Before you look at the content of identity stores you want to migrate, you should consider other sources of information that will help you identify a definitive set of legitimate users. For example, it can be useful to contact people in other departments who have reliable knowledge about the current organization or historical knowledge about how the organization has evolved. Individuals with information about a segment of the user population can help you identify accounts that are obsolete or were created for testing, or belong to users who have left the company or moved to another department.
As a starting point for collecting information about existing users and groups, consider doing the following:
- Contact HR to get an up-to-date list of current active-duty employees, contractors, and consultants. You can use this information to compare personnel records to the UNIX accounts to be migrated. After you identify which accounts correspond to people in the organization, you can create a spreadsheet to record the UNIX user names, UIDs, and other useful fields for those accounts.
- Contact enterprise security administrators or department-level UNIX administrators to determine whether all of the accounts defined for a computer still need access to that computer. For example, you should determine if any users validated as current employees have changed departments or job functions. If a user no longer needs access to some computers, you may not need to add a profile for that user.
- Identify any conventions used in defining the namespace. For example, is there a standard format for the contents of the GECOS field? How do the conventions used for UNIX, Linux, or Mac OS X accounts compare to the conventions used in Active Directory? For example, is the convention used for the UNIX login name the same as the convention used for the user’s sAMAcountName in Active Directory? Does the GECOS field follow the same conventions as the user’s displayName in Active Directory?
- Identify which user attribute fields that can be used as primary keys for identifying a unique user. Depending on the conventions you use for creating new accounts, the user name, user identifier (UID), or the GECOS field may be a reliable field for identifying real users and mapping them to Active Directory accounts. If you use a standard provisioning convention across platforms for an attribute such as the GECOS field or user name, the convention makes it much easier to identify unique users and map user profiles to Active Directory accounts.