Check the running version from the command line:
redis-cli -p 6370 info | grep -i redis_version
Check that the Redis command-line interface (CLI) connects to the Redis server.
redis-cli -p 6370 ping PONG
Check max memory
By default redis is not set with a memory limit, so it's best to set one grep ^maxmemory /etc/redis/redis.conf maxmemory 536870912 Set with 0 by default which means unlimited redis-cli -p 6370 127.0.0.1:6370> config get maxmemory 1) "maxmemory" 2) "0" Emergency increase Redis maxmemory redis-cli -p 6370 127.0.0.1:6370> CONFIG SET maxmemory 8GB OK
Troubleshooting Memory
'maxmemory_human' must be larger than 'used_memory_peak_human' in case of using the 'volatile-lru' policy. redis-cli -p 6370 info | grep memory | grep 'human\|policy' When memory is fully used, the number of 'evicted_keys' will be large and growing fast. The maxmemory should be increased in this case even when it's using 'allkeys-lru' to prevent high amount of cache miss. redis-cli -p 6370 info | grep evicted_keys More info here https://redis.io/topics/lru-cache Logging errors that show that the maxmemory limit is being hit. cron.log: OOM command not allowed when used memory > 'maxmemory' error.log: PHP Fatal error: Uncaught Exception: Warning: session_write_close(): Failed to write session data using user defined save handler.
Troubleshooting Throughput
# if input or output is greater than 100 mbps then there can be an application issue redis-cli -p 6370 info stats | grep instantaneous
Troubleshooting Latency
# Latency monitoring is enabled by setting a threshold measured in milliseconds redis-cli -p 6370 CONFIG SET latency-monitor-threshold 100 # this command will generate a human-readable report of the logged latency events and recommendations on how to resolve them latency doctor # More info here https://redis.io/topics/latency-monitor
Check keys expiration
$ redis-cli -p 6370 KEYS *session* TTL name_of_session_key