After a node crashes and rejoins the cluster, it does not receive client load at the same level as the other nodes.
This issue does not affect data integrity and distribution in the cluster.