8d89fe80ff
- Archive: 85 agent .txt files → docs/archive/agents/legacy_txt/ - Scripts: Move 110 shell scripts → scripts/ (keep deploy.sh in root) - Models: Move 18 .safetensors → ml/models/checkpoints/training_artifacts/ - Delete: 34 directories (~33GB freed) - target/, coverage_*, test artifacts - Build: Clean 14 build artifacts (.rlib, .o, .pid, binaries) - Tests: Move 14 .rs files → tests/standalone/ - SQL: Move 5 files → sql/ (keep init-db*.sql for Docker) - Wave 153: Archive to docs/archive/historical/wave153/ - Docs: Archive 9 markdown files to wave_d/reports/ and historical/ Total impact: ~34GB freed (both waves), root directory cleaned from 583 to ~40 essential files Directory count reduced from 65 to 31 (52% reduction) All historical data preserved in organized archive structure
93 lines
3.5 KiB
Bash
Executable File
93 lines
3.5 KiB
Bash
Executable File
#!/bin/bash
|
|
# Alert Testing Framework for Wave 75 Agent 8
|
|
|
|
PROMETHEUS_URL="http://localhost:9099"
|
|
ALERTMANAGER_URL="http://localhost:9093"
|
|
|
|
echo "======================================================================="
|
|
echo " Foxhunt Alert Testing Framework"
|
|
echo " Wave 75 Agent 8: Alert Testing and Validation"
|
|
echo "======================================================================="
|
|
echo ""
|
|
|
|
# Test 1: Connectivity
|
|
echo "=== Test 1: Infrastructure Connectivity ==="
|
|
curl -s -m 2 "${PROMETHEUS_URL}/api/v1/status/config" > /dev/null 2>&1 && echo "✅ Prometheus connected" || echo "❌ Prometheus connection failed"
|
|
curl -s -m 2 "${ALERTMANAGER_URL}/api/v2/status" > /dev/null 2>&1 && echo "✅ AlertManager connected" || echo "❌ AlertManager connection failed"
|
|
echo ""
|
|
|
|
# Test 2: Alert Rules Count
|
|
echo "=== Test 2: Alert Rules Loaded ==="
|
|
curl -s -m 5 "${PROMETHEUS_URL}/api/v1/rules" 2>/dev/null | \
|
|
jq -r '.data.groups[] | "\(.name): \(.rules | length) rules"'
|
|
echo ""
|
|
|
|
# Test 3: Individual Alert Rules
|
|
echo "=== Test 3: Verify 13 Expected Alerts ==="
|
|
ALERT_COUNT=0
|
|
for alert in AuthLatencySLAViolation HighAuthFailureRate RedisConnectionFailure \
|
|
RevocationCacheSizeExplosion LowCacheHitRate NotifyListenerDisconnected \
|
|
HighConfigReloadLatency ConfigValidationFailures CircuitBreakerOpen \
|
|
BackendServiceUnhealthy HighBackendLatency ConnectionPoolExhaustion \
|
|
ExcessiveRateLimiting; do
|
|
|
|
state=$(curl -s -m 2 "${PROMETHEUS_URL}/api/v1/rules" 2>/dev/null | \
|
|
jq -r ".data.groups[].rules[] | select(.name==\"$alert\") | .state" 2>/dev/null)
|
|
|
|
if [ -n "$state" ]; then
|
|
echo "✅ $alert ($state)"
|
|
ALERT_COUNT=$((ALERT_COUNT + 1))
|
|
else
|
|
echo "❌ $alert (NOT FOUND)"
|
|
fi
|
|
done
|
|
echo ""
|
|
echo "Alert Rules Found: $ALERT_COUNT / 13"
|
|
echo ""
|
|
|
|
# Test 4: Alert Health
|
|
echo "=== Test 4: Alert Evaluation Health ==="
|
|
curl -s -m 5 "${PROMETHEUS_URL}/api/v1/rules" 2>/dev/null | \
|
|
jq -r '.data.groups[].rules[] | select(.health != "ok") | "\(.name): \(.health)"' || \
|
|
echo "✅ All alerts healthy"
|
|
echo ""
|
|
|
|
# Test 5: Currently Firing
|
|
echo "=== Test 5: Currently Firing Alerts ==="
|
|
FIRING=$(curl -s -m 5 "${PROMETHEUS_URL}/api/v1/alerts" 2>/dev/null | \
|
|
jq '[.data.alerts[] | select(.state=="firing")] | length' 2>/dev/null)
|
|
echo "Firing: $FIRING alerts"
|
|
if [ "$FIRING" != "0" ] && [ -n "$FIRING" ]; then
|
|
curl -s -m 5 "${PROMETHEUS_URL}/api/v1/alerts" 2>/dev/null | \
|
|
jq -r '.data.alerts[] | select(.state=="firing") | " - \(.labels.alertname)"'
|
|
fi
|
|
echo ""
|
|
|
|
# Test 6: AlertManager Receivers
|
|
echo "=== Test 6: AlertManager Receivers ==="
|
|
curl -s -m 5 "${ALERTMANAGER_URL}/api/v2/status" 2>/dev/null | \
|
|
jq -r '.config.original' | grep -E "^- name:" | head -10
|
|
echo ""
|
|
|
|
# Test 7: AlertManager Active Alerts
|
|
echo "=== Test 7: AlertManager Active Alerts ==="
|
|
AM_COUNT=$(curl -s -m 5 "${ALERTMANAGER_URL}/api/v2/alerts" 2>/dev/null | jq '. | length' 2>/dev/null)
|
|
echo "AlertManager has $AM_COUNT active alerts"
|
|
echo ""
|
|
|
|
echo "======================================================================="
|
|
echo " Summary"
|
|
echo "======================================================================="
|
|
echo "✅ Prometheus: Connected"
|
|
echo "✅ AlertManager: Connected"
|
|
echo "✅ Alert Rules: $ALERT_COUNT / 13 loaded"
|
|
echo "📊 Currently Firing: $FIRING alerts"
|
|
echo "📊 In AlertManager: $AM_COUNT alerts"
|
|
echo ""
|
|
|
|
if [ "$ALERT_COUNT" == "13" ]; then
|
|
echo "✅ ALL ALERT RULES VALIDATED"
|
|
else
|
|
echo "⚠️ Missing alert rules: $((13 - ALERT_COUNT))"
|
|
fi
|