Network Working Group Alan DeKok INTERNET-DRAFT FreeRADIUS Category: Informational 23 February 2007 Use of Status-Server Packets in RADIUS By submitting this Internet-Draft, each author represents that any applicable patent or other IPR claims of which he or she is aware have been or will be disclosed, and any of which he or she becomes aware will be disclosed, in accordance with Section 6 of BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet- Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at http://www.ietf.org/ietf/1id-abstracts.txt. The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. This Internet-Draft will expire on July 5, 2007. Copyright Notice Copyright (C) The IETF Trust (2007). Abstract [RFC2865] defines a Status-Server code for use in RADIUS, but labels it as "Experimental" without further discussion. This document describes practical uses for Status-Server that have been implemented as a method of querying the status of a RADIUS server. DeKok, Alan Informational [Page 1] INTERNET-DRAFT Status-Server Practices 23 February 2007 Table of Contents 1. Introduction ............................................. 3 1.1. Terminology ......................................... 3 1.2. Requirements Language ............................... 4 2. Problem Statement ........................................ 5 2.1. Sending Access-Request "pings" ...................... 5 2.1.1. Recommendation against Access-Request .......... 6 2.2. Sending Accounting-Request "pings" .................. 6 2.2.1. Recommendation against Accounting-Request ...... 7 2.3. Status-Server as a Solution ......................... 7 2.3.1. Status-Server instead of Access-Request ........ 7 2.3.2. Status-Server instead of Accounting-Request .... 7 3. Packet Format ............................................ 8 3.1. Consistent definition for Status-Server ............. 10 4. Implementation notes ..................................... 10 4.1. Client Requirements ................................. 11 4.2. Server Requirements ................................. 12 4.3. More Robust Fail-over with Status-Server ............ 13 4.4. Proxy Server handling of Status-Server .............. 14 4.5. MIB Considerations .................................. 14 4.5.1. Interaction with RADIUS Server MIBs ............ 14 4.5.2. Interaction with RADIUS Client MIBs ............ 15 5. Additional considerations ................................ 15 5.1. Local site testing .................................. 15 5.2. RADIUS over reliable transports ..................... 17 5.3. Other uses for Status-Server ........................ 17 5.4. Potential Uses for Status-Client .................... 17 6. Table of Attributes ...................................... 18 7. Examples ................................................. 18 7.1. Minimal Query to Authentication Port ................ 18 7.2. Minimal Query to Accounting Port .................... 19 7.3. Verbose Query and Response .......................... 20 8. IANA Considerations ...................................... 21 9. Security Considerations .................................. 21 10. References .............................................. 21 10.1. Normative references ............................... 21 10.2. Informative references ............................. 21 Intellectual Property Statement .............................. 22 Disclaimer of Validity ....................................... 24 Full Copyright Statement ..................................... 24 DeKok, Alan Informational [Page 2] INTERNET-DRAFT Status-Server Practices 23 February 2007 1. Introduction The RADIUS Working Group was formed in 1995 to document the protocol of the same name, and created a number of standards surrounding the protocol. It also defined experimental commands within the protocol, without elaborating further on the potential uses of those commands. This document describes how some of the current implementations on the market are using Status-Server packets as an application-layer method of "pinging" a RADIUS server to see if it is responding to requests. These queries do not affect the normal operation of the server, and do not result in any side effects other than incrementing internal packet counters. These "pings" are not intended to be the application-layer watchdog messages described in [RFC3539] Section 3.4. That document describes AAA protocols that run over reliable transports, which handle retransmissions internally. Since RADIUS runs over UDP rather than TCP, the full watchdog mechanism is not applicable here. The rest of this document is laid out as follows. Section 2 contains the problem statement, and explanations as to why some possible solutions have unwanted side effects. Section 3 defines the Status- Server packet format. Section 4 contains implementation notes with detailed client and server requirements. Section 5 lists additional considerations not covered in the other sections. The remaining text lists the typical RADIUS table of attributes, and covers some additional security considerations not covered in the rest of the document. 