[RFE] - Make it possible to set requested value for fluentd daemonset

Issue

• The logging playbook is just using the value of the variable `fluentd_memory_limit' .
• Is possible to define the value for requests too?
• This behavior already cause and outage in our environment, as this was blocking the router pods to start.
• Affected Line:
  https://github.com/openshift/openshift-ansible/blob/ee9ba74ff556e74b30109c1945a3f0a02c7d7d92/roles/openshift_logging_fluentd/templates/2.x/fluentd.j2#L52