The group graph

In order for users to be able to synchronize their data using DME, they must be members of a certain group in the directory (LDAP or Active Directory), usually called DME_User. The group name can be different, but in the following it will be referred to as DME_User.

(A note on terminology: In DME, LDAP and Active Directory are called directory servers. Sometimes the term LDAP is used to cover both - AD is just a specific implementation of LDAP.)

Some LDAP servers only return the group of which a particular user is a direct member. In most enterprises, however, the DME_User group would consist of other groups, resulting in a nested structure. In order to be able to verify that users belong in the DME_User group, even when the group is nested several layers deep, DME system builds a so-called "group graph". The group graph provides an efficient way to evaluate if a subgroup is actually given the right to synchronize or not.

For example: You want to allow the use of DME for the following groups: Sales, Marketing, Development.

You then create a DME_User group, and add those subgroups into DME_User. This can be represented as follows:

When DME requests user information about user Tubbs, the LDAP server might return that Tubbs is member of the following groups: Sales, All Users, US Users, SaaS.

Using the group graph, DME is able to recognize that the group Sales has actually been given DME_User rights.

In the Domain setup panel of the connector you can specify if the current connector should be used for reading the group graph. For more information, see Domain.

Next topic

Load balancing and failover