Memory cleanup after crash dump taken

[splitice]

We are currently exploring scaling out (downsize from 8GB ram nodes to 4GB ram nodes) our kubernetes cluster and we noticed that theres quite the variation in core-dump-handler pods memory usages.

core-dump-handler-8g52g                    0m           130Mi
core-dump-handler-9vwk9                    0m           58Mi
core-dump-handler-jw5fv                    0m           86Mi
core-dump-handler-q9hz5                    0m           21Mi
core-dump-handler-v98sh                    0m           102Mi
core-dump-handler-vd7kd                    0m           14Mi
core-dump-handler-vsv2r                    0m           38Mi
core-dump-handler-zgvjz                    0m           114Mi

This appears to be the case that handlers that have handled a crash have significantly higher memory usage than those that have not.

Could this be resolved?