We are working our ingestion performances, and we are unable to check the database behaviour as metrics concerns all the cluster, and are not representative of the node leader status.