1.1. Terminology This document uses the following terms: Network Access Server (NAS) The device providing access to the network. Also known as the Authenticator (IEEE 802.1X or EAP terminology) or RADIUS client. Home Server A RADIUS server that is authoritative for user authorization and authentication. Proxy Server A RADIUS server that acts as a Home Server to the NAS, but in turn proxies the request to another Proxy Server, or to a Home Server. silently discard This means the implementation discards the packet without further DeKok, Alan Informational [Page 3] INTERNET-DRAFT Status-Server Practices 23 February 2007 processing. The implementation SHOULD provide the capability of logging the error, including the contents of the silently discarded packet, and SHOULD record the event in a statistics counter. 1.2. Requirements Language In this document, several words are used to signify the requirements of the specification. The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC2119]. DeKok, Alan Informational [Page 4] INTERNET-DRAFT Status-Server Practices 23 February 2007 2. Problem Statement It is often useful to know if a RADIUS server is alive and responding to requests. The most accurate way to obtain this information is to query the server via normal protocol traffic, as other methods are either less accurate, or cannot be performed remotely. The reasons for wanting to know the status of a server are many. The administrator may simply be curious if the server is responding, and may not have access to NAS or traffic data that would give him that information. The queries may also be performed automatically by a NAS or proxy server, which is configured to send packets to a RADIUS server, and where that server may not be responding. That is, while [RFC2865] Section 2.6 indicates that sending Keep-Alives is harmful, it is useful to send "Are you Alive" queries to a server once it has been marked "dead" due to unresponsiveness. The occasional query to a "dead" server has little impact on the network or server load, and permits clients to more quickly discover when the server returns to a responsive state. In all, status queries are a useful part of a network server deployment. 2.1. Sending Access-Request "pings" One possible solution to the problem of querying server status is to send Access-Request packets as a kind of application-level "ping" to the RADIUS port (1812), and then to look for an Access-Accept response. However, the server may then conclude that a real user has logged onto a NAS, and perform local site actions that are undesirable for a simple status query. The server may otherwise respond with an Access-Challenge, indicating that it believes that it is engaging in an extended RADIUS authentication conversation with a user via the NAS. This side effect is also unwanted, as the request was intended as a simple status query, rather than a desire to start a longer protocol conversation. Or, the server may respond with an Access-Reject, indicating that the user is not authorized to gain access to network. As above, the server may also perform local site actions, such as warning an administrator of failed login attempts. The server may also delay the Access-Reject response, in the traditional manner of rate- limiting failed authentication attempts. This delay in response to the "ping" means that for a period of time, the querying administrator is unsure as to whether or not the server is down, is slow to respond, or is intentionally delaying its response to the query. DeKok, Alan Informational [Page 5] INTERNET-DRAFT Status-Server Practices 23 February 2007 In addition, using Access-Request packets as "pings" may mean that the server has to have local users configured whose sole reason for existence is to enable the "ping" requests. Unless the server policy is carefully designed for those users, it may be possible for an attacker to use that users credentials to gain unauthorized network access. We note that some NAS implementations currently use Access-Request packets as described above, with fixed (and non configurable) user name and password. Issues with those implementations sometimes mean that if a RADIUS server does not respond to those "ping" request, the server may be marked as unresponsive, even though it is actively responding to other Access-Requests. This behavior is confusing to administrators who have to understand why a live server is marked "unresponsive". 2.1.1. Recommendation against Access-Request For the reasons outlined above, NAS implementors SHOULD NOT use Access-Request packets as "pings" to see if a server is alive. Similarly, site administrators SHOULD NOT configure test users whose sole reason for existence is to enable "pings" via Access-Request packets. Note that it still may be useful to configure test users for the purpose of performing end-to-end or in-depth testing of a servers policy. We do not recommend against this practice, though we do warn administrators to use it with caution. 