Managing Keydb

KeyDB Management Skill

Monitor, analyze, and optimize KeyDB instances safely.

MANDATORY: Discovery-First Pattern

Always run INFO ALL before any operations. Never assume threading model or active replication topology.

Phase 1: Discovery

#!/bin/bash

keydb_cmd() {
    keydb-cli -h "${KEYDB_HOST:-localhost}" -p "${KEYDB_PORT:-6379}" \
              ${KEYDB_PASSWORD:+-a "$KEYDB_PASSWORD"} \
              --no-auth-warning "$@" 2>/dev/null || \
    redis-cli -h "${KEYDB_HOST:-localhost}" -p "${KEYDB_PORT:-6379}" \
              ${KEYDB_PASSWORD:+-a "$KEYDB_PASSWORD"} \
              --no-auth-warning "$@"
}

echo "=== Server Info ==="
keydb_cmd INFO server | grep -E 'keydb_version|redis_version|uptime_in_days|server_threads|os'

echo ""
echo "=== Threading ==="
keydb_cmd CONFIG GET server-threads 2>/dev/null
keydb_cmd CONFIG GET server-thread-affinity 2>/dev/null

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

echo ""
echo "=== Keyspace ==="
keydb_cmd INFO keyspace

echo ""
echo "=== Active Replication ==="
keydb_cmd INFO replication | grep -E 'role|connected_slaves|master_link_status|master_host|active_replica'

echo ""
echo "=== FLASH Tier ==="
keydb_cmd CONFIG GET storage-provider 2>/dev/null || echo "FLASH storage not configured"

Phase 1 outputs: KeyDB version, thread count, memory usage, keyspace, replication topology, FLASH status.

Phase 2: Analysis

#!/bin/bash

keydb_cmd() {
    keydb-cli -h "${KEYDB_HOST:-localhost}" -p "${KEYDB_PORT:-6379}" \
              ${KEYDB_PASSWORD:+-a "$KEYDB_PASSWORD"} \
              --no-auth-warning "$@" 2>/dev/null || \
    redis-cli -h "${KEYDB_HOST:-localhost}" -p "${KEYDB_PORT:-6379}" \
              ${KEYDB_PASSWORD:+-a "$KEYDB_PASSWORD"} \
              --no-auth-warning "$@"
}

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

echo ""
echo "=== Multi-Master Status ==="
keydb_cmd INFO replication | grep -E 'active-replica|multi_master'

echo ""
echo "=== Slow Log ==="
keydb_cmd SLOWLOG GET 10

echo ""
echo "=== Client Connections ==="
keydb_cmd CLIENT LIST | awk -F'[ =]' '{for(i=1;i<=NF;i++) if($i=="addr") print $(i+1)}' | \
    cut -d: -f1 | sort | uniq -c | sort -rn | head -10

echo ""
echo "=== Sub-Key Expires ==="
keydb_cmd INFO stats | grep -E 'subkey|expired' | head -5

echo ""
echo "=== Persistence ==="
keydb_cmd INFO persistence | grep -E 'rdb_last_save|rdb_last_bgsave_status|aof_enabled'

Output Format

KEYDB ANALYSIS
==============
Version: [version] | Threads: [count] | Mode: [standalone/active-replica]
Memory: [used]/[max] | Keys: [count] | FLASH: [enabled/disabled]

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