[Security and Access Control][CE] [Security] - Symphony should immediately reflect changes in Active Directory

The current Active Directory behavior in Symphony 7.1 is flawed. Feedback from support is that the expected practice is after a new user or group is added to AD or an existing user is added to an existing group that the cluster should be restarted in order to acquire the changes to Active Directory. This is not correct and is unrepresentative of the manner in which other AD based security models work. For example it is not necessary to restart a file server or SQL server when a user is added into an AD group. Similarly, e-mail systems (such as Lotus Notes) can use AD groups to maintain distribution lists that can be updated without requiring a restart of the server software.
The purpose of Symphony AD integration is to enable IT support staff (who do not have knowledge or experience of Symphony) to grant or revoke Symphony access to a user based on that user's membership of groups to which the Symphony administrator has assigned appropriate roles within the cluster. This enables security to be centrally managed by IT support staff in line with most other IT systems. For the design behavior to require a cluster restart following an AD change, this forces the Symphony administrator to be involved in what should be a routine operation.
The other principal issue with the design behavior is that a cluster restart is an invasive operation that requires coordinating a period of downtime within the business. For a dynamic organisation where users may need to be added at short notice during a busy time for the cluster this represents a problem, forcing the Symphony administrator to assign roles directly to the user within the cluster.
It is clear that Symphony is capable of eventually reflecting AD changes without a cluster restart, however these changes can take anything from a hour to some days to register. This behavior needs to change such that AD changes are reflected in Symphony within seconds and without any additional actions or commands required on the cluster itself.