The flow continuously monitors system and NiFi metrics (CPU, RAM, Threads, Disk usage, etc.), evaluates them against predefined thresholds, and triggers alerts via HTTP and Email when anomalies are detected.
The alerting pipeline is built around the following logical stages:
- Flow Triggering
- Metrics Collection
- Metrics Extraction
- Threshold Evaluation
- Routing
- Notification
- GenerateFlowFile
Acts as a scheduler to trigger the monitoring flow periodically.
- InvokeHTTP
Metrics Collection
Calls a monitoring endpoint (NiFi API / system metrics API).
GET /nifi-api/system-diagnostics
- EvaluateJsonPath
Extracts metrics from the JSON response.
- UpdateAttribute
Normalizes metrics and defines the necessary attributes, such as metric thresholds for alerts.
- RouteOnAttribute
Alert Detection : Compares metrics against thresholds and routes FlowFiles.
- ReplaceText
Alert Message Formatting : Builds a clean and readable alert message.
- PutEmail
Sends alert notifications via email.
- InvokeHTTP (Webhook / API)
Sends alerts to external systems (Slack, Teams, Monitoring API, etc.).
In this sub-process group, we use the same processors; the only difference is the API endpoint used to extract different NiFi metrics.
- InvokeHTTP
Metrics Collection
Calls a monitoring endpoint (NiFi API / system metrics API).
GET /nifi-api/flow/status
- EvaluateJsonPath
Extracts metrics from the JSON response.
- CPU metric alert
- Threads metric alert
- Que FlowFile metric alert
- CPU metric alert
- Threads metric alert
- Que FlowFile metric alert
- Apache NiFi 1.22
- Access to metrics endpoint
- SMTP server (for email alerts)
- Optional webhook endpoint