- Overview
- Using the aregcmd Commands
- Using the Graphical User Interface
- Cisco Prime Access Registrar Server Objects
- Using the radclient Command
- Configuring Local Authentication and Authorization
- RADIUS Accounting
- Diameter
- Extensible Authentication Protocols
- Using WiMAX in Cisco Prime Access Registrar
- Using Extension Points
- Using Replication
- Using On-Demand Address Pools
- Using Identity Caching
- Using Trusted ID Authorization with SESM
- Using Prepaid Billing
- Using Cisco Prime Access Registrar Server Features
- Directing RADIUS Requests
- Using FastRules to Process Packet Flow
- Wireless Support
- Using LDAP
- Using Open Database Connectivity
- SIGTRAN-M3UA
- Using SNMP
- Enforcement of Licensing Models
- Backing Up the Database
- Using the REX Accounting Script
- Logging Syslog Messages
- Troubleshooting Cisco Prime Access Registrar
- Cisco Prime Access Registrar Tcl, REX, and Java Dictionaries
- Environment Dictionary
- RADIUS Attributes
- Support for REST API in Cisco Prime Access Registrar
Cisco Prime Access Registrar 7.1 User Guide
Bias-Free Language
The documentation set for this product strives to use bias-free language. For the purposes of this documentation set, bias-free is defined as language that does not imply discrimination based on age, disability, gender, racial identity, ethnic identity, sexual orientation, socioeconomic status, and intersectionality. Exceptions may be present in the documentation due to language that is hardcoded in the user interfaces of the product software, language used based on RFP documentation, or language that is used by a referenced third-party product. Learn more about how Cisco is using Inclusive Language.
- Updated:
- February 7, 2018
Chapter: Extensible Authentication Protocols
Extensible Authentication Protocols
Cisco Prime Access Registrar (Prime Access Registrar) supports the Extensible Authentication Protocol (EAP) to provide a common protocol for differing authentication mechanisms. EAP enables the dynamic selection of the authentication mechanism at authentication time based on information transmitted in the Access-Request. (This type of EAP authentication mechanism is called an authentication exchange.)
Extensible Authentication Protocols (EAP) provide for support of multiple authentication methods. Cisco Prime Access Registrar supports the following EAP authentication methods:
- EAP-AKA
- EAP-AKA-Prime (EAP-AKA’)
- EAP-FAST
- EAP-GTC
- EAP-LEAP
- EAP-MD5
- EAP-Negotiate
- EAP-MSChapV2
- EAP-SIM
- EAP-Transport Level Security (TLS)
- EAP-TTLS
- Protected EAP
– PEAP Version 0 (Microsoft PEAP)
– PEAP Version 1 (Cisco PEAP)
In general, you enable each EAP method by creating and configuring a service of the desired type. Use the radclient test tool to confirm that the EAP service has been properly configured and is operational.
Both versions of Protected EAP (PEAP) are able to use other EAP methods as the authentication mechanism that is protected by PEAP encryption. For PEAP Version 0, the supported authentication methods are EAP-MSChapV2, EAP-SIM, EAP-TLS and EAP-Negotiate. For PEAP Version 1, the supported authentication methods are EAP-GTC, EAP-SIM, EAP-TLS and EAP-Negotiate.
The PEAP protocol consists of two phases: an authentication handshake phase and a tunnel phase where another complete EAP authentication exchange takes place protected by the session keys negotiated by phase one. Cisco Prime Access Registrar supports the tunneling of other EAP methods within the PEAP phase two exchange.
Prime Access Registrar supports rolling encryption, which involves generating rolling pseudonym secrets for EAP-SIM, EAP-AKA, and EAP-AKA’ services. For more details, see Rolling Encryption Support for Pseudonym Generation in EAP-SIM, EAP-AKA, and EAP-AKA’ Services.
EAP-AKA
Authentication and Key Agreement (AKA) is an EAP mechanism for authentication and session key distribution. It is used in the 3rd generation mobile networks Universal Mobile Telecommunications System (UMTS) and CDMA2000. AKA is based on symmetric keys, and typically runs in a UMTS Subscriber Identity Module (USIM), or a (Removable) User Identity Module ((R) UIM), similar to a smart card. EAP-AKA (Extensible Authentication Protocol Method for UMTS Authentication and Key Agreement) includes optional identity privacy support, optional result indications, and an optional fast reauthentication procedure.
In support of EAP-AKA, the following features are supported:
- support of MAP protocol over SIGTRAN
- support of HLR and/or HSS (3GPP compliant)
- Wx interface
- Support M3UA-SIGTRAN over IP
For more information on Wx Interface Support, see the Wx Interface Support for SubscriberDB Lookup.
Prime Access Registrar server supports migration to a converged IP Next Generation Networks (IP NGN) by supporting SS7 and SIGTRAN (SS7 over IP) for HLR communication to enable the seamlessly transition to next-generation IP-based signaling networks.
Prime Access Registrar supports M3UA-SIGTRAN to fetch the authentication vectors from HLR for EAP-AKA authentication, See SIGTRAN-M3UA for more information.
EAP-AKA is based on rfc-4187 (http://www.ietf.org/rfc/rfc4187.txt). This document specifies the details of the algorithms and messages.
This section contains the following topics:
Configuring EAP-AKA
You can use aregcmd to create and configure a service of type eap-aka.
Table 9-1 lists and describes the EAP-AKA service properties.
When True, enables the server to obtain the subscriber's identity via EAP/AKA messages instead of relying on the EAP messages alone. This might be useful in cases where intermediate software layers can modify the identity field of the EAP-Response/Identity message. The default value is False. |
|
When True, the identity privacy feature is enabled. The default value is False. |
|
The secret string that is used as the basis for protecting identities when identity privacy is enabled. This should be at least 16 characters long and have a value that is impossible for an outsider to guess. The default value is secret. This field is not available if the EnableRollingPseudonymSecret field is checked. Note It is very important to change PseudonymSecret from its default value to a more secure value when identity privacy is enabled for the first time. |
|
Specifies the maximum age a pseudonym can have before it is renewed. When the server receives a valid pseudonym that is older than this, it generates a new pseudonym for that subscriber. The value is specified as a string consisting of pairs of numbers and units, where the units might be of the following: M, Minute, Minutes, H, Hour, Hours, D, Day, Days, W, Week, Weeks. The default value is "24 Hours". Examples are: "8 Hours", "10 Hours 30 Minutes", "5 D 6 H 10 M" |
|
Specifies the maximum age a pseudonym can have before it is rejected by the server, forcing the subscriber to authenticate using it's permanent identity. The value is specified as a string consisting of pairs of numbers and units, where the units might be one of the following: M, Minute, Minutes, H, Hour, Hours, D, Day, Days, W, Week, Weeks. It can also be Forever, in which case, pseudonyms do not have a maximum age. The default value is "Forever". Examples are: "Forever", "3 Days 12 Hours 15 Minutes", "52 Weeks" |
|
(Optional); Notification service is an authorization service and is used to send a notification code to the client in case of an authorization failure. For more information about the Notification-Code variable, see Appendix B, “Environment Dictionary.” This is applicable for RADIIUS and Diameter and can be any of the services configured under /radius/services/ except eap services, accounting services, radius-session, radius-query, and diameter. |
|
When True, the fast reauthentication option is enabled. The default value is False. |
|
Default value is FALSE. When set to TRUE, Prime Access Registrar drops or rejects reauthentication requests as per outage policy when the remote server is down. This can be processed only when there is at least one failed full authentication before proceeding with reauthentication. |
|
Specifies the maximum number of times a reauthentication identity might be reused before it must be renewed. The default value is 16. |
|
Specifies the time in seconds that reauthentication identities are cached by the server. Subscribers that attempt to reauthenticate using identities that are older than this value will be forced to use full authentication instead. The default value is 3600 (one hour). |
|
Optional. If you configure the realm, this value is appended to the FastReauthenticationUserId. |
|
Time in seconds to wait for authentication to complete. The default is 2 minutes; range is 10 seconds to 10 minutes. |
|
Optional. If the script is set, the custom scripting point can be used to read the quintets from a flat file or generate quintets instead of fetching the quintets from HLR.If the script is not set, the Prime Access Registrar sends the request to HLR configured in remote server to fetch the quintets. |
|
Enables or disables the use of protected results messages. Results messages indicate the state of the authentication but are cryptographically protected. |
|
Required. Must be set to either DIAMETER or SIGTRAN-M3UA. When set to DIAMETER, the HSS lookup happens using the Diameter Wx Interface. You need to configure the DestinationRealm to send the Diameter packets to the RemoteServer. When set to SIGTRAN-M3UA, the HLR/HSS lookup happens using the SIGTRAN protocol. You need to configure the SIGTRAN remote server. |
|
Required. When set True, it fetches MSISDN from HLR. |
|
Optional script Prime Access Registrar server runs when it receives a request from a client for an EAP-AKA/EAP-SIM service. |
|
Optional script Prime Access Registrar server runs before it sends a response to a client using an EAP-AKA/EAP-SIM service. |
|
Optional. If set to the name of a script, Prime Access Registrar runs the script when an outage occurs. This property allows you to create a script that notifies you when the server detects a failure. |
|
Check this box to activate rolling encryption process that involves generating rolling pseudonym secrets for the service. For more information about rolling encryption support, see Rolling Encryption Support for Pseudonym Generation in EAP-SIM, EAP-AKA, and EAP-AKA’ Services. |
To enable EAP-AKA authentication:
Step 1 Launch aregcmd and create an EAP-AKA service.
Step 2 Change directory to the service and set its type to eap-aka.
The following example shows the default configuration for an EAP-AKA service:
The following example shows the default configuration for an EAP-AKA Wx service:
Testing EAP-AKA with radclient
To test the EAP-AKA service, launch radclient and use the simple_eap_aka_test command. The simple_eap_aka_test command sends an Access-Request for the designated user with the user's secret key and sequence number.
