident openedUnavailable3 failures · threshold met
  7. Passing CheckRecoveringGET /health · 200 · 119 ms
  8. Recovery confirmedAvailable2 successes · recovered
  9. Passing CheckAvailableGET /health · 200 · 102 ms

3 failures confirm

Three consecutive Failing Checks open an Incident.

2 passes recover

Two consecutive Passing Checks confirm Recovery.

Manual Checks never change Availability.

Only consecutive Scheduled Checks can change the state.

One evidence chain, three decisions.

PointWarden turns what a Scheduled Check observed into a confirmed Availability change, then carries the same evidence into Delivery.

  1. Observe

    Run Scheduled Checks against public HTTP Endpoints.

    Example Check EvidencePassing
    Endpoint
    https://api.example.com/health
    Observed
    18 May 2024 · 10:18:12 UTC
    Final status
    200 · 112 ms · 0 redirects
    Monitor Revision
    14

    Checks run from the public internet against the Endpoint you configured, so the evidence reflects real reachability.

    Scheduled Checks can affect Availability. Manual Checks cannot.

  2. Confirm

    Use consecutive Check Evidence before changing Availability.

    Confirmation sequenceConsecutive evidence
    1. Failing Checktimeout
    2. Failing Check2 failures · confirming
    3. Incident opened3 failures · threshold met
    4. Passing Check1 success · recovering
    5. Recovery confirmed2 successes · recovered

    The first failure starts observed impact. The third consecutive failure confirms an Incident. Recovery follows the same careful pattern in reverse.

    Single failures never flip the state.

  3. Notify

    Deliver Incident and Recovery events to verified Notifiers.

    NotifierDelivery health
    Generic
    Healthy
    Slack
    Delivering
    Discord
    Retrying
    Microsoft Teams
    No Delivery Data

    A successful Test Delivery verifies a Notifier. Enabled decides whether it can receive Incident and Recovery events. Delivery health shows what happened afterward.

    Verification and enablement are separate decisions.

Evidence without exposure.

PointWarden retains the minimum facts needed to explain Availability, while leaving sensitive response material alone.

Retained to explain Availability

  • Timestamp, Monitor Revision, and outcome
  • Final status when one is available
  • Latency to final response headers and redirect count
  • A sanitized failure category

Never retained

  • Request or response headers and bodies
  • Redirect URLs, query strings, or secrets
  • Cookies or authorization credentials
  • Certificates, IP addresses, or server banners

Delivery proof, not just configuration.

Follow the evidence from a public Endpoint to a confirmed event your Workspace can act on.

Start monitoring