Skip to main content
FeaturesMonitoring
Monitoring

Every protocol that matters. One scheduler. One timeline.

From HTTP endpoints to gRPC services, from browser journeys to database probes — the same retry, backoff and SSL side-channel logic runs every check.

14
Monitor types shipped
100
Default concurrent workers
1 file
To add a new monitor type
Multi-region
Global PoPsRoadmap

HTTP / HTTPS endpoints

Status code matching, latency thresholds, body assertions, header expectations and follow-redirect rules — with full request/response capture on failure.

  • Status-code allowlists with multi-code support (e.g. 200, 201, 204)
  • Latency assertion with per-monitor SLO threshold
  • Response-body substring and regex assertions
  • Request header injection and Bearer-token templating
  • Configurable timeout, follow-redirects and TLS-skip flag
  • Per-check screenshot of response body for failure forensics
json
{
  "name": "checkout-api",
  "type": "http",
  "url": "https://api.acme.io/v1/health",
  "interval": 60,
  "assertions": [
    { "type": "status", "in": [200, 204] },
    { "type": "latency_ms", "lt": 800 },
    { "type": "body", "contains": "ok" }
  ]
}

TCP, ICMP & DNS

Reach the things HTTP can't — internal services, hosts behind L4 load balancers, resolver health.

  • TCP port-open checks with optional banner match
  • ICMP ping with packet-loss and median-latency thresholds
  • DNS resolution against A, AAAA, CNAME, MX, NS, TXT records
  • DNS answer assertion (e.g. A record must equal a specific IP)
  • Resolver override for split-horizon DNS validation

SSL certificate expiry

Every HTTPS check opportunistically grabs cert details and upserts them — so the cert inventory is always live, not a sidecar product.

  • Subject, SAN, issuer and serial captured per check
  • Expiry alerts at 30/14/7/1 day thresholds
  • Mixed-host inventory across services, automatically deduped
  • Self-signed and chain-incomplete detection
  • Standalone SSL-only monitor for hosts without HTTP checks

Database probes

Real driver-level connections, not synthetic TCP. We open the protocol, run a probe query, close cleanly — exactly like your app does.

  • PostgreSQL — connect + `SELECT 1` with timeout
  • MySQL — protocol handshake + ping
  • MongoDB — replica-set aware `serverStatus`
  • Redis — `PING` with auth and TLS support
  • Credentials encrypted at rest (AES-256-GCM)

Protocol checks (SMTP / IMAP / FTP / SSH)

Mail, file transfer and admin endpoints deserve the same first-class treatment as web traffic.

  • SMTP banner + HELO/STARTTLS handshake
  • IMAP login + CAPABILITY assertion
  • FTP / FTPS login + directory listing probe
  • SSH host-key verification with optional command execution

gRPC & WebSocket

Modern protocols, monitored modernly — health-check service for gRPC, message echo for WebSocket.

  • gRPC `grpc.health.v1.Health/Check` against any service descriptor
  • TLS, mTLS and metadata-header support
  • WebSocket handshake, message echo and latency
  • Per-frame body assertion on the WebSocket response

Heartbeat / cron-job monitoring

Inverse monitoring for jobs that should phone home. Miss a window and we alert — even if the job never started.

  • Custom expected interval (e.g. every 30 minutes)
  • Configurable grace window before declaring miss
  • Stateless ping URL — `curl` from cron, no SDK required
  • Per-heartbeat alert routing — different jobs, different oncall
  • Last-seen and miss-streak surfaced on the dashboard
bash
# In your crontab:
*/30 * * * * /usr/local/bin/nightly-backup.sh \
  && curl -fsS https://hb.alertifypro.com/h/abc123 \
  > /dev/null

Synthetic browser journeys

Headless Chromium runs the steps your users would — login, navigate, assert. Failures capture a screenshot and DOM snapshot.

  • Step-by-step scripting: click, type, wait-for-selector, assert
  • Per-step latency and full journey runtime
  • Screenshot on failure with DOM snapshot
  • Configurable viewport and user-agent
  • Stored credentials are encrypted (AES-256-GCM)
  • Multi-region browser execution from a global PoP networkRoadmap

Custom metrics ingest

Push any number from any system; we'll baseline, alert and chart it. Closes the loop on the metrics monitoring tools never quite covered.

  • Define a metric — name, unit, expected range
  • Push values via authenticated API or webhook
  • Tagged dimensions for per-segment alerting
  • Statistical alerting integrates automatically
  • OpenTelemetry-native ingest pathRoadmap

Statistical anomaly detection

Rolling baselines per service with z-score scoring — catch the regressions a static threshold would miss.

  • Per-service latency baseline with 7-day rolling window
  • Configurable z-score sensitivity
  • Warning-level alerts (won't page primary on-call)
  • Baseline reset on legitimate latency shifts
  • Anomaly events captured in the incident timeline

Region toggles & multi-region execution

Pick which regions a check runs from. Per-service overrides today; full PoP-mesh execution is on the roadmap.

  • Per-service region allowlist UI
  • Region preference saved with the monitor definition
  • Multi-region check execution from a global PoP networkRoadmap
  • Region-of-failure attribution in the incident timelineRoadmap
Monitoring vs the market

How we stack up on monitoring

Only rows relevant to this capability. See the full comparison →

FeatureAlertifyProDatadogUptimeRobotPingdom
14 monitor types (incl. gRPC, WebSocket, browser)
Database probes (Postgres, MySQL, MongoDB, Redis)
Heartbeat / cron-job monitoring
Statistical anomaly detection
Custom metrics ingest
Multi-region global probesRoadmap

Frequently asked

Ready to see monitoring on your stack?

Spin up your first monitor in under a minute. Free forever for the first 5 services.