redis-lens

Redis diagnostics CLI for engineers who don't have Grafana yet.

No agents, no dashboards, no setup — just run it against your Redis instance.

Usage

# Analyze cache stampede risk
redis-lens scan

# Analyze memory usage by key pattern
redis-lens memory

# With options
redis-lens --url redis://:password@host:6379/ scan
redis-lens --url redis://127.0.0.1/ scan --bucket 30 --sample 0.5
redis-lens --url redis://127.0.0.1/ scan --pattern "user:*"
redis-lens --url redis://127.0.0.1/ scan --output json

Commands

scan

Analyzes key expiry distribution and detects cache stampede risk windows.

Options:
  --bucket <SEC>     Bucket size in seconds (default: 60)
  --sample <RATE>    Sample rate 0.0~1.0 (default: 1.0)
  --pattern <GLOB>   Only scan keys matching this pattern (e.g. user:*)
  --output <FORMAT>  Output format: text or json (default: text)

memory

Analyzes memory usage grouped by key namespace prefix.

Options:
  --sample <RATE>    Sample rate 0.0~1.0 (default: 1.0)
  --output <FORMAT>  Output format: text or json (default: text)

Environment Variables

export REDIS_URL=redis://:password@host:6379/
redis-lens scan

CI/CD Integration

SCORE=$(redis-lens --url $REDIS_URL scan --output json | jq '.risk_score')
if [ "$SCORE" -gt 70 ]; then
  echo "⚠ High stampede risk detected ($SCORE/100), consider adding TTL jitter"
  exit 1
fi

Production Tips

• Use --sample 0.1 on large Redis instances to minimize load
• Run during low-traffic hours for full scans
• Use --pattern to focus on specific namespaces
• SCAN is non-blocking — safe to interrupt anytime

When to use redis-lens vs Grafana

	redis-lens	Grafana + Redis Exporter
Setup	none	complex
Real-time monitoring	no	yes
One-off diagnostics	yes	overkill
CI/CD integration	yes (--output json)	possible
Stampede risk analysis	yes	no
Memory pattern analysis	yes	partial

Roadmap

• Cache stampede risk detection
• Memory usage by key pattern
• JSON output for CI/CD
• REDIS_URL env var
• --pattern flag
• --watch mode
• --throttle option
• Hot key detection