Dynamic Access Control (DAC) is a complete, end-to-end solution to secure information access and is not just another new feature of Windows Server 2012. DAC can really help you to solve some daily problems you may have in giving access to data on distributed file servers. For example, Jack works on a project called Ikarus, and he needs some information from the marketing department, but Jack is not really a member of that department. Therefore, you are going to build some security groups to solve this request, and a complex group scenario starts to exist, because the groups and their memberships will grow and in each case become more and more complex. In addition, it is always a challenge to audit and monitor such a solution. You might know situations such as "Who had access to the sensitive finance information on June 1, 2013?" Or the wonderful "access denied" message a user encounters that leads them to ask for access to a particular piece of information. Immediately you start searching to provide the Chief Information Security Officer (CISO) of the organization the right information for evidence on who the owner of this information is for the CISO or the data owner to decide whether or not to give the user proper access. These are a few short examples that we will discuss in the following chapters to give you a broad overview. Do not forget that we will go in deep in the following chapters.
The topics we will cover in this chapter are:
Business needs, purpose, and benefits
Inside the architecture of DAC
Building your smart test lab
Getting started with your first real-life solution
In today's complex IT environments, file servers play an increasingly vital role. We store tonnes of data and information on them, which is distributed for many individuals in an organization. Additionally all of this data needs to be secure, accessible across varied networks, devices, and applications, and needs to enact with strategies like Bring Your Own Device (BYOD), Direct Access, and different Cloud solutions.
To hold the costs down while meeting the security requirements is always a challenge for those responsible.
The main challenges for data owners or file server administrators are as follows:
The numbering and management of security groups needs to be reduced as illustrated in the simple example consisting of the Account—Global Groups—Domain Local Groups—Permissions principles shown in the following diagram:
The idea of the following list is to show a part of the current challenges with respect to managing, securing, and maintaining information. Feel free to extend the list infinitely for your notes:
Central access and audit management of business and compliance needs
Building enhanced authentication and authorization scenarios (for example, BYOD)
Sensitive information needs to be protected wherever it goes
The productivity of information workers should not be affected
The content owners should be responsible for their information
To provide access-denied assistance messages to provide a managed end-to-end scenario
So the million-dollar question is, "How can Dynamic Access Control help you to address and solve these requirements?".
Dynamic Access Control provides you with the following enhanced ways to control and manage access in your distributed file server environment:
Classification: Identify and classify your information based on their content. There are four ways to tag information; by location, manually, automatically, and using application APIs.
Control access: Build up the precise definitions of the right person, with the right permission, at the right time, from the defined device. Usage of the Central Access Policy (CAP) will help you to address the following common security policies, compliance (general, organization-wide, departmental, specific-data) and the need-to-know principle.
Compliance: This is a response to governmental regulations, but it can also be a response to industrial or organizational requirements:
Policy staging: This allows you to control changes to CAPs by comparing current settings against new settings by firing event log entries into the system log. Information can be analyzed using Event Viewer or by connecting with System Center Operations Manager.
Access denied remediation: In current environments, you get just a very simple access-denied message, which is not very helpful for the helpdesk or the user. DAC provides additional information and the opportunity to send information that is more useful to the data owner.
Audit: Defining policies based on information security, organizational and departmental requirements for reporting, analysis, and forensic investigation. Central Audit Policies form the key answer provided by Dynamic Access Control for those requirements.
Protection: Dynamic Access Control integrates with Active Directory Rights Management Services (AD RMS) for classification-based automatic encryption of sensitive tagged information. This option helps in any transmission aspect to protect the content against any unauthorized person.
Now that you have had a little recap about the business needs, the purpose, and the benefits of Windows 2012 Dynamic Access Control, we can dive into the technical details.
As promised in the previous section, Dynamic Access Control is not just a single feature, but an end-to-end file server solution based on the following features in Windows Server 2012:
Windows authorization and audit engine supporting expression-based access control
Kerberos version 5 support for user and device claims
File classification infrastructure that supports claims
RMS support that can be extended for further file types from third-party vendors
API to extend the solution with custom classification and audit tools
The Dynamic Access Control solution can be logically divided into the following main components to get a better, granular overview:
Infrastructure requirements
User and device claims
Expression-based ACEs
Classification enhancements
Central access and audit policies
Access-denied assistance
These different building blocks are explained in the following sections with all the details. But first, you need to get a quick overview of the most important facts of Dynamic Access Control. We will start the overview with the infrastructure requirements.
