managing-dragonfly

name	managing-dragonfly
description	Use when working with Dragonfly — dragonfly in-memory datastore management, multi-threaded performance analysis, memory efficiency monitoring, and compatibility checks. Covers thread utilization, snapshot operations, memory usage per thread, replication status, and Redis/Memcached protocol compatibility. Read this skill before any Dragonfly operations.
connection_type	dragonfly
preload	false

Dragonfly Management Skill

Monitor, analyze, and optimize Dragonfly instances safely.

MANDATORY: Discovery-First Pattern

Always run INFO ALL before any operations. Never assume threading model or memory layout.

Phase 1: Discovery

#!/bin/bash

df_cmd() {
    redis-cli -h "${DRAGONFLY_HOST:-localhost}" -p "${DRAGONFLY_PORT:-6379}" \
              ${DRAGONFLY_PASSWORD:+-a "$DRAGONFLY_PASSWORD"} \
              --no-auth-warning "$@"
}

echo "=== Server Info ==="
df_cmd INFO server | grep -E 'dragonfly_version|redis_version|uptime_in_days|tcp_port|os|used_threads'

echo ""
echo "=== Memory Overview ==="
df_cmd INFO memory | grep -E 'used_memory_human|used_memory_peak_human|maxmemory_human|mem_fragmentation_ratio'

echo ""
echo "=== Keyspace ==="
df_cmd INFO keyspace

echo ""
echo "=== Database Size ==="
df_cmd DBSIZE

echo ""
echo "=== Clients ==="
df_cmd INFO clients | grep -E 'connected_clients|blocked_clients'

echo ""
echo "=== Replication ==="
df_cmd INFO replication | grep -E 'role|connected_slaves|master_link_status'

Phase 1 outputs: Dragonfly version, thread count, memory usage, keyspace, client count.

Phase 2: Analysis

#!/bin/bash

df_cmd() {
    redis-cli -h "${DRAGONFLY_HOST:-localhost}" -p "${DRAGONFLY_PORT:-6379}" \
              ${DRAGONFLY_PASSWORD:+-a "$DRAGONFLY_PASSWORD"} \
              --no-auth-warning "$@"
}

echo "=== Performance Stats ==="
df_cmd INFO stats | grep -E 'instantaneous_ops_per_sec|total_commands_processed|keyspace_hits|keyspace_misses|evicted_keys'

echo ""
echo "=== Hit Rate ==="
df_cmd INFO stats | awk -F: '
    /keyspace_hits/ { hits=$2 }
    /keyspace_misses/ { misses=$2 }
    END {
        total = hits + misses
        if (total > 0) printf "Hit rate: %.1f%%\n", hits/total*100
    }'

echo ""
echo "=== Slow Log ==="
df_cmd SLOWLOG GET 10 2>/dev/null || echo "Slowlog not available"

echo ""
echo "=== Snapshot Status ==="
df_cmd INFO persistence | grep -E 'rdb_|aof_|save_|snapshot'

echo ""
echo "=== Key Type Distribution (sample) ==="
df_cmd SCAN 0 COUNT 200 | tail -n +2 | while read key; do
    df_cmd TYPE "$key"
done | sort | uniq -c | sort -rn

Output Format

DRAGONFLY ANALYSIS
==================
Version: [version] | Threads: [count]
Memory: [used]/[max] | Keys: [count] | Ops/sec: [count]

ISSUES FOUND:
- [issue with affected component]

RECOMMENDATIONS:
- [actionable recommendation]

Anti-Hallucination Rules

1. NEVER assume resource names — always discover via CLI/API in Phase 1 before referencing in Phase 2.
2. NEVER fabricate metric names or dimensions — verify against the service documentation or --help output.
3. NEVER mix CLI commands between service versions — confirm which version/API you are targeting.
4. ALWAYS use the discovery → verify → analyze chain — every resource referenced must have been discovered first.
5. ALWAYS handle empty results gracefully — an empty response is valid data, not an error to retry.

Counter-Rationalizations

Shortcut	Counter	Why
"I'll skip discovery and check known resources"	Always run Phase 1 discovery first	Resource names change, new resources appear — assumed names cause errors
"The user only asked for a quick check"	Follow the full discovery → analysis flow	Quick checks miss critical issues; structured analysis catches silent failures
"Default configuration is probably fine"	Audit configuration explicitly	Defaults often leave logging, security, and optimization features disabled
"Metrics aren't needed for this"	Always check relevant metrics when available	API/CLI responses show current state; metrics reveal trends and intermittent issues
"I don't have access to that"	Try the command and report the actual error	Assumed permission failures prevent useful investigation; actual errors are informative