Database operator running as single instance in a Kubernetes cluster


When we use operator to manage the database in a Kuberenetes cluster, For instance, mongodb enterprise operator, we run a single instance of the operator in a pod on the cluster. In case of a failure of the instance and before it gets recreated, would the orchestration operation performed by the operator is stopped?

What effect it can have on database?