For basic deployment of Dynamic Access Control, you do not need to put in a big effort. To use claims for authorization and auditing, there is only a need for the following components:
At least one Windows 2012 or newer domain controller
Configure DAC objects, which are:
Claim Types
Central Access Rules
Central Access Policies
Administering with Active Directory Administrative Center (ADAC) or Remote Server Administration Tools (RSAT) installed on Windows 8/ Windows Server 2012 or newer
Group policy to deploy Central Access Policies to your file servers
Group policy to enable the KDC support for claims
Group policy to enable the Kerberos client support for claims
All the file servers that use DAC must be 2012 or newer
Windows 8 or newer client computers must be part of that domain (only required when using device claims)
AD RMS role must be enabled and configured if you want to use automatic encryption
You need to enable claims support on domain controllers and clients (disabled by default)
DAC stores all configurations in the Active Directory configuration partition
Group policies are used to configure DAC on file servers and clients
The File Server Resource Manager (FSRM) brings up many features such as File Server Classification Infrastructure (FCI)
Dynamic Access Control also works over organization boundaries with Claims Transformation Policies (CTP)
The following figure shows the basic deployment and configuration that needs to be done
However, what happens if you don't use Windows 8 clients?
Note
For non-Windows 8 / Windows Server 2012, such as XP, Vista or Windows 7, the user doesn't need to worry about claims. In that case, the 2012-based file server will query the Active Directory services and forward the claims request to get information about the claims the user or the machine provides.
As you can see in the figure above, DAC works between different Active Directory Forests (Active Directory instance of an organization), and Claims Transformation Policies will provide the functionality to translate the claims definitions between two or more organizations. To prepare for this scenario, you need to establish a Forest Trust between the Active Directory Forests and the Domain Function Level (DFL), which in both the Forest Root domains must be Windows 2012 or higher. Right now, this is a challenge but also a necessary requirement. There is no need for Claim Transformation Rules inside a Forest. This works fine out of the box because Dynamic Access Control objects are stored in the configuration part of the Active Directory and the whole Forest knows the relevant information.
Traditionally, you may have secured access to files by using NTFS file permissions and security groups. With this configuration, we were restricted to making policy decisions based on the user's group membership and the number of groups will explode. Therefore, if we wanted to include the device to control access, there was no chance to do this in an earlier version of the Windows Server. Another limitation was the requirement for folder or file access based on a certificate. Before Windows 2012 Dynamic Access Control, there was no way for the built-in functionality to include devices or certificates. DAC now integrates claims into Windows Authentication so that we can use Active Directory attributes from users and computers to control access to our information stored on file servers such as a location, department, or project.
The following figure shows the new combinations you can use for authorization:
This opens new ways of giving permissions on files and folders, such as:
Allow | Read, Write | If (@User.Department == @File.Department) AND (@Device.Managed == True)
By using expression-based access control, users or devices must satisfy conditions that we define to access files in a given classification.
To explain the major benefits, we use a very easy and common example. Let us consider that 200 projects, 20 countries, and two divisions are part of an organization. So in fact, this results in something like 8,000 groups to solve the access control in this scenario using the traditional approach. Reducing security groups is always a vital task in the current IT environment. For example:
Project Budget2014 CH Finance Users
Project Budget2014 UK Finance Users
Windows Server 2012, without claims, already allows multiple groups with a Boolean logic (expression-based Access Control lists). This helps us to reduce the groups in an effective way. Let us look at the following example of using the AND
operations to build up a permission model:
Allow Modify IF MemberOf(ProjectA) AND MemberOf(CH) AND MemberOf(Finance)
The result is 222 groups instead of something like 8,000 security groups. Yeah!
Finally, by using claims inside the expression-based access rules, we can convert the groups into exactly three user claims.
The first task in every Dynamic Access Control project is to identify and classify files based on their content. With Windows 2008 R2, we could already fulfil the following tasks:
Define classification properties
Automatically classify files based on location and content
Apply file management tasks (file expiration / custom commands) based on classification
Produce reports
With Windows Server 2012, the following classification improvements are added:
Manual classification (Windows Explorer)
Continuous classification (File Server Resource Manager)
Folder-based inherited classification
Conditional access control entries (additional authorization layer)
The next figure gives you an introduction to the processes carried out in a file classification scenario and shows the continuous classification:
Define resource properties in Active Directory such as a department or company, and apply them to your file servers.
The File Classification Infrastructure checks the file content and classifies the information with the correct classification.
After classifying the information, the classification can be used for authorizing access to the information.
With the Windows Server 2012 File Classification Infrastructure (FCI) feature, you can identify sensitive files and encrypt them automatically with RMS.
Some possible scenarios include:
Access to all documents on the file server must be limited to active, full-time employees of the company—even if an employee distributes copies to different places, such as Skydrive, Dropbox, or SharePoint
The AD RMS-policy of Finance read only must be applied to all files containing more than 10 credit card numbers or other Personal Identifiable Information (PII)
The AD RMS-policy of Sales Managers only to all Excel files larger than 100 MB containing Personal Identifiable Information (PII) and 10 contract numbers being created by the CRM system
This technology also gives you the possibility of supporting file types other than Office documents. You just need to install and configure a combination of FCI with Rights Protected Folder Explorer from http://blogs.technet.com/b/rms/archive/2012/06/29/official-release-of-rights-protected-folder-explorer.aspx.