The response packet should indicate an Access-Accept if authentication was successful. View the response packet to ensure the authentication was successful.
simple_eap_aka_test bob secret 2
To test from radclient, you have to configure /cisco-ar/conf/imsi.conf file on radius server and reload the server. This file content should have imsi users in the format below:
EAP-AKA-Prime (EAP-AKA’)
EAP-AKA-Prime (EAP-AKA') is a new EAP authentication method, with a small revision to the existing EAP-AKA method. EAP- AKA' has a new key derivation function, which binds the keys derived within the method to the name of the access network. This limits the effects of compromised access network nodes and keys.
EAP- AKA' is similar to EAP-AKA in all aspects except the following:
- Key derivation involves an AT_KDF_INPUT attribute, which is mapped to the NetworkName attribute, and an AT_KDF attribute, which takes the key derivation function in the configuration, to ensure that the peer and the server know the name of the access network.
- EAP-AKA' employs SHA-256 (Secure Hash Algorithm) instead of SHA-1 as used in EAP-AKA, to ensure more security.
EAP-AKA' is based on rfc-5448 (http://www.ietf.org/rfc/rfc5448.txt). This document specifies the details of the algorithms and messages.
This section contains the following topics:
Configuring EAP-AKA’
You can use aregcmd to create and configure a service of type eap-aka-prime. EAP-AKA' service has the following attribute in addition to the service properties listed in Table 9-1 .
To enable EAP-AKA' authentication:
Step 1 Launch aregcmd and create an EAP-AKA' service.
Step 2 Change directory to the service and set its type to eap-aka-prime.
The following example shows the default configuration for an EAP-AKA’ service:
Testing EAP-AKA’ with radclient
To test the EAP-AKA’ service, launch radclient and use the simple_eap_aka_prime_test command. The simple_eap_aka_prime_test command sends an Access-Request for the designated user with the user's secret key and sequence number.
The response packet should indicate an Access-Accept if authentication was successful. View the response packet to ensure the authentication was successful.
simple_eap_aka_prime_test bob secret 2
To test from radclient, you have to configure /cisco-ar/conf/imsi.conf file on radius server and reload the server. This file content should have imsi users in the format below:
EAP-FAST
Cisco Prime Access Registrar supports the EAP-FAST authentication method. EAP-FAST uses the EAP-MSChapV2 method for credential provisioning and EAP-GTC for authentication. Credential provisioning typically occurs only during the client’s initial EAP-FAST authentication. Subsequent authentications rely on the provisioned credential and will usually omit the provisioning step.
EAP-FAST is an authentication protocol designed to address the performance shortcomings of prior TLS-based EAP methods while retaining features such as identity privacy and support for password-based protocols. The EAP-FAST protocol is described by the IETF draft draft-cam-winget-eap-fast-00.txt .
The EAP-FAST credential is known as a Protected Access Credential (PAC) and contains information used to secure the authentication operations. Parts of the PAC are encrypted by the server and are not visible to other entities. Clients are expected to securely store PACs locally for use during authentication.
Configuring EAP-FAST involves creating and configuring the required EAP-MSChapV2 and EAP-GTC services as well as the EAP-FAST service with the appropriate parameters.
You can use the radclient test tool to confirm that the EAP services are properly configured and operational.
This section contains the following topics:
- Configuring EAP-FAST
- EAP-FAST Keystores
- Testing EAP-FAST with radclient
- Parameters Used for Certificate-Based Authentication
- PAC—Credential Export Utility
Configuring EAP-FAST
You can use aregcmd to create and configure a service of type eap-fast .
Step 1 Launch aregcmd and create an EAP-FAST service.
Step 2 Change directory to the service and set its type to eap-fast.
Step 3 Set the AuthorityIdentifier:
set AuthorityIdentifier authority-identifier
Step 4 : Set the AuthorityInformation:
set AuthorityInformation authority-information
Step 5 : Set the AuthentitcationService:
set AuthenticationService eap-gtc-service
Step 6 :Set the ProvisionService:
set ProvisionService eap-mschapv2-service
The follow example shows the default configuration for an EAP-FAST service:
Table 9-2 lists and describes the EAP-FAST service properties.
Optional script Prime Access Registrar server runs when it receives a request from a client for EAP-FAST service. |
|
Optional script Prime Access Registrar server runs before it sends a response to a client using EAP-FAST. |
|
A string that uniquely identifies the credential (PAC) issuer. The client uses this value to select the correct PAC to use with a particular server from the set of PACs it might have stored locally. Ensure that the AuthorityIdentifier is globally unique and that it does not conflict with identifiers used by other EAP-FAST servers or PAC issuers. |
|
A string that provides a descriptive text for this credential issuer. The value can be displayed to the client for identification purposes and might contain the enterprise or server names. |
|
Indicates the maximum length in bytes that a PEAP or EAP-TLS message can have before it is fragmented. |
|
The full pathname of the file containing the server’s certificate or certificate chain used during the TLS exchange. The pathname can be optionally prefixed with a special string that indicates the type of encoding used for the certificate. The two valid encoding prefixes are PEM and DER. If an encoding prefix is not present, the file is assumed to be in PEM format. |
|
The full pathname of the file containing the server’s RSA or ECC private key. The pathname can be optionally prefixed with a special string that indicates the type of encoding used for the certificate. The two valid encoding prefixes are “PEM” and “DER”. If an encoding prefix is not present, the file is assumed to be in PEM format. The following example assumes that the subdirectory pki under /cisco-ar contains the server’s certificate file. The file server-key.pem is assumed to be in PEM format. The file extension .pem is not significant. |
|
The full pathname of the file containing trusted CA certificates used for client verification. The file can contain more than one certificate, but all certificates must be in PEM format. DER encoding is not allowed. |
|
The name of a directory containing trusted CA certificates (in PEM format) used for client verification. This parameter is optional, and if it is used there are some special preparations required for the directory it references. Each certificate file in this directory must contain exactly one certificate in PEM format. The server looks up the certificate files using the MD5 hash value of the certificate’s subject name as a key. The directory must therefore also contain a set of symbolic links each of which points to an actual certificate file. The name of each symbolic link is the hash of the subject name of the certificate. For example, if a certificate file named ca-cert.pem is located in the CACertificatePath directory, and the MD5 hash of the subject name contained in ca-cert.path.pem is 1b96dd93, then a symbolic link named 1b96dd93 must point to ca-cert.pem . If there are subject name collisions such as multiple certificates with the same subject name, each link name must be indexed with a numeric extension as in 1b96dd93.0 and 1b96dd93.1. |
|
Optional. Enter the URL that Prime Access Registrar should use to retrieve the CRL.You can specify a URL that uses HTTP or LDAP.
The following is an example for an HTTP URL:
The following is an example for an LDAP URL: |
|
Specifies the type of verification used for client certificates. Must be set to one of RequireCertificate, None, or Optional. |
|
Specifies the maximum length of the certificate chain used for client verification. |
|
Determines the applicability of the authentication mechanism in SmartGrid Solutions, see the Smart Grid Solution Management for more information. When UseECCCertificates is set to True, it can use the ECC, RSA, or combination of both certificate for certificate based verification. When UseECCCertificates is set to False, it can only use the RSA certificate for certificate based verification. The default location to fetch the certificate file is /cisco-ar/pki. |
|
Specifies whether TLS session caching (fast reconnect) is enabled or not. Set to True to enable session caching; otherwise set to False. |
|
If TLS session caching (fast reconnect) is enabled, SessionTimeout specifies the maximum lifetime of a TLS session. Expired sessions are removed from the cache and will require a subsequent full authentication. SessionTimeout is specified as a string consisting of pairs of numbers and units, where units might be one of the following: M, Minute, Minutes, H, Hour, Hours, D, Day, Days, W, Week, Weeks, as in the following: |
|
Mandatory; specifies time (in seconds) to wait before an authentication request times out; defaults to 120. |
|
Specifies the maximum lifetime of a Protected Access Credential (PAC). Clients that successfully authenticate with an expired PAC will be reprovisioned with a new PAC. CredentialLifetime is specified as a string consisting of pairs of numbers and units, where units might be one of the following: M, Minute, Minutes, H, Hour, Hours, D, Day, Days, W, Week, Weeks. Credentials that never expire should be specified as Forever. |
|
Specifies the name of the EAP-GTC service is used for authentication. The named service must have the UseLabels parameter set to True. |
|
Specifies the TLS mode used for provisioning. Clients only support the default Anonymous mode. |
|
Specifies the name of the EAP-MSChapV2 service used for provisioning. |
|
Indicates whether provisioning should always automatically rollover into authentication without relying on a separate session. Most environments, particularly wireless, will perform better when this parameter is set to True, the default value. |
Note Prime Access Registrar verifies the certificate during the TLS-based authentication. CRL validation is done before accepting a client certificate during the TLS authentication.
EAP-FAST Keystores
The EAP-FAST service manages a set of keys used to protect the security and integrity of the PACs it issues. The keys are stored in /Radius/Advanced/KeyStores/EAP-FAST and are maintained automatically requiring minimal administration. Administrators can specify the maximum number of keys that are stored and the frequency of key updates.
The following is the default KeyStores settings:
Table 9-3 defines the KeyStores properties.
Testing EAP-FAST with radclient
There are two distinct phases to testing EAP-FAST: provisioning and authentication. In the instructions below, Step 2 and Step 3 test provisioning and Steps 4 and Step 5 test authentication. At least one successful provisioning phase must be completed prior to testing authentication. Testing EAP-FAST with radclient requires that the EAP-MSChapV2 and EAP-GTC services be configured and functional.
