Endpoint Monitoring

insightd can monitor HTTP/HTTPS endpoints for uptime, response time, and availability — similar to Uptime Kuma but integrated with your container monitoring.

Adding Endpoints

1. Log in to the web UI (click Settings in the sidebar to enter your admin password)
2. Go to Endpoints and click Add Endpoint
3. Fill in the form:

Field	Description	Default
Name	Display name for this endpoint	Required
URL	Full URL to monitor (must start with http:// or https://)	Required
Method	HTTP method	GET
Expected Status	Status code that means “up”	200
Interval	Seconds between checks (10-3600)	60
Timeout	Request timeout in milliseconds (1000-30000)	10000
Headers	Custom headers as JSON (e.g. {"Authorization":"Bearer xxx"})	(none)
Enabled	Whether this endpoint is actively monitored	Yes

How It Works

• The hub probes each endpoint directly (not agents)
• Checks run on a 1-minute scheduler tick; each endpoint’s interval determines when it’s actually due
• Results are stored in the http_checks table (30-day retention)
• The checker respects timeouts and follows redirects

Private Network Endpoints

Since the hub makes HTTP requests directly, it can monitor endpoints on private networks it has access to. This is a major advantage over external monitoring services.

http://192.168.1.50:8080/health
http://my-nas:9090/api/status

Alerts

When an endpoint fails 3 consecutive checks (configurable), an endpoint_down alert fires. The alert resolves automatically when the endpoint recovers.

Configure the threshold via:

• Environment variable: INSIGHTD_ALERT_ENDPOINT_FAILURES=3
• Settings page: Endpoint Failure Threshold

Dashboard

When you have endpoints configured, the dashboard shows an Endpoints Up X/Y stat card. The Endpoints page shows a table with:

• Status (up/down dot)
• 24-hour uptime percentage
• Average response time
• Last check time

Endpoint Detail

Click any endpoint to see:

• Status badge (Up/Down) with color-coded uptime and response time stats
• Check status strip — 24 hours of checks bucketed into ~15-minute slots. Hover any slot to see pass/fail count, average response time, and error details. Red slots indicate at least one failure in that window.
• Response time chart — full TimeSeriesChart with time axis, hover tooltips, and area fill (replaces the old bar chart)
• Failed checks table — only failed checks are shown in detail. If all checks passed, a summary message is shown instead.

Keyboard shortcuts: b to go back, e to edit (when logged in). Press ? for the full shortcut list.

Weekly Digest

Endpoints with less than 99% uptime are flagged in the weekly digest email and webhook notifications.