database resource_error ai_generated true

redis.exceptions.ResponseError: OOM command not allowed when used memory > 'maxmemory'.

ID: database/redis-oom

Also available as: JSON · Markdown
90%Fix Rate
92%Confidence
60Evidence
2023-01-01First Seen

Version Compatibility

VersionStatusIntroducedDeprecatedNotes
7 active — — —

Root Cause

Redis has reached its maxmemory limit and the configured eviction policy does not allow freeing memory for the current command. Write commands are rejected. Caused by storing too much data without an eviction policy, memory leaks from accumulating keys, or maxmemory set too low.

generic

Workarounds

  1. 92% success Configure an appropriate maxmemory-policy eviction strategy 
    Set an eviction policy: redis-cli CONFIG SET maxmemory-policy allkeys-lru (for cache use cases) or volatile-lru (to only evict keys with TTL). Make persistent in redis.conf. Choose the policy based on your use case: allkeys-lru for caches, volatile-ttl for session stores.
  2. 88% success Identify and remove large or unnecessary keys to free memory 
    Run: redis-cli --bigkeys to find the largest keys. Use redis-cli MEMORY USAGE <key> to check specific keys. Delete unnecessary keys with DEL or UNLINK (async). Set TTL on keys that should expire: EXPIRE key seconds. Monitor memory: redis-cli INFO memory.

Dead Ends

Common approaches that don't work:

  1. Setting maxmemory to 0 (unlimited) on a server with limited RAM 85% fail

    Without a maxmemory limit, Redis will consume all available system memory until the Linux OOM killer terminates the process. This causes data loss and affects other processes on the same server.

  2. Restarting Redis without persistence to clear memory 80% fail

    If Redis is configured without AOF or RDB persistence, restarting loses all data. Even with persistence, the same data will be reloaded and memory will fill up again immediately after restart.

Error Chain

Leads to:
Preceded by:
Frequently confused with: