FLUX MONITOR
Real-time infrastructure monitoring with live dashboards, threshold alerting, and deep system metrics for your operations team.
// OVERVIEW
Flux Monitor provides continuous health monitoring for all hosts and services in your infrastructure. It collects metrics via SNMP, agent-based checks, and API polling — presenting everything on a unified real-time dashboard.
When thresholds are breached, Flux Monitor automatically creates alerts and (when configured) triggers incident notifications via Flux Notify.
// HOST MONITORING
HOST STATUS STATES
| State | Condition | Action |
|---|---|---|
| OK | All metrics within normal thresholds | No action |
| WARNING | One or more metrics in warning range | Alert generated, logged |
| CRITICAL | One or more metrics in critical range | Alert + optional auto-notify |
| UNKNOWN | Check failed or unreachable | Alert, investigate connectivity |
| MAINTENANCE | Host in maintenance window | Suppress all alerts |
DEFAULT METRIC THRESHOLDS
| Metric | Warning | Critical |
|---|---|---|
| CPU Usage | > 75% | > 90% |
| Memory Usage | > 80% | > 95% |
| Disk Usage | > 80% | > 90% |
| Disk I/O Wait | > 20% | > 40% |
| Network Error Rate | > 1% | > 5% |
| Load Average (1m) | > 2× CPU cores | > 4× CPU cores |
// SERVICE CHECKS
In addition to host-level metrics, Flux Monitor performs active service checks to verify application availability.
| Check Type | Description | Config |
|---|---|---|
| HTTP/HTTPS | URL availability, status code, response time | check_http |
| TCP Port | Port open/closed, connection time | check_tcp |
| ICMP Ping | Host reachability, round-trip time | check_ping |
| MySQL/MariaDB | Connection test, query latency | check_mysql |
| Process | Process running, CPU/memory usage | check_process |
| Custom Script | Any shell script returning 0/1/2/3 | check_custom |
// ALERT RULES
Alert rules define the conditions under which notifications are generated. Rules support:
- Threshold conditions — metric value greater/less than a value
- Duration requirements — condition must persist for N minutes before alerting
- Re-alert intervals — repeat notification every N minutes if unacknowledged
- Recovery notifications — optional alert when condition clears
# Example rule definition
rule_name: "High CPU"
host_pattern: "db-prod-*"
metric: cpu_usage
condition: ">"
warning_threshold: 75
critical_threshold: 90
duration_minutes: 5
re_alert_minutes: 30
auto_notify: true
notify_contact_list: "DBA Team"// FLUX NOTIFY INTEGRATION
When a critical alert is triggered, Flux Monitor can automatically create an incident in Flux Notify and send notifications to the configured contact list.
# Settings → Integrations → Flux Notify
FLUX_NOTIFY_URL=http://flux-notify:8080
FLUX_NOTIFY_API_KEY=your_api_key
AUTO_INCIDENT_SEVERITY=CRITICAL # CRITICAL, HIGH, or both
DEFAULT_CONTACT_LIST=NOC Team// MAINTENANCE WINDOWS
Maintenance windows suppress alerts for scheduled maintenance, preventing false positive notifications during planned downtime.
# Create a maintenance window via API
POST /api/maintenance
{
"name": "DB Patching - Feb 25",
"host_pattern": "db-prod-*",
"start": "2026-02-25T02:00:00Z",
"end": "2026-02-25T06:00:00Z",
"suppress_notifications": true,
"created_by": "j.smith"
}