Redis Cache Key Convention Generator

A cache key naming convention is a structured pattern for Redis keys that makes them self-documenting, easy to namespace, and query-friendly with SCAN. Common patterns: colon-separated ('user:123:profile'), pipe-separated ('user|123|profile'), or dot-separated ('user.123.profile'). Colon-separated is the most widely adopted convention.

General guidelines: User sessions: 30 min–24 hours (depending on security needs). API responses: 60–300 seconds (balance freshness vs load). Auth tokens: match token expiry (15 min for access tokens, 7 days for refresh). Rate limit counters: 1 minute or sliding window. Search results: 5–30 minutes. Product catalog: 1–24 hours. Use shorter TTLs for frequently changing data.

Use SCAN with a glob pattern: SCAN 0 MATCH user:* COUNT 100. Never use KEYS * in production as it blocks the server. For atomic bulk deletion, use a Lua script: redis.call('DEL', unpack(redis.call('KEYS', ARGV[1]))). Better yet, use key prefixes designed for SCAN and delete in batches.

Redis supports keys up to 512 MB, but best practice is to keep keys under 1KB for memory efficiency. Very long keys increase memory usage and slow down comparisons. Use SHA hashes for keys derived from large inputs (like full URL paths): 'api:response:'+sha256(url).

EXPIRE key seconds — sets TTL in seconds from now. EXPIREAT key timestamp — sets expiry at a Unix timestamp. PEXPIRE key ms — sets TTL in milliseconds. SETEX key seconds value — atomically SET + EXPIRE in one command. SET key value EX seconds — modern equivalent with additional options (NX, XX, GET). Always prefer atomic SET...EX over separate SET+EXPIRE.