To define policy rules based on user or group, first you create an LDAP server profile that defines how the firewall connects and authenticates to your directory server. The firewall supports a variety of directory servers, including Microsoft Active Directory (AD), Novell eDirectory, and Sun ONE Directory Server. The server profile also defines how the firewall searches the directory to retrieve the list of groups and the corresponding list of members. If you are using a directory server that is not natively supported by the firewall, you can integrate the group mapping function using the XML API. You can then create a group mapping configuration to
Map Users to Groups
Enable User- and Group-Based Policy.
Defining policy rules based on group membership rather than on individual users simplifies administration because you don’t have to update the rules whenever new users are added to a group. When configuring group mapping, you can limit which groups will be available in policy rules. You can specify groups that already exist in your directory service or define custom groups based on LDAP filters. Defining custom groups can be quicker than creating new groups or changing existing ones on an LDAP server, and doesn’t require an LDAP administrator to intervene. User-ID maps all the LDAP directory users who match the filter to the custom group. For example, you might want a security policy that allows contractors in the Marketing Department to access social networking sites. If no Active Directory group exists for that department, you can configure an LDAP filter that matches users for whom the LDAP attribute Department is set to Marketing. Log queries and reports that are based on user groups will include custom groups.
Knowing user and groups names is only one piece of the puzzle. The firewall also needs to know which IP addresses map to which users so that security rules can be enforced appropriately.
illustrates the different methods that are used to identify users and groups on your network and shows how user mapping and group mapping work together to enable user- and group-based security enforcement and visibility. The following topics describe the different methods of user mapping:
With server monitoring a User-ID agent—either a Windows-based agent running on a domain server in your network, or the integrated PAN-OS User-ID agent running on the firewall—monitors the security event logs for specified Microsoft Exchange Servers, Domain Controllers, or Novell eDirectory servers for login events. For example, in an AD environment, you can configure the User-ID agent to monitor the security logs for Kerberos ticket grants or renewals, Exchange server access (if configured), and file and print service connections. Note that for these events to be recorded in the security log, the AD domain must be configured to log successful account login events. In addition, because users can log in to any of the servers in the domain, you must set up server monitoring for all servers to capture all user login events. See
Configure User Mapping Using the Windows User-ID Agent
Configure User Mapping Using the PAN-OS Integrated User-ID Agent
In environments with multi-user systems—such as Microsoft Terminal Server or Citrix environments—many users share the same IP address. In this case, the user-to-IP address mapping process requires knowledge of the source port of each client. To perform this type of mapping, you must install the Palo Alto Networks Terminal Services Agent on the Windows/Citrix terminal server itself to intermediate the assignment of source ports to the various user processes. For terminal servers that do not support the Terminal Services agent, such as Linux terminal servers, you can use the XML API to send user mapping information from login and logout events to User-ID. See
Configure User Mapping for Terminal Server Users
for configuration details.
User-ID can read the IPv4 or IPv6 addresses of users from the X-Forwarded-For (XFF) header in HTTP client requests when the firewall is deployed between the Internet and a proxy server that would otherwise hide the user IP addresses. User-ID matches the true user IP addresses with usernames. See
Configure the firewall to obtain the user IP address from the X-Forwarded-For (XFF) header.
If the firewall or the User-ID agent can’t map an IP address to a username—for example, if the user isn’t logged in or uses an operating system such as Linux that your domain servers don’t support—you can configure Captive Portal. Any web traffic (HTTP or HTTPS) that matches a Captive Portal policy rule requires user authentication. You can base the authentication on a transparent browser-challenge (
Kerberos Single Sign-On (SSO)
NT LAN Manager (NTLM)
authentication), web form (for RADIUS, TACACS+, LDAP, Kerberos, or local database authentication), or client certificates. For details, see
Map IP Addresses to Usernames Using Captive Portal.
Your environment might have existing network services that authenticate users. These services include wireless controllers, 802.1x devices, Apple Open Directory servers, proxy servers, and other Network Access Control (NAC) mechanisms. You can configure these services to send syslog messages and configure the User-ID agent to parse the messages for login events. The agent then maps IP addresses to usernames based on the login events.
Both the PAN-OS integrated User-ID agent and Windows-based User-ID agent use Syslog Parse profiles to parse syslog messages. In environments where services send the messages in different formats, you can create a custom profile for each format. If you use the PAN-OS integrated User-ID agent, you can also use predefined Syslog Parse profiles that Palo Alto Networks provides through Applications content updates.
Configure User-ID to Receive User Mappings from a Syslog Sender
for configuration details.
For mobile or roaming users, the GlobalProtect client provides the user mapping information to the firewall directly. In this case, every GlobalProtect user has an agent or app running on the client that requires the user to enter login credentials for VPN access to the firewall. This login information is then added to the User-ID user mapping table on the firewall for visibility and user-based security policy enforcement. Because GlobalProtect users must authenticate to gain access to the network, the IP address-to-username mapping is explicitly known. This is the best solution in sensitive environments where you must be certain of who a user is in order to allow access to an application or service. For more information on setting up GlobalProtect, refer to the
GlobalProtect Administrator’s Guide.
Captive Portal and the other standard user mapping methods might not work for certain types of user access. For example, the standard methods cannot add mappings of users connecting from a third-party VPN solution or users connecting to a 802.1x-enabled wireless network. For such cases, you can use the PAN-OS XML API to capture login events and send them to the PAN-OS integrated User-ID agent . See
Send User Mappings to User-ID Using the XML API
In a Microsoft Windows environment, you can configure the User-ID agent to probe client systems using Windows Management Instrumentation (WMI) and/or NetBIOS probing at regular intervals to verify that an existing user mapping is still valid or to obtain the username for an IP address that is not yet mapped.
Client probing was designed for legacy networks where most users were on Windows workstations on the internal network, but is not ideal for today’s more modern networks that support a roaming and mobile user base on a variety of devices and operating systems. Additionally, client probing can generate a large amount of network traffic (based on the total number of mapped IP addresses) and can pose a security threat when misconfigured. Therefore, client probing is no longer a recommended method for user mapping. Instead collect user mapping information from more isolated and trusted sources, such as domain controllers and through integrations with
XML API, which allow you to safely capture user mapping information from any device type or operating system. If you have sensitive applications that require you to know exactly who a user is, configure
to ensure that you are only allowing access to authorized users.
If you do choose to enable probing in your trusted zones, the agent will probe each learned IP address periodically (every 20 minutes by default, but this is configurable) to verify that the same user is still logged in. In addition, when the firewall encounters an IP address for which it has no user mapping, it will send the address to the agent for an immediate probe.