Otherwise, you need to add a third-party solution to provide support for other file types.
Central Access Policies (CAPs) play an essential role in a Dynamic Access Control scenario. CAPs are a set of authorization policies that we manage in the Active Directory and deploy them to the file servers over Group Policies. You can think about a CAP-like safety net policy to give you another idea of what you can expect from that element.
A CAP has two logical parts:
Defined conditions as to which files the policy will be applied
List of one or more Access Control Entries (ACEs)
The next figure should provide you with some information on how the different solution components interact and where the information of the DAC objects is stored. Furthermore, it gives you the necessary tasks in the right order and the tools that you can use to configure CAPs, claims, and property definitions.
Obviously, if you change policy, you want to check the consequences of your work. For this reason there is a function called "policy staging" available, which lets you run a new policy parallel to your current configuration to evaluate the results.
On the left-hand side of the following figure, you see the tasks that need to be done to configure Dynamic Access Control, and on the right-hand side, the results on the system.
Also, a new tab is present in the Advanced Security Setting for Finance Documents called Central Policy.
After applying Central Access Policies, we need to think about Auditing Policies. With Windows Server 2012, you can author audit policies by combining claims and resource properties. It enables scenarios for you that were impossible or very hard to implement until now. The next figure shows you the file-access auditing workflow to give you a better understanding of this process:
A quick look at how much power is inside these new audit improvements:
Note
Auditing everyone who is not working on a specific project and trying to access information tagged as only accessible for full-time employees and a project member working on that project is now possible.
To view and analyze audit events you can use the common Event Viewer or if available, the System Center Operations Manager with the Audit Collection Service configured.
Access-denied assistance is a role service of the File and Storage Services role in Windows Server 2012 and helps us in the following use cases:
Users get more than just an Access-denied message. They are provided with detailed information for the data owner, helpdesk, or file server administrators.
Allows user to request access from the data owner.
There are two ways to configure the Access-denied assistance:
E-mail – The user gets a customized access-denied message with a button to request assistance and an e-mail fired to the data owner
Web service - The user gets a customized access-denied message with a link included and gets redirected to a self-service portal, such as Forefront Identity Manager 2010 R2
While building our smart and straightforward test lab, we will start to apply our knowledge in a practical way. Not wanting to spend hours, we start with a minimal lab and extend it step-by-step for our needs.
We start with the following configuration:
A domain controller Windows 2012 R2 (build your own Forest, such as
inovit.ch
)A domain-joined File Server Windows 2012 R2
A domain-joined Client Computer Windows 8.1 Pro
You might have noticed that we are using the latest versions. IT professionals always like to touch the newest one! In fact, we need this version because in further labs, we will show you how to integrate Dynamic Access Control in a Bring Your Own Device scenario including a Work Folders configuration.
There are no special requirements on the virtual environment, such as disk, CPU, or memory configuration. Just use your common configurations. Feel free to start as well with the Base Windows 2012 R2 Test Lab Guide at http://www.microsoft.com/en-us/download/details.aspx?id=39638.
On the file server, add an additional virtual disk to provide Shared Folders for our little test company and create a file structure as follows:
Create a shared folder for each country (
CH
,FR
,andMA
).Additionally, create a folder for each office location (
Zurich
,Paris
,Rabat
, andCasablanca
).Additionally, create a folder for each department (
Sales
,Human Resources
,Engineering
,Marketing
, andHelp Desk
).Under the department folders, create a folder called
Sensitive
.The structure looks like MA | Casablanca | Marketing | Sensitive.
Create a shared folder for some example projects (Project A, Project B, Project C).
Create a shared folder for some public information.
The next steps will provide you with the main tasks to implement your first Dynamic Access Control configuration.
Create some test users in your Active Directory with a minimum of 10 users and:
Define the Active Directory claim types.
Country
,Department
, andLocation
for the folder structure decided earlier.Populate the three attributes for the 10 test users.
Define the Resource properties for
Country
,Department
, andLocation
.Define the Active Directory Access Rule as follows:
(Resource.Country equals User.Country) AND (Resource.Location equals User.Location) AND (Resource.Department equals User.Department)
Build a Central Access Policy and deploy the Access Rule to the file servers.
Build a Resource Property list, and deploy it to the file servers.
Open an administrative PowerShell, and fire
gpupdate /force
andUpdate-FSRMClassificationPropertyDefinition
on the file server.On the resources, apply the Resource properties correctly.
Apply the Central Access Policy to the file shares.
Apply the Access Rule to all the
Country
shares and theLocation
andDepartment
folders.Try out whether access is allowed or not.
This chapter introduced a lot of new concepts and information such as the business needs, purpose, and benefits and the main components of Dynamic Access Control. It is always important to get a general overview of a technology to get a better understanding about the scope of what we need to go through. Don't worry if it sounds like you will get many things to study, for after understanding the main principles, you will see a successful deployment of Dynamic Access Control very soon.