The following instructions and examples assume that the AlwaysAuthenticate parameter has been set to False for testing purposes. This permits the provisioning and authentication steps to be tested separately. Most installations will set AlwaysAuthenticate to True for production use, and radclient works with that setting, but might display extra error messages that you can ignore.
To test EAP-FAST using radclient :
Step 2 Specify the inner provisioning method
The only allowable method for provisioning is eap-mschapv2.
simple_eap_fast_test user-name password
Step 4 Specify the inner authentication method.
The only allowable method for authentication is eap-gtc.
Step 5 Authenticate using the PAC.
simple_eap_fast_test user-name password
The simple_eap_fast_test command passes its arguments to the inner authentication mechanism which in turn treats the arguments as a username and a password. The command in Step 3 should result in provisioning a new PAC, and Step 5 should result in successful authentication using that PAC.
PAC Provisioning
The following example provisions a PAC for user bob.
In this example the simple_eap_fast_test command indicates that it did not receive an AccessAccept. This is normal because the provisioning step always results in an AccessReject even when a new PAC has been successfully provisioned. The last pac show command displayed some status information from the new PAC and is used to verify that provisioning succeeded and authentication can now be tested. The PAC information displayed will vary and depends on how EAP-FAST is configured.
Authentication
The following example authenticates user bob (continuing from the PAC Provisioning example).
In this example, the EAP_FAST authentication using the PAC from the previous provisioning step succeeded. The AccessAccept packet received from Prime Access Registrar can be displayed to confirm that it contains the expected attributes including the MS-MPPE session keys.
Parameters Used for Certificate-Based Authentication
EAP-FAST might optionally use RSA or ECC certificates to securely create the tunnel that is used for PAC provisioning. However, the Cisco client does not support the use of certificates and the following parameters will be ignored and should be left at their default values:
- PrivateKeyPassword
- ServerCertificateFile
- ServerKeyFile
- CACertificateFile
- CACertificatePath
- ClientVerificationMode
- VerificationDepth
- UseECCCertificates
- EnableSessionCache
- SessionTimeout
The parameters for configuring certificate-based operation are identical to those used for PEAP and EAP-TLS.
Table 9-4 describes the parameters used for certificate-based authentication.
radclient Command Reference
This section describes the radclient commands you can use to test EAP-FAST.
eap-trace
Use the eap-trace command to display additional client protocol trace information for EAP methods. Level is a number from 1 to 5 inclusively. Level 5 shows detailed hex dumps of all messages, level 4 shows a message trace without hex dumps, and levels 3 and below show status and error information. To turn off trace displays, set the level to 0.
Set the trace level for all EAP methods.
For example, the following command sets the trace level to 4 for all EAP methods.
Set the trace level for the specified EAP method.
The following example sets the trace level to 5 for EAP-FAST only. The trace level for other EAP methods is not affected.
Note The eap-trace command is for client-side trace information only and is independent of the server trace level that can be set using aregcmd.
tunnel
The tunnel command is used to specify the inner provisioning and authentication methods for EAP-FAST. The specified EAP method type must agree with the server’s configured methods or authentication will fail.
For EAP-FAST provisioning, the only allowable tunnel method is eap-mchavp2. For EAP-FAST authentication, the only allowable tunnel method is eap-gtc.
simple_eap_fast_test
The arguments are passed to the inner authentication method as its authentication parameters. If a PAC is not present, the tunnel method should be eap-mschapv2 and provisioning will occur. If a PAC is present, the tunnel method should be eap-gtc and authentication will occur.
simple_eap_fast_test username password
There are also variants for the simple test command for other EAP methods as shown in the following examples:
pac
The pac command is used display, save, and delete PACs that are received from the server during testing. radclient maintains a cache of PACs that it knows about and that can be used for authentication testing. The current PAC cache can be displayed with the pac show command. PACs created during a test session can be stored to files with the pac save command, and reloaded in another session with the pac load command. The contents of the PAC cache are completely deleted with pac delete . If the optional parameter cache is included, PACs are also erased from disk.
pac load | save | show { hex } | delete { cache }
The pac show command displays the currently cached PACs. If the optional parameter hex is included, additional detailed information including hex dumps are included in the display output.
The pac load command loads any previously saved PACS from disk into the active cache.
The pac save command saves all PACs from the active cache to disk. Any previously existing PACS for the same user will be over-written.
The pac delete command deletes all PACs from the active cache. If the optional cache parameter is included then PACs are also erased from disk.
PAC—Credential Export Utility
You can manually provision EAP-FAST PACs to clients and avoid the use of the protocol provisioning phase. This might be desirable from a security perspective since the default provisioning protocol uses an anonymous (unauthenticated) method to construct the tunnel used to download the PAC to the client.
Manual provisioning involves exporting a PAC from Prime Access Registrar to a file which is then copied to the client machine and used by the import utility. After a PAC has been manually imported, the client should be able to authenticate via EAP-FAST while bypassing the initial provisioning phase. Care should be taken while storing and transporting PAC files since they contain information that potentially allows a client to authenticate via EAP-FAST.
PACs are exported from Prime Access Registrar via the pac command which is a new utility for this release. (Note that this pac command is a standalone executable which is different from the Radclient pac command.) The pac command has two capabilities:
PAC Export
Use the pac export command to create a new PAC file. In the following example, eap-fast is the name of the Prime Access Registrar service configured for EAP-FAST authentication, bob is the name of the user this PAC will be used for, and password is the password used to derive a key for encrypting the resulting file. (This password is not the same as the administrator’s password). The PAC file will be named bob.pac by default. You can use the –f option to give the file a different name.
pac –s export eap-fast bob password
If you omit the password parameter, a default password will be used.
Note Using the default password is strongly discouraged for security reasons.
PAC Display
Use the pac show command to display information about a PAC file. In the following example, bob.pac is the name of the PAC file and password is the password used to decrypt the file contents.
Syntax Summary
The complete pac command syntax is as follows:
pac { options } export <service-name> <user-name> <file-password>
pac { options } show <file-name> file-<password>
-C <cluster> —Specifies the cluster to be used.
-N <user> —Specifies the user.
EAP-GTC
EAP-GTC, defined in RFC 2284, is a simple method for transmitting a user’s name and password to an authentication server. EAP-GTC should not be used except as an authentication method for PEAP Version 1 because the password is not protected.
This section contains the following topics:
Configuring EAP-GTC
Table 9-5 lists and describes the EAP-GTC specific properties for EAP-GTC authentication.
To enable EAP-GTC, use aregcmd to create and configure a service of type eap-gtc
Step 1 Launch aregcmd and create an EAP-GTC service.
Step 2 Change directory to the service and set its type to eap-gtc.
The follow example shows the default configuration for an EAP-GTC service:
Step 3 Set the service’s UserService to local-users or another local authentication service that is able to authenticate using clear-text passwords.
Step 4 If configuring for EAP-FAST, set the UseLabels property to TRUE.
Testing EAP-GTC with radclient
To test the EAP-GTC service, launch radclient and use the simple_eap_gtc_test command. The simple_eap_gtc_test command sends an Access-Request for the designated user with the user’s password.
The response packet should indicate an Access-Accept if authentication was successful. View the response packet to ensure the authentication was successful.
EAP-LEAP
Prime Access Registrar supports the new AAA Cisco-proprietary protocol called Light Extensible Authentication Protocol (LEAP), a proprietary Cisco authentication protocol designed for use in IEEE 802.11 wireless local area network (WLAN) environments. Important features of LEAP include:
- Mutual authentication between the network infrastructure and the user
- Secure derivation of random, user-specific cryptographic session keys
- Compatibility with existing and widespread network authentication mechanisms (e.g., RADIUS)
- Computational speed
Note Prime Access Registrar supports a subset of EAP to support LEAP. This is not a general implementation of EAP for Prime Access Registrar.
The Cisco-Wireless or Lightweight Extensible Authentication Protocol is an EAP authentication mechanism where the user password is hashed based on an MD4 algorithm and verified by a challenge from both client and server.
Configuring EAP-LEAP
You can use aregcmd to create and configure a service of type eap-leap . When you create an EAP-LEAP service type, you must also specify a UserService to perform AAA service. The UserService can be any configured authentication service.
Step 1 Launch aregcmd and create an EAP-LEAP service.
Step 2 Set the service type to eap-leap .
Step 3 Set the UserService property to a configured authentication service.
EAP-MD5
Cisco Prime Access Registrar supports EAP-MD5, or MD5-Challenge, another EAP authentication exchange. In EAP-MD5 there is a CHAP-like exchange and the password is hashed by a challenge from both client and server to verify the password is correct. After verified correct, the connection proceeds, although the connection is periodically re-challenged (per RFC 1994).
Configuring EAP-MD5
Specify type eap-md5 when you create an EAP-MD5 service. When you create an EAP-MD5 service type, you must also specify a UserService to perform AAA service. The UserService can be any configured authentication service.
You can use aregcmd to create and configure a service of type eap-md5 . When you create an EAP-MD5 service type, you must also specify a UserService to perform AAA service. The UserService can be any configured authentication service.
Step 1 Launch aregcmd and create an EAP-LEAP service.
Step 2 Set the service type to eap-md5 .
Step 3 Set the UserService property to a configured authentication service.
EAP-Negotiate
EAP-Negotiate is a special service used to select at runtime the EAP service to be used to authenticate the client. EAP-Negotiate is configured with a list of candidate EAP services that represent the allowable authentication methods in preference order. When an EAP session begins, the EAP-Negotiate service tires the first service in the list. If the client does not support that method, it will respond with an EAP-Nak message which triggers EAP-Negotiate to try the next method on the list until a valid method is found or the list is exhausted in which case authentication fails.
EAP-Negotiate is useful when the client population has deployed a mix of different EAP methods that must be simultaneously supported by Prime Access Registrar. It can be difficult or impossible to reliably distinguish which clients require which methods simply by examining RADIUS attributes or other packet properties. EAP-Negotiate solves this problem by using the method negotiation feature of the EAP protocol. Negotiation can be used to select the primary EAP method used for authentication and also to select the inner method for PEAP.
