Subscribe to Windows IT Pro
May 30, 2007 12:00 AM

Microsoft’s NAP Option

Windows IT Pro
InstantDoc ID #95954
Rating: (2)

Microsoft Network Access Protection (NAP) is a policy-based platform that lets you vet a system’s configuration according to a set of standards designed to protect healthy systems and networks by detecting and optionally limiting access to network resources those systems deemed vulnerable. NAP uses a client/server architecture (as do the other products in the main review) and includes an API that lets developers and vendors customize the capabilities of NAP with additional health validation, compliance, and enforcement mechanisms. NAP is incorporated into Windows Vista and the upcoming Longhorn Server. It will also be supported in Windows XP by a NAP Client for Windows XP.

Microsoft is quick to point out that NAP isn't designed to protect networks from malicious users, and NAC does nothing to prevent malware from running on a policy-compliant system. Rather, the idea is to promote network health by monitoring the configuration of managed systems to ensure, among other things, the presence of policy-compliant versions of security applications, firewall implementations, and anti-spyware, for example. As provided by Microsoft, NAP monitors the settings of the Microsoft Security Center, including Windows Firewall, Automatic Updates, and Windows Defender. Using the NAP API, others can extend NAP to support policy compliance for third-party products. NAP is designed to enable reporting or enforcement of policy compliance for dial-up, VPN, wireless, and wired network connections.

NAP supports IPsec enforcement, IEEE 802.1X enforcement, VPN enforcement, and DHCP enforcement. IPsec is NAP’s strongest form of enforcement, letting you configure Ipsec-secured communication between network endpoints, and control by IP address and TCP or UDP port who a compliant system may communicate with. Endpoints connecting through an 802.1x-authenticating device are enforced according to 802.1x protocols, receiving a limited-access profile when non-compliant and unlimited network access when compliant. With VPN enforcement, NAP assigns a set of IP packet filters to a non-compliant computer’s VPN connection at the VPN server. DHCP enforcement, implemented in the DHCP Client and DHCP Server services on supported platforms, causes non-compliant systems to receive a restricted-access IPv4 network configuration. Because anyone with administrative rights can assign fixed IP addressing to Windows systems, DHCP is the weakest enforcement mode. Administrators are free to implement any and all enforcement modes in their networks. NAP actively monitors policy compliance for all compliant systems and enforces limited access when a system becomes non-compliant.

Network Policy Server (NPS) in Longhorn Server acts as a health-policy server for all of these NAP enforcement methods. It also acts as a RADIUS server and proxy, replacing the Internet Authentication Service (IAS) present in Windows Sever 2003. On an NPS server, you configure health policies, specifying requirements for compliant systems. For non-compliant systems, you can also configure a set of remediation actions that a NAP client must perform to become compliant.

Related Content:

ARTICLE TOOLS

Comments
  • ROB
    5 years ago
    Aug 23, 2007

    Why can't someone develop a simple DHCP/802.1x solution based on the "computer" being a member of a AD domain (ie trusted) or a trusted MAC address (whitelist)? Seems so simple in concept. Why do we need to pre-scan and post-scan and add agents to systems we already trust and manage through other means? Why do I care about users...don't we already have enough methods to authenticate a user? NAC/NAP cannot stop malicious users. The issue is how to simply identify trusted devices. Most networks and budgets would be just fine with such a solution.

    My impression is the vendors have tried to pack everything into a single solution, suitable for high security environments, when in fact most networks are not high security environments, such as military, government or financial networks. Where's a simple solution that solves the the simple problem of denying access to untrusted systems? Let's face it, a well managed network, shouldn't require scanning of trusted systems. Scanning should limited to un-trusted systems whether through automation or manually.

    Seems to me we are ignoring the pareto principle, when we try to focus on everything instead of the meaningful few ie 80/20. I'm tempted to dig out my programming books and refresh some old skills.

  • Darren
    5 years ago
    Jun 08, 2007

    Can only see half the document?

You must log on before posting a comment.

Are you a new visitor? Register Here

advertisement

advertisement

Windows is a trademark of the Microsoft group of companies. Windows IT Pro is used by Penton Media Inc. under license from owner.