Skip to main content

​
GET /health

Comprehensive health check endpoint that verifies API and InfluxDB connectivity.

​
Response

​
status
string
System health status: healthy, degraded, or unhealthy
​
database
string
Database status: connected, mock, or unavailable
​
message
string
Human-readable status description

​
Example

cURL
curl https://predictive-maintenance-uhlb.onrender.com/health
Response (Healthy)
{
  "status": "healthy",
  "database": "connected",
  "message": "All systems operational. InfluxDB connected."
}
Response (Mock Mode)
{
  "status": "healthy",
  "database": "mock",
  "message": "Running in mock mode. Data persisted to console. Set INFLUX_TOKEN for full persistence."
}
Response (Degraded)
{
  "status": "degraded",
  "database": "unavailable",
  "message": "Running in degraded mode. Demo features available via /system/* endpoints."
}

​
Status Codes

  • 200 - Health check succeeded (even in mock/degraded mode)
  • 503 - Critical failure (rare, only if API itself is unresponsive)
The /health endpoint returns 200 even in mock mode. This is intentional - the system uses graceful degradation to allow demo/testing without a live InfluxDB instance.

​
GET /ping

Lightweight heartbeat endpoint with no database or ML dependencies. Used by the frontend’s keep-alive mechanism to prevent Render free-tier cold starts.

​
Response

​
status
string
Always returns ok

​
Example

cURL
curl https://predictive-maintenance-uhlb.onrender.com/ping
Response
{
  "status": "ok"
}

​
Status Codes

  • 200 - Pong received

​
Keep-Alive Strategy

The frontend dashboard sends a /ping request every 10 minutes to keep the Render backend warm: 
// Frontend keep-alive implementation
setInterval(() => {
  fetch(`${API_URL}/ping`).catch(() => {});
}, 10 * 60 * 1000); // 10 minutes
This prevents the free-tier backend from spinning down after 15 minutes of inactivity.
Use /ping for uptime monitoring and load balancer health checks. It has minimal overhead (1ms) and doesn’t touch the database.

​
Comparison

EndpointDatabase CheckML CheckUse CaseLatency
/health✅ Yes❌ NoDeployment verification, admin dashboards~50-200ms
/ping❌ No❌ NoKeep-alive, uptime monitoring1ms
Use /health for comprehensive status checks before deployments. Use /ping for lightweight heartbeats.

Build docs developers (and LLMs) love

Get started for freeTalk to us