OOMKilled in InfluxDB, how to properly set memory settings?

Hello,
Since I’ve migrated from Docker to Openshift, I have problem with OOMKilled in InfluxDB.
It looks like InfluxDB is killed everyday at the same time (there is cron tasks inside InfluxDB?).
I set up continous queries and retention policies for metrics but it didn’t help.
I set up memory settings (INFLUXDB_DATA_CACHE_MAX_MEMORY_SIZE, INFLUXDB_DATA_CACHE_SNAPSHOT_MEMORY_SIZE, INFLUXDB_DATA_CACHE_SNAPSHOT_WRITE_COLD_DURATION) to have control at memory usage, but it didn’t help too.
Do You know how to overcome this problem? Or maybe some advice what settings can I add to avoid OOMKilled in InfluxDB?

Best regards,

Hi @marazm ,

this one may be interesting … avoid oom kill

but oom will choose a new victim to kill …