This section contains the following topics:
Configuring EAP-Negotiate
You may first use aregcmd to create and configure the EAP services that will be used for authentication, then create and configure a service of type eap-negotiate.
Step 1 Launch aregcmd and create an EAP-LEAP service.
Step 2 Set the service type to eap-negotiate .
Step 3 Set the ServiceList property to a list of preconfigured EAP authentication services.
The ServiceList property lists the names of the EAP services that can be negotiated with this instance of EAP-Negotiate. The ServiceList property is a space-separated list and must consist of valid EAP service name, not service types , in preference order from left to right. Each service and type on the list must be unique; duplicates are not allowed.
set ServiceList “eap-leap-service eap-md5-service peap-v1-service”
Negotiating PEAP Tunnel Services
EAP-Negotiate can also be used to negotiate the inner tunnel service used for phase two of PEAP-V0 or PEAP-V1. To do this, create and configure a service of type eap-negotiate. The ServiceList can only contain services that are legal for the version of PEAP that it is used with. Set the PEAP service’s TunnelService parameter to the name of the eap-negotiate service.
Note Not all supplicants support negotiation of the PEAP inner method. EAP-Negotiate can only be used with supplicants that can use EAP-Nak to reject an unsupported inner method.
EAP-MSChapV2
EAP-MSChapv2 is based on draft-kamath-pppext-eap-mschapv2-00.txt , an informational IETF draft document. EAP-MSChapv2 encapsulates the MSChapV2 protocol (specified by RFC 2759) and can be used either as an independent authentication mechanism or as an inner method for PEAP Version 0 (recommended).
This section contains the following topics:
Configuring EAP-MSChapV2
To enable EAP-MSChapv2, use aregcmd to create and configure a service of type eap-mschapv2
Step 1 Launch aregcmd and create an EAP-MSChapV2 service.
Note This example named the service eap-mschapv2, but you can use any valid name for your service.
Step 2 Set the service’s type to eap-mschapv2.
Step 3 Set the service’s UserService to local-users or another local authentication service that is able to authenticate using MSChapV2.
Step 4 You might (optionally) set a string for System ID that identifies the sender of the MSChapV2 challenge message, as in the following:
Testing EAP-MSChapV2 with radclient
To test the EAP-MSChapVersion 2 service using radclient :
Step 2 Use the simple_eap_mschapv2_test command to authenticate using EAP-MSChapV2, as in the following:
simple_eap_mschapv2_test bob bob
The simple_eap_mschapv2_test command above sends an Access-Request for user bob with the user’s password. The response packet should indicate an Access-Accept if authentication was successful.
Step 3 View the response packet to ensure the authentication was successful.
EAP-SIM
Cisco Prime Access Registrar supports EAP-SIMv16. In a GSM network a subscriber is issued a smart card called the subscriber identity module (SIM) that contains a secret key (Ki) and an International Mobile Subscriber Identity (IMSI). The key (Ki) is also stored in the GSM authentication center located with the Home Location Registry (HLR).
An access point uses the Prime Access Registrar RADIUS server to perform EAP-SIM authentication of mobile clients. Prime Access Registrar must obtain authentication information from the HLR. Prime Access Registrar contacts the MAP gateway that performs the MAP protocol over SS7 to the HLR, see SIGTRAN-M3UA for more information.
In support of EAP-SIM, the Wx Interface feature will be supported. For more information on Wx Interface Support, see the Wx Interface Support for SubscriberDB Lookup.
Configuring EAP-SIM
You can use aregcmd to create and configure a service of type eap-sim .
Table 9-6 lists and describes the EAP-SIM specific properties.
When True, enables the server to obtain the subscriber's identity via EAP/AKA messages instead of relying on the EAP messages alone. This might be useful in cases where intermediate software layers can modify the identity field of the EAP-Response/Identity message. The default value is False. |
|
When True, the identity privacy feature is enabled. The default value is False. |
|
The secret string that is used as the basis for protecting identities when identity privacy is enabled. This should be at least 16 characters long and have a value that is impossible for an outsider to guess. The default value is secret. This field is not available if the EnableRollingPseudonymSecret field is checked. Note It is very important to change PseudonymSecret from its default value to a more secure value when identity privacy is enabled for the first time. |
|
Specifies the maximum age a pseudonym can have before it is renewed. When the server receives a valid pseudonym that is older than this, it generates a new pseudonym for that subscriber. The value is specified as a string consisting of pairs of numbers and units, where the units might be of the following: M, Minute, Minutes, H, Hour, Hours, D, Day, Days, W, Week, Weeks. The default value is "24 Hours". Examples are: "8 Hours", "10 Hours 30 Minutes", "5 D 6 H 10 M" |
|
Specifies the maximum age a pseudonym can have before it is rejected by the server, forcing the subscriber to authenticate using it's permanent identity. The value is specified as a string consisting of pairs of numbers and units, where the units might be one of the following: M, Minute, Minutes, H, Hour, Hours, D, Day, Days, W, Week, Weeks. It can also be Forever, in which case, pseudonyms do not have a maximum age. The default value is "Forever". Examples are: "Forever", "3 Days 12 Hours 15 Minutes", "52 Weeks" |
|
(Optional); Notification service is an authorization service and is used to send a notification code to the client in case of an authorization failure. For more information about the Notification-Code variable, see Appendix B, “Environment Dictionary.” This can be any of the services configured under /radius/services/ except eap services, accounting services, radius-session, radius-query, and diameter. |
|
When True, the fast reauthentication option is enabled. The default value is False. |
|
Default value is FALSE. When set to TRUE, Prime Access Registrar drops or rejects reauthentication requests as per outage policy when the remote server is down. This can be processed only when there is at least one failed full authentication before proceeding with reauthentication. |
|
Specifies the maximum number of times a reauthentication identity might be reused before it must be renewed. The default value is 16. |
|
Specifies the time in seconds that reauthentication identities are cached by the server. Subscribers that attempt to reauthenticate using identities that are older than this value will be forced to use full authentication instead. The default value is 3600 (one hour). |
|
Optional. If you configure the realm, this value is appended to the FastReauthenticationUserId. |
|
Time in seconds to wait for authentication to complete. The default is 2 minutes; range is 10 seconds to 10 minutes. |
|
Optional. If the script is set, the custom scripting point can be used to read the quintets from a flat file or generate quintets instead of fetching the quintets from HLR.If the script is not set, the Prime Access Registrar sends the request to HLR configured in remote server to fetch the quintets. |
|
Enables or disables the use of protected results messages. Results messages indicate the state of the authentication but are cryptographically protected. |
|
Required. Must be set to either DIAMETER or SIGTRAN-M3UA. When set to DIAMETER, the HSS lookup happens using the Diameter Wx Interface. You need to configure the DestinationRealm to send the Diameter packets to the RemoteServer. When set to SIGTRAN-M3UA, the HLR/HSS lookup happens using the SIGTRAN protocol. You need to configure the SIGTRAN remote server. |
|
Required. When set True, it fetches MSISDN from HLR. |
|
Optional script Prime Access Registrar server runs when it receives a request from a client for an EAP-AKA/EAP-SIM service. |
|
Optional script Prime Access Registrar server runs before it sends a response to a client using an EAP-AKA/EAP-SIM service. |
|
Optional. If set to the name of a script, Prime Access Registrar runs the script when an outage occurs. This property allows you to create a script that notifies you when the server detects a failure. |
|
Check this box to activate rolling encryption process that involves generating rolling pseudonym secrets for the service. For more information about rolling encryption support, see Rolling Encryption Support for Pseudonym Generation in EAP-SIM, EAP-AKA, and EAP-AKA’ Services. |
To enable EAP-SIM authentication using aregcmd:
Step 1 Launch aregcmd and create an EAP-SIM service.
Step 2 Change directory to the service and set its type to eap-sim .
Note The EAP-SIM property OutagePolicy present in earlier versions of Prime Access Registrar is no longer part of the EAP-SIM configuration.
To enable EAP-SIM authentication using radclient :
Step 1 Create an EAP-SIM service.
Step 2 Change directory to the service and set its type to eap-sim .
Step 3 Execute the below command in radclient to set session keys in the server.
simple_eap_sim_test 987456321123654 secret
Note The IMSI number that is stored in HLR is used for EAP-SIM authentication.
Step 4 Enter the server name in which the session key is created to view the eap-sim service details.
MS-MPPE-Send-Key = 9c:56:e5:36:9f:fe:84:a2:26:16:80:0a:13:74:fb:b7:87:30:00:5c:45:99:ea:78:af:7d:ae:37:0e:b1:3a:2e:2b:b1:c8:4f:20:39:33:04:eb:dc:ba:27:e7:6f:56:08:21:56
MS-MPPE-Recv-Key = 8b:27:42:c5:47:79:ce:6a:41:ae:34:1f:15:2f:cf:b8:ee:18:e7:b5:1c:64:41:26:f7:4b:bc:53:bd:54:57:70:a3:3b:df:78:9e:34:33:47:b3:a2:ff:4e:f1:fe:6f:8f:ee:aa
Quintets to Triplets Conversion
Prime Access Registrar provides a configuration option in EAP-SIM service, which allows conversion of quintets received from a Universal Mobile Telecommunications Service (UMTS) subscriber to triplets. This feature facilitates backward compatibility by allowing to perform EAP-SIM authentication from an EAP-AKA or EAP-AKA’ source.
EAP-Transport Level Security (TLS)
EAP-Transport Level Security (EAP-TLS), described in RFC 2716, is an authentication method designed to mitigate several weaknesses of EAP. EAP-TLS leverages TLS, described in RFC 2246, to achieve certificate-based authentication of the server and (optionally) the client. EAP-TLS provides many of the same benefits as PEAP but differs from it in the lack of support for legacy authentication methods.
