Check why cluster-controller consume up to 400Mi of memory
Using the native image this is the memory consumption for the cluster-controller is too high:
POD NAME CPU(cores) MEMORY(bytes)
postgres-0 pgbouncer 1m 1Mi
postgres-0 patroni 6m 145Mi
postgres-0 postgres-util 1m 0Mi
postgres-0 prometheus-postgres-exporter 2m 4Mi
postgres-0 cluster-controller 5m 523Mi
postgres-0 envoy 6m 17Mi
Check why this is happening. Probably using the JVM version could shed some light on which internal component is using so much memory.