Alarms & Health¶

OCI Monitoring Alarms¶

Alarm	MQL Query	Severity
High error rate	`app.errors.rate[1m].rate() > 5`	CRITICAL
DB latency	`app.db.latency_ms[1m].percentile(0.95) > 2000`	WARNING
Health down	`app.health[1m].min() < 1`	CRITICAL
CRM sync stale	`app.crm.sync_age_s[5m].max() > 600`	WARNING
Low stock	`app.inventory.low_stock_products[5m].max() > 3`	WARNING

All alarms deliver via OCI Notifications to configured email/webhook.

Health Endpoints¶

Liveness (`/health`)¶

{"status": "ok", "service": "octo-drone-shop"}

Always returns 200. K8s restarts the pod if this fails.

Readiness (`/ready`)¶

{
  "ready": true,
  "database": "connected",
  "db_type": "oracle_atp",
  "apm_configured": true,
  "rum_configured": true,
  "runtime": { "host.name": "...", "process.pid": 1 }
}

Checks DB connectivity. K8s removes the pod from the service if this fails.

OCI Health Checks¶

HTTP monitor on /ready endpoint every 30 seconds. Configured via deploy/oci/ensure_monitoring.sh.

Provisioning¶

COMPARTMENT_ID="<compartment-ocid>" \
SHOP_PUBLIC_URL="https://shop.yourcompany.cloud" \
ALARM_EMAIL="ops@yourcompany.cloud" \
./deploy/oci/ensure_monitoring.sh