This section contains the following topics:
- Configuring EAP-TLS
- Configuring EAP-TLS with OCSP Support
- Testing EAP-TLS with RSA or ECC Certificate using radclient
- Testing EAP-TLS with Client Certificates
Configuring EAP-TLS
You can use aregcmd to create and configure a service of type eap-tls . Table 9-7 describes the EAP-TLS configuration properties:
Optional script Prime Access Registrar server runs when it receives a request from a client for EAP-TLS service |
|
Optional script Prime Access Registrar server runs before it sends a response to a client using EAP-TLS |
|
Indicates the maximum length in bytes that a PEAP or EAP-TLS message can have before it is fragmented. |
|
The full pathname of the file containing the server’s certificate or certificate chain used during the TLS exchange. The pathname can be optionally prefixed with a special string that indicates the type of encoding used for the certificate. The two valid encoding prefixes are PEM and DER. If an encoding prefix is not present, the file is assumed to be in PEM format. |
|
The full pathname of the file containing the server’s RSA or ECC (remove for Diameter) private key. The pathname can be optionally prefixed with a special string that indicates the type of encoding used for the certificate. The two valid encoding prefixes are “PEM” and “DER”. If an encoding prefix is not present, the file is assumed to be in PEM format. The following example assumes that the subdirectory pki under /cisco-ar contains the server’s certificate file. The file server-key.pem is assumed to be in PEM format. The file extension .pem is not significant. |
|
The full pathname of the file containing trusted CA certificates used for client verification. The file can contain more than one certificate, but all certificates must be in PEM format. DER encoding is not allowed. |
|
The name of a directory containing trusted CA certificates (in PEM format) used for client verification. This parameter is optional, and if it is used there are some special preparations required for the directory it references. Each certificate file in this directory must contain exactly one certificate in PEM format. The server looks up the certificate files using the MD5 hash value of the certificate’s subject name as a key. The directory must therefore also contain a set of symbolic links each of which points to an actual certificate file. The name of each symbolic link is the hash of the subject name of the certificate. For example, if a certificate file named ca-cert.pem is located in the CACertificatePath directory, and the MD5 hash of the subject name contained in ca-cert.path.pem is 1b96dd93, then a symbolic link named 1b96dd93 must point to ca-cert.pem . If there are subject name collisions such as multiple certificates with the same subject name, each link name must be indexed with a numeric extension as in 1b96dd93.0 and 1b96dd93.1. |
|
Optional. The URL that Prime Access Registrar should use to retrieve the CRL.You can specify a URL that uses HTTP or LDAP.
The following is an example for an HTTP URL:
The following is an example for an LDAP URL: |
|
Specifies the type of verification used for client certificates. Must be set to one of RequireCertificate, None, or Optional. |
|
Specifies the maximum length (in bytes?) of the certificate chain used for client verification. |
|
Determines the applicability of the authentication mechanism in SmartGrid Solutions, see the Smart Grid Solution Management for more information. When UseECCCertificates is set to True, it can use the ECC, RSA, or combination of both certificate for certificate based verification. When UseECCCertificates is set to False, it can only use the RSA certificate for certificate based verification. The default location to fetch the certificate file is /cisco-ar/pki. |
|
Specifies whether TLS session caching (fast reconnect) is enabled or not. Set to True to enable session caching; otherwise set to False. |
|
If TLS session caching (fast reconnect) is enabled, SessionTimeout specifies the maximum lifetime of a TLS session. Expired sessions are removed from the cache and will require a subsequent full authentication. SessionTimeout is specified as a string consisting of pairs of numbers and units, where units might be one of the following: M, Minute, Minutes, H, Hour, Hours, D, Day, Days, W, Week, Weeks, as in the following: |
|
Mandatory; specifies time (in seconds) to wait before an authentication request times out; defaults to 120. |
|
When set to TRUE, Prime Access Registrar sends its server certificate chain (Server-Cert -> IntermediateCA -> RootCA) while presenting the server certificate to the client for server side authentication. When set to FALSE, Prime Access Registrar sends only the server certificate (Server-Cert) to the client. |
To enable EAP-TLS authentication:
Step 1 Launch aregcmd and create an EAP-TLS service.
Step 2 Change directory to the service and set its type to eap-tls.
Note Prime Access Registrar verifies the certificate during the TLS-based authentication. CRL validation is done before accepting a client certificate during the TLS authentication.
Configuring EAP-TLS with OCSP Support
You can configure an EAP-TLS service to support Online Certificate Status Protocol (OCSP), which is used to check the status of X.509 digital certificates. This protocol can be used as an alternate to the certificate revocation list (CRL). For more information on CRL, see CRL Support for Cisco Prime Access Registrar.
Prime Access Registrar queries any number of OCSP servers to check the revocation status based on the URLs present in the incoming packet.
OCSP can return the following three values for a given certificate request:
- Good—The certificate is good for usage. This OCSP response is taken as a final response and Access-Accept will be sent to the client.
- Revoked—The certificate is revoked. This OCSP response is taken as a final response and Access-Reject will be sent to the client.
- Unknown —If the certificate status is unknown or if none of the OCSP servers respond, failover to CRL happens. In that case, response from CRL is considered as final and an Access-Accept or Access-Reject is sent to the client accordingly.
Table 9-8 describes the EAP-TLS configuration property with OCSP support:
Testing EAP-TLS with RSA or ECC Certificate using radclient
To test the EAP-TLS service, launch radclient and use the simple_eap_tls_test command, as in the following:
The argument is arbitrary for the simple_eap_tls_test command and can be anything. You can either select RSA or ECC client certificates using this argument.
Testing EAP-TLS with Client Certificates
You can test EAP-TLS using client certificates verified by the server during the TLS exchange. The client certificate file and RSA or ECC key file must reside in /cisco-ar/pki and be named client-cert.pem and client-key.pem respectively. Both files must be in PEM format.
EAP-TTLS
Prime Access Registrar supports the Extensible Authentication Protocol Tunneled TLS (EAP-TTLS). EAP-TTLS is an EAP protocol that extends EAP-TLS. In EAP-TLS, a TLS handshake is used to mutually authenticate a client and server. EAP- TTLS extends this authentication negotiation by using the secure connection established by the TLS handshake to exchange additional information between client and server.
EAP-TTLS leverages TLS (RFC 2246) to achieve certificate-based authentication of the server (and optionally the client) and creation of a secure session that can then be used to authentication the client using a legacy mechanism. EAP-TTLS provides several benefits:
- Industry standard authentication of the server using certificates (TLS)
- Standardized method for session key generation using TLS PRF
- Strong mutual authentication
- Identity privacy
- Fast reconnect using TLS session caching
- EAP message fragmentation
- Secure support for legacy client authentication methods
EAP-TTLS is a two-phase protocol. Phase 1 conducts a complete TLS session and derives the session keys used in Phase 2 to securely tunnel attributes between the server and the client. The attributes tunneled during Phase 2 can be used to perform additional authentication(s) via a number of different mechanisms.
The authentication mechanisms that can be used during Phase 2 include PAP, CHAP, MS-CHAP, MS-CHAPv2, and EAP. If the mechanism is EAP, then several different EAP methods are possible.
The Phase 2 authentication can be performed by the local AAA Server (the same server running EAP-TTLS) or it can be forwarded to another server (known as the home AAA Server). In the latter case, the home server has no involvement in the EAP-TTLS protocol and can be any AAA service that understands the authentication mechanism in use and is able to authenticate the user. It is not necessary for the home server to understand EAP-TTLS.
This section contains the following topics:
Configuring EAP-TTLS
Configuring EAP-TTLS involves two major tasks:
1. Configuring the TLS parameters used for Phase 1
2. Selecting the Phase 2 authentication methods and specifying whether authentication is performed locally or forwarded to the home server.
If authentication is forwarded, the configuration must include the identity of the remote home server and its shared secret.
You configure EAP-TTLS using the aregcmd CLI to create the appropriate services and specify their parameters. Use the radclient test tool to confirm that the services have been properly configured and are operational.
Creating an EAP-TTLS Service
You can use aregcmd to create and configure a service of type eap-ttls . Table 9-9 describes the EAP-TTLS configuration properties:
Optional script Prime Access Registrar server runs when it receives a request from a client for EAP-TTLS service. |
|
Optional script Prime Access Registrar server runs before it sends a response to a client using EAP-TTLS. |
|
Indicates the maximum length in bytes that a PEAP or EAP-TLS message can have before it is fragmented. |
|
The full pathname of the file containing the server’s certificate or certificate chain used during the TLS exchange. The pathname can be optionally prefixed with a special string that indicates the type of encoding used for the certificate. The two valid encoding prefixes are PEM and DER. If an encoding prefix is not present, the file is assumed to be in PEM format. |
|
The full pathname of the file containing the server’s RSA or ECC private key. The pathname can be optionally prefixed with a special string that indicates the type of encoding used for the certificate. The two valid encoding prefixes are “PEM” and “DER”. If an encoding prefix is not present, the file is assumed to be in PEM format. The following example assumes that the subdirectory pki under /cisco-ar contains the server’s certificate file. The file server-key.pem is assumed to be in PEM format. The file extension .pem is not significant. |
|
The full pathname of the file containing trusted CA certificates used for client verification. The file can contain more than one certificate, but all certificates must be in PEM format. |
|
The name of a directory containing trusted CA certificates (in PEM format) used for client verification. This parameter is optional, and if used, there are some special preparations required for the directory it references. Each certificate file in this directory must contain exactly one certificate in PEM format. The server looks up the certificate files using the MD5 hash value of the certificate’s subject name as a key. The directory must therefore also contain a set of symbolic links each of which points to an actual certificate file. The name of each symbolic link is the hash of the subject name of the certificate. For example, if a certificate file named ca-cert.pem is located in the CACertificatePath directory, and the MD5 hash of the subject name contained in ca-cert.path.pem is 1b96dd93, then a symbolic link named 1b96dd93 must point to ca-cert.pem . If there are subject name collisions such as multiple certificates with the same subject name, each link name must be indexed with a numeric extension as in 1b96dd93.0 and 1b96dd93.1. See rehash-ca-certs Utility for information about how to create the required certificate file hash links. |
|
Optional. The URL that Prime Access Registrar should use to retrieve the CRL.You can specify a URL that uses HTTP or LDAP.
