Cluster-Wide mode does not support backups of multiple Mongo-Instances in v1.13.0

Prior to v1.13.0 it was possible to create backup-tasks for multiple clusters in a cluster-wide mode (with adapted roles) thanks to the usage of Kubernetes CronJobs.

Since PerconaServer MongoDB Operator v1.13.0 the cronjobs are not used anymore but instead the jobs are executed inside the one and only Operator Pod. It is now not possible anymore to generate backups via the PerconaServer MongoDB Operator for multiple mongo instances anymore.

The problem is, that the backup-jobs are not using the namespaces as differentiators. It is also not possible to use different names for the backup-jobs for the different instances. This results in the jobs being deleted and recreated all the time.