I originally posted this at http://www.318.com/TechJournal
In Leopard, Workgroup Manager supports rudimentary ACLs for the LDAP database. We’re all familiar with Access Control Lists by now. Especially in the Mac OS X Server community. However, we might not all be familiar with ACLs as they’re implemented in LDAP. But we should be, because LDAP is being used more and more as an address book, and with the new Directory application being shipped in Leopard it is conceivable that environments aren’t just going to use ACLs to secure LDAP but they’re also going to use them to allow users to self update their information in the directory. So in the interest of security and making the most out of the technologies build into LDAP, let’s cover LDAP ACLs for a bit. So to push beyond what you can do in Workgroup Manager, let’s take a look at building out more finely grained ACLs manually.
First, like with most things in LDAP ACLs are configured using the /etc/openldap/slapd.conf file. Below is the pertinent portion of this file that we will be looking at:
# Define global ACLs to disable default read access.
# Do not enable referrals until AFTER you have a working directory
# service AND an understanding of referrals.
#referral ldap://root.openldap.org
# Sample access control policy:
# Root DSE: allow anyone to read it
# Subschema (sub)entry DSE: allow anyone to read it
# Other DSEs:
# Allow self write access
# Allow authenticated users read access
# Allow anonymous users to authenticate
# Directives needed to implement policy:
# access to dn.base="" by * read
# access to dn.base="cn=Subschema" by * read
# access to *
# by self write
# by users read
# by anonymous auth
#
# if no access controls are present, the default policy
# allows anyone and everyone to read anything but restricts
# updates to rootdn. (e.g., "access to * by * read")
#
# rootdn can always read and write EVERYTHING!
Now, if we remove the commented out portions of the file or add more lines we can start to limit who has access to read and/or change what information in the LDAP database. Keep in mind that you always want to back up your slapd.conf file prior to doing so.
You can control access to each element in the database. Each ACL has an “access to†which is the elements in the LDAP database that you are granting or denying access for and then a “by†portion that lists who can do what to that portion of the database. An entire ACL can be listed on one line, as is done with policies that have only one user or group associated to them. For example, the following line gives anyone and everyone read access to the database:
access to dn.base=— by * read
For ease of use and reviewing, we typically put the “access to†on one line and the subsequent users or groups with access in their own “by†lines for more complicated ACL rule sets. Slapd parses the file in such a way that it realizes that “access to†means the beginning of a new ACL. The following is an example of some more complicated ACLs:
access to attrs=userPassword
by dn="cn=users,dc=318,dc=com" write
by self write
by * compare
access to *
by dn="cn=computers,dc=318,dc=com" write
by users read
by * auth
Access levels in ACLs are hierarchical. Levels that are used are none, auth, compare, search, read and write. None is the lowest level of access and write is the highest. Each level includes the rights of all lower levels. In the above example, a user is able to write to their own userPassword record. This means that the user is also able to auth, compare, search and read that record.
ACLs are prosessed from top to bottom. This makes it important to put specific ACLs and by statements above more general ones. ACLs that restrict access to the userPassword attribute, followed by one applicable to *, that is, the entire LDAP database. In the above example, placing the userPassword ACL first causes the rule that allows users to change their own passwords to process before the wildcard that specifies everyone. When a * is used as a wildcard in the access to line of slapd.conf it means the entire database or tree of the LDAP database. When the * is used in the by line it typically denotes all users.
Access levels in ACLs are hierarchical. Levels that are used are none, auth, compare, search, read and write. None is the lowest level of access and write is the highest. Each level includes the rights of all lower levels. These two points, the first match wins rule and the inclusive nature of access levels, are crucial to understanding how ACLs are parsed. They also are important for making sure your ACLs don’t lead to either greater or lesser levels of access than you intend in a given situation.
It can be time consuming to go through every possible attribute by group and determine who has access to what. However, if you want to have users updating their own addresses, phone numbers, and other information, as can be done with the Directory application, this is often one way to accomplish this goal. You could also provide help desk users the ability to update the database using the Directory application but not allow them to access other records in the LDAP database, such as group memberships. Having a very granular ACL environment for records can also allow you to obtain a maximum level of security.
This can also be put into the schema in order to force replication between hosts. Keep an eye out for that article at a later date. 
For what it’s worth, at 318 we’ve found that commenting out each ACL helps us to keep track of who did what, why and what they were thinking when they did it. Happy OD everyone!!!