The following is an example for an HTTP URL:
The following is an example for an LDAP URL: |
|
Specifies the type of verification used for client certificates. Must be set to one of RequireCertificate, None, or Optional. |
|
Specifies the maximum length of the certificate chain used for client verification. |
|
Determines the applicability of the authentication mechanism in SmartGrid Solutions, see the Smart Grid Solution Management for more information. When UseECCCertificates is set to True, it can use the ECC, RSA, or combination of both certificate for certificate based verification. When UseECCCertificates is set to False, it can only use the RSA certificate for certificate based verification. The default location to fetch the certificate file is /cisco-ar/pki. |
|
Specifies whether TLS session caching (fast reconnect) is enabled or not. Set to True to enable session caching; otherwise set to False. |
|
If TLS session caching (fast reconnect) is enabled, SessionTimeout specifies the maximum lifetime of a TLS session. Expired sessions are removed from the cache and require a subsequent full authentication. SessionTimeout is specified as a string consisting of pairs of numbers and units, where units might be one of the following: M, Minute, Minutes, H, Hour, Hours, D, Day, Days, W, Week, Weeks, as in the following: |
|
Mandatory; specifies time (in seconds) to wait before an authentication request times out. The default is 120. |
|
Mandatory; specifies the authentication service to use to authenticate users. See Configuring an EAP-TTLS Authentication Service for more information. Note The authentication service must exist before you can save the EAP-TTLS service configuration. |
To enable EAP-TTLS authentication:
Step 1 Launch aregcmd and create an EAP-TTLS service.
Step 2 Change directory to the service and set its type to eap-ttls.
Note Prime Access Registrar verifies the certificate during the TLS-based authentication. CRL validation is done before accepting a client certificate during the TLS authentication.
Configuring an EAP-TTLS Authentication Service
The EAP-TTLS service can authenticate users with either a legacy method such as PAP, CHAP, MSCHAP, or MSCHAPv2 or with an EAP method such as EAP-MSCHAPv2 or EAP-GTC. The authentication can be performed by the local server (the same server running EAP-TTLS) or it can be forwarded to a remote AAA Server (the home server for the user’s domain).
This section provides examples of several different ways to configure an EAP-TTLS authentication service. The following examples assume that you are using aregcmd and have already created the EAP-TTLS service.
Note After you make a configuration change, you must save the configuration before it can be used.
Authenticating Local Users with a Legacy Method
You can use a service like the local-users service (created as part of the example configuration) to authenticate users in the local UserList.
set AuthenticationService local-users
This service can be used to authenticate using PAP, CHAP, MSCHAP, and MSCHAPv2.
Authenticating Users with EAP-MSChapV2
This example uses a service named eap-mschapv2 for authentication. Attempts to authenticate using any other method than EAP-MSChapV2 (assuming the service type is also eap-mschapv2) will fail.
Authenticating Users with EAP Negotiate
You can use the EAP-negotiate method to authenticate using more than one EAP type. The following example defines an EAP service named eap-negotiate that can negotiate EAP-MSChapV2 or EAP-GTC then configures an EAP-TTLS service to authenticate using that service.
To configures an EAP-TTLS service to authenticate using eap-negotiate:
Step 1 Create a service of type eap-negotiate .
set ServiceList “eap-mschapv2 eap-gtc”
Step 2 Configure the EAP-TTLS AuthenticationService.
set AuthenticationService eap-nego
Authenticating Users with Legacy and EAP Methods
You can configure EAP-TTLS to authenticate using both legacy and EAP methods with a Group service using an OR result rule. A configuration like that shown in the following example first attempts to authenticate with the eap-negotiate service. If that fails, the server attempts to authenticate with the local-users service.
To authenticate with the eap-negotiate service;
Step 1 Create the Group service
Step 2 Configure the EAP-TTLS AuthenticationService.
set AuthenticationService local-or-eap
Authenticating Using a Remote AAA Server
You can configure an EAP-TTLS service to forward authentication to a remote AAA Server known (or the home server). The following configures a RADIUS service to use a remote server, then configures EAP-TTLS to use that service for authentication.
The first step in the following example configures a remote RADIUS server (aaa-remote) with its IP address and the shared secret that it shares with the local server. You might also specify other important parameters such as ports, timeouts, and maximum number of retries. See Services, for information about configuring RADIUS services.
To configure a remote RADIUS server (aaa-remote) with its IP address and a shared secret:
Step 1 Configure a remote AAA Server.
The following step configures a RADIUS service to use the remote server created in the previous step. You might also configure other important parameters such as the failover strategy. See Services, for information about configuring RADIUS services.
Step 2 Configure an AAA service.
Step 3 Configure the EAP-TTLS AuthenticationService:
set AuthenticationService home
Other configurations are also possible. For example, a group service can be used to perform some authentications locally and forward others to a remote server.
Testing EAP-TTLS with radclient
To test the EAP-TLS service, launch radclient and use the simple_eap_ttls_test command. The simple_eap_ttls_test command has the following syntax:
simple_eap_ttls_test identity password { method }
password is the user’s password
method is one of: PAP, CHAP, MSChap, MSChapV2, or PEAP.
Note If the method parameter is EAP, the tunnel command must be used to specify the EAP method type.
Testing EAP-TTLS Using Legacy Methods
To authenticate a user using EAP-TTLS with PAP:
Step 2 Authenticate using EAP-TTLS PAP.
simple_eap_ttls_test bob bob pap
The following commands show how to test the other valid legacy methods.
simple_eap_ttls_test bob bob chap
simple_eap_ttls_test bob bob mschap
simple_eap_ttls_test bob bob mschapv2
Testing EAP-TTLS Using EAP Methods
The following example uses EAP-TTLS with EAP-MSChapV2 as the Phase 2 method to authenticate a user named bob whose password is bob (from the example configuration). Issue the tunnel command to specify the Phase 2 EAP method, then issue the simple_eap_ttls_test command with eap as a method type.
To authenticate a user using EAP-TTLS with EAP-MSChapV2 as the Phase 2 method:
Step 2 Authenticate using EAP-TTLS and EAP-MSChapV2.
simple_eap_ttls_test bob bob eap
To test with a different EAP method, use the tunnel command to specify the method as shown in the following command to specify EAP-TLS.
simple_eap_ttls_test bob bob eap
rehash-ca-certs Utility
The rehash-ca-certs utility works with the CACertificatePath property and enables you to create the required certificate file hash links (similar to those used with PEAP and EAP-TLS). The rehash-ca-certs utility is only used when the server is validating certificates from the client (which is optional and not a common case for EAP-TTLS).
The syntax for the rehash-ca-certs utility is:
rehash-ca-certs { -v } path1 { path2 … pathn }
Each directory path specified on the command line is scanned by the rehash-ca-certs utility for filenames with the pem extension (such as ca-cert.pem ) and the appropriate hash link is created as described above. Before creating links, rehash-ca-certs first removes all existing links in the directory, so each invocation creates fresh links. The –v option enables verbose output.
The following is an example of the rehash-ca-certs utility:
The rehash-ca-certs utility warns about PEM files that do not contain certificates. On Cisco Prime Access Registrar, intermediate/chained certificates cannot be imported.
To run Prime Access Registrar with PEAP authentication:
Step 1 Add both root and intermediate CA in the directory /opt/CSCOar/pki (as configured for CACertificatePath in the service NYU-NetIDs-PEAPService).
Step 2 Change the directory to pki:
Step 3 run /opt/CSCOar/bin/rehash-ca-certs
Step 4 Stop ARserver and restart.
radclient Command Reference
This section provides a summary of the radclient commands you can use to test PEAP and EAP-TLS. It contains the following topics:
eap-trace
Use the eap-trace command to display additional client protocol trace information for EAP methods. Set the level to a number from 1 to 5 inclusively. Level 5 shows detailed hexadecimal dumps of all messages. Level 4 shows a message trace without hexadecimal dumps. Levels 3 and below show status and error information. To turn off trace displays, set the level to 0.
Use eap-trace level to set the trace level for all EAP methods. The following example command sets the trace level to 4 for all EAP methods:
Use eap-trace method level to set the trace level for the specified EAP method. The following example command sets the trace level to 5 for PEAP Version0 only. The trace level for other EAP methods is not affected.
Note The eap-trace command is for client-side trace information only and is independent of the server trace level you set using aregcmd.
tunnel
Use the tunnel command to specify the inner authentication method for PEAP. The specified EAP method type must agree with the server’s configured authentication method or authentication will fail.
For PEAP Version 0, the allowable tunnel methods are EAP-MSCHAPV2 and EAP-SIM. For PEAP Version 1, the allowable tunnel methods are EAP-GTC and EAP-SIM.
simple_eap_mschapv2_test username password
simple_eap_gtc_test username password
simple_eap_peapv0_test arg1 arg2
The arguments are passed to the inner authentication method as its authentication parameters. For EAP-MSChapv2 the arguments are username and password; for EAP-SIM they are IMSI and key.
simple_eap_peapv1_test arg1 arg2
The arguments are passed to the inner authentication method as its authentication parameters. For EAP-GTC the arguments are username and password; for EAP-SIM they are IMSI and key.
