Monitor types and behavior
Skip to Content
Sraraa docs

Monitor types and behavior

How each Sraraa check works and what it needs to connect.

Choose the monitor type that matches the failure you need to see. Each section lists what the check observes and what to configure in the dashboard.

HTTP and website checks

Use for public routes where response status, timing and redirects matter.

  • Enter the public https:// URL.
  • Set the expected status code, timeout and whether redirects should be followed.
  • Run the first check and confirm the result timestamp before enabling alerts.

Keyword checks

Use when a page can return 200 with the wrong content.

  • Enter the page URL and the exact keyword or phrase.
  • Choose whether the keyword must be present or absent.
  • Verify against the deployed page, not a draft or staging copy.

API contract checks

Use when the response shape matters as much as the status code.

  • Enter the API endpoint and any required authentication.
  • Add JSON assertions: required keys, expected type or a value at a path.
  • Test against a known-good response before relying on the assertion.

Authenticated requests

Use the workspace Secrets Store for credentials; do not paste secrets into the URL or public fields.

  • Create a secret in the workspace Secrets Store.
  • Reference it from request headers or body as ${secret:NAME}.
  • Confirm the result metadata does not contain the secret value.

TCP port checks

Use for databases, SSH, message brokers and other non-HTTP services.

  • Enter the target as host:port.
  • Allow the Sraraa probe egress range through the target firewall.
  • Confirm the port is reachable from outside your private network.

ICMP ping checks

Use for network-layer reachability and round-trip time.

  • Enter the hostname or public IP address.
  • Confirm the target allows ICMP echo requests.
  • Use a TCP monitor when the provider blocks ICMP.

UDP checks

Use for services that answer over UDP.

  • Enter the target as host:port.
  • Set the payload and, when required, the expected response.
  • Allow UDP traffic from the probe through the firewall.

DNS checks

Use to catch wrong answers, missing records and unintended DNS changes.

  • Enter the hostname and select the record type.
  • Set the expected value when the answer must match exactly.
  • Choose a resolver only when you specifically need that network path.

TLS certificate checks

Use to catch expiry and chain problems before browsers do.

  • Enter the public hostname or HTTPS endpoint.
  • Confirm the certificate is served for the hostname clients use.
  • Set the expiry warning threshold and verify the first result.

Domain and WHOIS checks

Use to track registrar expiry and public registration data.

  • Enter the registered domain name without a path.
  • Confirm the registrar responds to the first lookup.
  • Use the expiry date in the result to plan renewal.

Page speed checks

Use to measure DNS, TLS, TTFB and total page time.

  • Enter the public page URL.
  • Set the latency budget the check should enforce.
  • Compare results from the same probe path over time.

Synthetic checks

Use for multi-step flows such as sign-in or checkout.

  • Enter the starting URL and define each step.
  • Keep credentials in the Secrets Store.
  • Verify the full flow against a staging account before enabling alerts.

Cron heartbeats

Use when scheduled work has no external request to monitor.

  • Create the heartbeat monitor and copy its unique URL from the create result.
  • Call that URL after the job completes successfully.
  • If the URL is lost, rotate it in monitor Settings → Heartbeat credential.

Webhook push

Use when an external service already knows its own status.

  • Create the webhook monitor and copy its unique endpoint.
  • Send a POST request with JSON: {"status":"up"} or {"status":"down","message":"..."}.
  • If the endpoint is exposed, rotate it in monitor Settings → Heartbeat credential.

Push and Docker agents

Use an agent when the target is behind a firewall or must report host metrics.

  • Create the push or Docker/host monitor and copy its unique report endpoint.
  • Configure the agent with that endpoint and the workspace token.
  • Confirm the first report appears before relying on the agent.

Plan-dependent checks

API contract, page speed, synthetic, agent and advanced assertion checks depend on the workspace plan. The dashboard and plan catalogue show what the current workspace can use.