Redis version :5.0.7 Amount of RAM on the server : 15G. maxmemory-policy : volatile-lru maxmemory : 6.44G mem_fragmentation_ratio : 1.14
We have an instance of Redis on our Linux server. This instance of Redis is used on a PHP application. We ‘discover’ that we’ve huge fall of memory usage when the maxmemory amount is hit. This brutal fall affects our db graph. See bellow a graph representing the fall of the memory usage when redis hit max memory usage limit :
Is this pattern a ‘normal’ pattern for the given maxmemory policy ?
Furthermore, i’m asking myself if it’s a good practice to restart the redis instances in order to purge the redis cache at specific and controlled time ? For example, every night.