Protected EAP
Protected EAP (PEAP) is an authentication method designed to mitigate several weaknesses of EAP. PEAP leverages TLS (RFC 2246) to achieve certificate-based authentication of the server (and optionally the client) and creation of a secure session that can then be used to authenticate the client. PEAP provides several benefits:
- Industry standard authentication of the server using certificates (TLS)
- Standardized method for session key generation using TLS PRF
- Strong mutual authentication
- Identity privacy
- Fast reconnect using TLS session caching
- EAP message fragmentation
- Secure support for legacy client authentication methods
Cisco Prime Access Registrar supports the two major existing variants of PEAP, PEAP Version 0 (Microsoft PEAP) and PEAP Version 1 (Cisco PEAP). PEAP Version 0 is described in IETF drafts, draft-kamath-pppext-peapv0-00.txt and draft-josefsson-pppext-eap-tls-eap-02.txt . This version of PEAP can use either EAP-MSChapV2 or EAP-SIM as an authentication method. PEAP Version 1 is described by IETF draft draft-zhou-pppext-peapv1-00.txt . PEAP Version 1 can use either EAP-GTC or EAP-SIM as an authentication method.
This section contains the following topics:
Configuring PEAP Version 0
You can use aregcmd to create and configure a service of type peap-v0 . Table 9-10 describes the PEAP service properties for PEAP Version 0.
Optional script Prime Access Registrar server runs when it receives a request from a client for PEAP-v0 service. |
|
Optional script Prime Access Registrar server runs before it sends a response to a client using PEAP-v0 |
|
Indicates the maximum length in bytes that a PEAP or EAP-TLS message can have before it is fragmented. |
|
The full pathname of the file containing the server’s certificate or certificate chain used during the TLS exchange. The pathname can be optionally prefixed with a special string that indicates the type of encoding used for the certificate. The two valid encoding prefixes are PEM and DER. If an encoding prefix is not present, the file is assumed to be in PEM format. The following example assumes that the subdirectory pki under /cisco-ar contains the server’s certificate file. The file server-cert.pem is assumed to be in PEM format; note that the file extension .pem is not significant. |
|
The full pathname of the file containing trusted CA certificates used for client verification. The file can contain more than one certificate, but all certificates must be in PEM format. DER encoding is not allowed. |
|
The name of a directory containing trusted CA certificates (in PEM format) used for client verification. This parameter is optional, and if it is used there are some special preparations required for the directory it references. Each certificate file in this directory must contain exactly one certificate in PEM format. The server looks up the certificate files using the MD5 hash value of the certificate’s subject name as a key. The directory must therefore also contain a set of symbolic links each of which points to an actual certificate file. The name of each symbolic link is the hash of the subject name of the certificate. For example, if a certificate file name ca-cert.pem is located in the CACertificatePath directory, and the MD5 hash of the subject name contained in ca-cert.path.pem is 1b96dd93, then a symbolic link named 1b96dd93 must point to the ca-cert.pem file. If there are subject name collisions such as multiple certificates with the same subject name, each link name must be indexed with a numeric extension as in 1b96dd93.0 and 1b96dd93.1. |
|
Optional. The URL that Prime Access Registrar should use to retrieve the CRL.You can specify a URL that uses HTTP or LDAP.
The following is an example for an HTTP URL:
The following is an example for an LDAP URL: |
|
Specifies the type of verification used for client certificates. Must be set to one of RequireCertificate, None, or Optional. |
|
Specifies the maximum length of the certificate chain used for client verification. |
|
Determines the applicability of the authentication mechanism in SmartGrid Solutions, see the Smart Grid Solution Management for more information. When UseECCCertificates is set to True, it can use the ECC, RSA, or combination of both certificate for certificate based verification. When UseECCCertificates is set to False, it can only use the RSA certificate for certificate based verification. The default location to fetch the certificate file is /cisco-ar/pki. |
|
Specifies whether TLS session caching (fast reconnect) is enabled or not. Set to True to enable session caching; otherwise set to False. |
|
If TLS session caching (fast reconnect) is enabled, SessionTimeout specifies the maximum lifetime of a TLS session. Expired sessions are removed from the cache and will require a subsequent full authentication. SessionTimeout is specified as a string consisting of pairs of numbers and units, where units might be one of the following: M, Minute, Minutes, H, Hour, Hours, D, Day, Days, W, Week, Weeks, as in the following: |
|
Mandatory; specifies time (in seconds) to wait before an authentication request times out; defaults to 120. |
|
Mandatory; must be the name of an existing EAP-MSCHAPv2 or EAP-SIM service for PEAP Version 0. |
|
When set to TRUE, enables Windows Provisioning Service (WPS) and provides two other properties, MasterURL and WPSGuestUserProfile. The default value is FALSE. |
|
When using WPS, specifies the URL of the provisioning server which is modified with the appropriate fragment and sent to the client. |
|
When using WPS, specifies a profile to be used as a guest user profile; must be a valid profile under /Radius/Profiles. This profile is used for guests and users whose account has expired. This profile normally contains attributes denoting the VLAN-id of the guest network (which has the provisioning server alone) and might contain IP-Filters that would restrict the access of the guest (to only the provisioning server). |
Step 1 Launch aregcmd and create a PEAP Version 0 service.
Step 2 Set the service’s type to peap-v0.
Step 3 Set the service’s TunnelService property to the name of an existing EAP-MSCHAPV2 or EAP-SIM service.
set TunnelService name_of_EAP-MSCHAPv2_service
set TunnelService name_of_EAP-SIM_service
Note Prime Access Registrar verifies the certificate during the TLS-based authentication. CRL validation is done before accepting a client certificate during the TLS authentication.
Testing PEAP Version 0 with radclient
Step 2 Specify the inner authentication method, eap-mschapv2 or eap-sim, as in the following.
Step 3 Use the simple_eap_peapv0_test command to authenticate using PEAP Version 0, as in the following:
simple_eap_peapv0_test arg1 arg2
The simple_eap_peapv0_test command passes its arguments to the inner authentication mechanism which treats the arguments as either a username and a password (for eap-mschapv2) or as an IMSI and a key (for eap-sim).
The following example tests PEAP Version 0 with EAP-MSCHAPV2 as the inner authentication mechanism using username bob and password bob:
simple_eap_peapv0_test bob bob
The following example tests PEAP Version 0 with EAP-SIM as the inner authentication mechanism using IMSI 1124567891 and key 0112456789ABCDEF:
simple_eap_peapv0_test 1124567891 0112456789ABCDEF
Testing PEAP Version 0 with Client Certificates
You can test PEAP Version 0 using client certificates verified by the server during the TLS exchange. The client certificate file and RSA or ECC key file must reside in /cisco-ar/pki and be named client-cert.pem and client-key.pem respectively. Both files must be in PEM format.
Configuring PEAP Version 1
You can use aregcmd to create and configure a service of type peap-v1 . Table 9-11 describes the PEAP service properties for both PEAP Version 1.
Optional script Prime Access Registrar server runs when it receives a request from a client for PEAP-v1 service. |
|
Optional script Prime Access Registrar server runs before it sends a response to a client using PEAP-v1. |
|
Indicates the maximum length in bytes that a PEAP or EAP-TLS message can have before it is fragmented. |
|
The full pathname of the file containing the server’s certificate or certificate chain used during the TLS exchange. The pathname can be optionally prefixed with a special string that indicates the type of encoding used for the certificate. The two valid encoding prefixes are PEM and DER. If an encoding prefix is not present, the file is assumed to be in PEM format. |
|
The full pathname of the file containing trusted CA certificates used for client verification. The file can contain more than one certificate but all certificates must be in PEM format. DER encoding is not allowed. |
|
The name of a directory containing trusted CA certificates (in PEM format) used for client verification. This parameter is optional, and if it is used there are some special preparations required for the directory it references. Each certificate file in this directory must contain exactly one certificate in PEM format. The server looks up the certificate files using the MD5 hash value of the certificate’s subject name as a key. The directory must therefore also contain a set of symbolic links each of which points to an actual certificate file. The name of each symbolic link is the hash of the subject name of the certificate. For example, if a certificate file named ca-cert.pem is located in the CACertificatePath directory, and the MD5 hash of the subject name contained in ca-cert.path.pem is 1b96dd93, then a symbolic link named 1b96dd93 must point to the ca-cert.pem file. If there are subject name collisions such as multiple certificates with the same subject name, each link name must be indexed with a numeric extension as in 1b96dd93.0 and 1b96dd93.1. |
|
Optional. The URL that Prime Access Registrar should use to retrieve the CRL.You can specify a URL that uses HTTP or LDAP.
The following is an example for an HTTP URL:
The following is an example for an LDAP URL: |
|
Specifies the type of verification used for client certificates. Must be set to one of RequireCertificate, None, or Optional. |
|
Specifies the maximum length of the certificate chain used for client verification. |
|
Determines the applicability of the authentication mechanism in SmartGrid Solutions, see the Smart Grid Solution Management for more information. When UseECCCertificates is set to True, it can use the ECC, RSA, or combination of both certificate for certificate based verification. When UseECCCertificates is set to False, it can only use the RSA certificate for certificate based verification. The default location to fetch the certificate file is /cisco-ar/pki. |
|
Specifies whether TLS session caching (fast reconnect) is enabled or not. Set to True to enable session caching; otherwise set to False. |
|
If TLS session caching (fast reconnect) is enabled, SessionTimeout specifies the maximum lifetime of a TLS session. Expired sessions are removed from the cache and will require a subsequent full authentication. SessionTimeout is specified as a string consisting of pairs of numbers and units, where units might be one of the following: M, Minute, Minutes, H, Hour, Hours, D, Day, Days, W, Week, Weeks, as in the following: |
|
Mandatory; specifies time (in seconds) to wait before an authentication request times out; defaults to 120. |
|
Mandatory; must be the name of an existing EAP-GTC or EAP-SIM service for PEAP Version 0. |
Step 1 Launch aregcmd and create a PEAP Version 1 service.
