Developer Tool · Monitoring

Alert Rules

Configure email or webhook notifications when nodes experience high latency, disconnects, or drop below compute throughput thresholds.

Intelligent Alert Suppression

Correlated Incidents

Overlapping anomalies grouped by root cause

No active incidents

All systems operating normally

Alert Suppression: Related anomalies detected within 30 minutes are correlated into incidents. Individual notifications are suppressed to reduce noise. Correlation uses temporal proximity, metric type, affected nodes, and anomaly patterns.

Resource Utilization Thresholds

Utilization Thresholds

Set CPU and memory utilization thresholds for each cluster environment. Alerts will trigger when utilization exceeds these limits.

production Environment

Strict

CPU Threshold75%

1%99%

Memory Threshold80%

1%99%

Alerts will trigger when CPU or memory exceeds 75% and 80% respectively.

staging Environment

Balanced

CPU Threshold80%

1%99%

Memory Threshold85%

1%99%

Alerts will trigger when CPU or memory exceeds 80% and 85% respectively.

development Environment

Relaxed

CPU Threshold85%

1%99%

Memory Threshold90%

1%99%

Alerts will trigger when CPU or memory exceeds 85% and 90% respectively.

📋 Recommended Thresholds

Production: 75% CPU, 80% Memory — strict limits to prevent cascading failures
Staging: 80% CPU, 85% Memory — balanced for realistic testing
Development: 85% CPU, 90% Memory — relaxed for rapid iteration

Webhook Payload Format

POST application/json

{
  "event":      "meshinfer.alert.triggered",
  "rule_id":    "rule_abc123",
  "rule_name":  "High Latency Alert",
  "trigger":    "high_latency",
  "threshold":  500,
  "observed":   731,
  "node_id":    "node-014",
  "ts":         "2026-04-20T14:22:03.112Z",
  "signature":  "sha256=hmac_hex..."   // present if signing secret set
}

Active Rules (0)

No alert rules yet

Create your first rule to get notified when node health degrades.