Step 2 Set the service’s type to peap-v1.
Step 3 Set the service’s TunnelService property to the name of an existing EAP-GTC or EAP-SIM service.
set TunnelService name_of_EAP-GTC_service
set TunnelService name_of_EAP-SIM_service
Testing PEAP Version 1 with radclient
Step 2 Specify the inner authentication method, EAP-GTC or EAP-SIM, as in the following.
Step 3 Use the simple_eap_peapv1_test command to authenticate using PEAP Version 1, as in the following:
simple_eap_peapv1_test arg1 arg2
The simple_eap_peapv1_test command passes its arguments to the inner authentication mechanism which treats the arguments as either a username and a password (for EAP-GTC) or as an IMSI and a key (for EAP-SIM).
Testing PEAP Version 1 with Client Certificates
You can test PEAP Version 1 using client certificates verified by the server during the TLS exchange. The client certificate file and RSA or ECC key file must reside in /cisco-ar/pki and be named client-cert.pem and client-key.pem respectively. Both files must be in PEM format.
How to Configure Oracle, Mysql Accounting with the Buffering Option Enabled
Prime Access Registrar provides support for MySQL to query user records from Oracle database using sql interface and enables you to write accounting records into Oracle database. You can use insert, update, and delete queries to
- add new details into database.
- modify the existing details in the database.
- remove the outdated details from the database.
To Select the SQL Statement in Run Time Accounting
Prime Access Registrar provides support to query user account details from SQL database and enables you to add, delete, and update accounting details into SQL when using Oracle accounting.
You can execute the following SQL statements to perform various actions:
Query
You can query the accounting details from Oracle by referring this service in /Radius/DefaultAuthenticationService and in /Radius/DefaultAuthorization.
The following example is an SQL statement used for Authentication and Authorization of the subscribed users. You can use the SQL and MarkerList properties statement to query the selected attributes from Oracle.
Insert
You can insert user details into SQL database by Oracle accounting. This service is used by referring the /Radius/DefaultAccountingService or Accounting-Service environment variable.
For instance, you can use the following SQL and MarkerList properties statement to insert the selected attributes:
Update
You can easily modify the details in an SQL table with the UPDATE statement.
For example, you can use the following SQL and MarkerList properties statement to update the selected attributes:
Delete
You can remove the unnecessary records from SQL database using DELETE statement.
For example, you can use the following SQL and MarkerList properties statement to delete the selected attributes:
Configuring Oracle, Mysql Accounting
The following script describes you how to configure Oracle, Mysql accounting with the buffering option enabled:
How Suffix and Prefix Rules Work with Prime Access Registrar
Prime Access Registrar includes several scripts that you can use with the rules. The following are the most commonly used rules:
- Prefix Rule, See ExecPrefixRule for more information
- Suffix Rule, See ExecSuffixRule for more information
Configuring Prefix and Suffix Policies
To configure prefix and suffix policies in Prime Access Registrar in order to provide authentication and authorization services for the subscribed users:
Step 1 Activate the Policy Engine by configuring SelectPolicy. This script explains you how to set a suffix and prefix policy in the grouping list.
Step 2 Run the configuration rules for Prefix and Suffix.
Step 3 Set Script = ExecSuffixRule in the prefix rule configuration.
Step 4 Specify Script = ExecRealmRule in the suffix configuration to scan.
CRL Support for Cisco Prime Access Registrar
Prime Access Registrar checks for various certificates for validation purposes in its authentication services. The client sends a certificate along with the access-challenge to Prime Access Registrar. Prime Access Registrar verifies the validity of the certificate and approves the request if the certificate is valid. For certificate validation, Prime Access Registrar uses an advanced verification mechanism, which uses Certificate Revocation Lists (CRLs).
A CRL, which uses the X.509 certification format, is the signed data structure that the certificate authority (CA) issues periodically. It contains a list of the serial numbers and the timestamp of the revoked certificates. These revoked certificates are not valid and Prime Access Registrar rejects any request that comes with these certificates. The CRLs are available in a public repository in Prime Access Registrar.
A certificate can be revoked because of the following reasons:
- Expiration of the validity period.
- Change in the name of the user to whom the certificate is issued.
- Change in the association between the CA and the user.
- Loss of the private they that is associated with the certificate.
Prime Access Registrar uses the Lightweight Dynamic Authentication Protocol (LDAP) and HTTP for validating the certificates using CRL. The CRLDistributionURL in the TLS based EAP authentication services, is used for the CRL support in Prime Access Registrar. When you configure this property, Prime Access Registrar fetches the CRL from the specified URL, at the startup. A background thread in Prime Access Registrar keeps track of these CRLs. When any of the CRLs expires, Prime Access Registrar fetches the latest version of CRL using the specified URL. Each CRL contain the information related to its expiry.
Prime Access Registrar places all the CRLs in a CRL store. It uses these CRLs while it does a TLS authentication for certificate validation. During an authentication service, the certificate verifier in Prime Access Registrar checks for the validity of the certificate against the CRL issued by the CA that signed the certificate. It looks for the serial number of the certificate in the list of revoked certificates in the appropriate CRL. If it finds a match in the CRL, it compares the revocation time that is encoded in the CRL against the current time. If the current time is later than the revocation time, Prime Access Registrar considers the certificate invalid.
This section contains the following topics:
- Configuring Certificate Validation Using CRL
- Using Intermediate Certificates in Prime Access Registrar
Note Prime Access Registrar uses the CRLDistributionURL property in the following services:
eap-tls
eap-ttls
eap-fast
peap-v0
peap-v1
Configuring Certificate Validation Using CRL
Prime Access Registrar uses the CRLDistributionURL property for the certificate validation using CRLs. The following shows a sample configuration for the certificate verification using CRLs in Prime Access Registrar:
Table 9-9 describes the properties in this sample configuration.
Using Intermediate Certificates in Prime Access Registrar
The rehash-ca-certs utility can be used to import intermediate certificates in Prime Access Registrar. See rehash-ca-certs Utility for information about how to create the required certificate file hash links.
To import intermediate certificates in Prime Access Registrar:
Step 1 Copy the Root CA, Intermediate CA of the client to a directory.
Step 2 Run /opt/CSCOar/bin/rehash-ca-certs –v <path of the client certificate store>
The utility creates the required hash links to maintain the chain between the Root CA certificate and Intermediate CA certificates.
Step 3 Set the CACertificateFile property in EAP service to the path where Root CA Certificate of the client is stored.
Step 4 Restart the Prime Access Registrar server.
The following shows an example to import intermediate certificates in Prime Access Registrar:
Step 1 Copy the Client Root CA and Intermediate CA Certificate in /cisco-ar/certs/wimax/ directory .
cp /tmp/wimax_device_root.pem /cisco-ar/certs/wimax/
cp /tmp/wimax_device_root_ca1.pem /cisco-ar/certs/wimax/
/opt/CSCOar/bin/rehash-ca-certs –v /cisco-ar/certs/wimax/
a. Configure the eap service which uses these client certificates.
set CACertificateFile PEM:/opt/CSCOar/pki/wimax_device_root.pem
Step 3 Save the configuration.
/opt/CSCOar/bin/arserver restart
Rolling Encryption Support for Pseudonym Generation in EAP-SIM, EAP-AKA, and EAP-AKA’ Services
Prime Access Registrar supports rolling encryption which involves generating rolling pseudonym secrets instead of a single pseudonym secret for EAP-SIM, EAP-AKA, and EAP-AKA’ services. This feature involves the following objects:
Prime Access Registrar allows you to configure pseudonym encryption keys. There are two types of user-defined keys:
- For generating 3GPP compliant pseudonym secrets—configured at /Radius/Advanced/KeyStores/3GPPKeys
- For generating non-3GPP compliant pseudonym secrets—configured at /Radius/Advanced/KeyStores/non3GPPKeys
If user-defined keys are not configured, Prime Access Registrar uses system generated keys for rolling encryption.
The Key Store maintains a set of keys that are user-defined and/or auto-generated. There is only one active key at any point in time that is used for generating the pseudonym secret. Rest of the keys are inactive and are used for decryption of old pseudonyms if used for pseudonym generation. After a rollover period, another key is selected as an active key and is used for pseudonym generation. Expiry of a pseudonym depends on the PseudonymLifetime property set for the corresponding EAP service.
For each key store, you can configure the number of keys and rollover period for each key, which is one week by default. For each type of EAP service, there are two key stores:
- For generating 3GPP compliant pseudonym secrets—number of keys is limited to 15. Default is 15. Each key should be 16 digits in length.
- For generating non-3GPP compliant pseudonym secrets—number of keys is limited to 1024. Default is 256. Each key should be 29 digits in length.
Following is the CLI configuration of a key store:
CLI to configure non-3GPP keys:
Rolling encryption is applicable for an EAP-SIM, EAP-AKA, or EAP-AKA’ service. This feature is applicable only when the parameter EnableRollingPseudonymSecret is set to TRUE for the service. Each service uses either the 3GPP compliant or non-3GPP compliant pseudonym key store based on the Generate3GPPCompliantPseudonym option set for the service.
The EAP service uses the current active key of the corresponding key store to generate the pseudonym. After rollover period of the current active key, the next active key is selected and used for pseudonym generation. When Prime Access Registrar receives the pseudonym back from the client, it decrypts the pseudonym using the same key that was used to generate the pseudonym. Based on the PseudonymRenewtime and PseudonymLifetime set for the service, if the pseudonym received from the client is expired or to be renewed, Prime Access Registrar uses the current active key from the corresponding key store,
Table 9-12 provides an example of the rolling encryption process.
Feedback