2.2. Sending Accounting-Request "pings" A similar solution for the problem of querying server status may be to send Accounting-Request packets as a kind of application-level "ping" to the RADIUS accounting port (1813), and then to look for an Accounting-Response packet. As above, the server may then conclude that a real user has logged onto a NAS, and perform local site actions that are an undesirable for a simple status query. In addition, some attributes are mandatory to include in an Accounting-Request. This requirement forces the administrator who desires to query the server to create "fake" values for those attributes in a test packet. These "fake" values increase the work required to perform a simple query, and may pollute the servers accounting database with invalid data. DeKok, Alan Informational [Page 6] INTERNET-DRAFT Status-Server Practices 23 February 2007 2.2.1. Recommendation against Accounting-Request For the reasons outlined above, NAS implementors SHOULD NOT use Accounting-Request packets as "pings" to see if a server is alive. Similarly, site administrators SHOULD NOT configure accounting policies whose sole reason for existence is to enable "pings" via Accounting-Request packets. Note that it still may be useful to configure test users for the purpose of performing end-to-end or in-depth testing of a servers policy. We do not recommend against this practice, though we do warn administrators to use it with caution. 2.3. Status-Server as a Solution A better solution to the problem is to use Status-Server. The name appears to be intended for packets that query the status of a server, the packets are previously undefined, and the problem statement outlined in Section 3 indicates that administrators need the ability to query the a servers status. The difficulty before now has been an inter-operable method of performing these queries. 2.3.1. Status-Server instead of Access-Request Status-Server SHOULD be used instead of Access-Request to query the responsiveness of a server. In this use-case, the protocol exchange between client and server is similar to the normal exchange of Access-Request and Access-Accept, as diagrammed below. NAS RADIUS server --- ------------- Status-Server/ Message-Authenticator -> <- Access-Accept/ Reply-Message With this design, the Status-Server packet replaces an Access-Request packet, without the previously mentioned side effects. 2.3.2. Status-Server instead of Accounting-Request Status-Server may be used instead of Accounting-Request to query the responsiveness of a server. In this use-case, the protocol exchange between client and server is similar to the normal exchange of Accounting-Request and Accounting-Response, as diagrammed below. NAS RADIUS server --- ------------- Status-Server/ DeKok, Alan Informational [Page 7] INTERNET-DRAFT Status-Server Practices 23 February 2007 Message-Authenticator -> <- Accounting-Response With this design, the Status-Server packet replaces an Accounting-Request packet, without the previously mentioned side effects. 3. Packet Format Status-Server packets have the normal RADIUS packet form, with the fields and values for those fields as defined [RFC2865] Section 3. We do not include all of the text or diagrams of that section here, but instead document the details specific to implementing Status- Server. The Authenticator field of Status-Server packets MUST be generated using the same method as that used for the Request Authenticator field of Access-Request packets. In more detail, the definition of Request Authenticator for Status-Server packets is given below. The role of the Identifier field is the same for Status-Server as for other packets. However, as Status-Server is taking the role of Access-Request or Accounting-Request packets, there is the potential for Status-Server requests to be in conflict with Access-Request or Accounting-Request packets with the same Identifier. In Section 4.2, below, we describe how these problems can be avoided. Request Authenticator In Status-Server Packets, the Authenticator value is a 16 octet random number, called the Request Authenticator. The value SHOULD be unpredictable and unique over the lifetime of a secret (the password shared between the client and the RADIUS server), since repetition of a request value in conjunction with the same secret would permit an attacker to reply with a previously intercepted response. Since it is expected that the same secret MAY be used to authenticate with servers in disparate geographic regions, the Request Authenticator field SHOULD exhibit global and temporal uniqueness. The Request Authenticator value in a Status-Server packet SHOULD also be unpredictable, lest an attacker trick a server into responding to a predicted future request, and then use the response to masquerade as that server to a future Status- Server. Although protocols such as RADIUS are incapable of protecting against theft of an authenticated session via real-time active DeKok, Alan Informational [Page 8] INTERNET-DRAFT Status-Server Practices 23 February 2007 wiretapping attacks, generation of unique unpredictable requests can protect against a wide range of active attacks against authentication. Similarly, the Response Authenticator field of Access-Accept packets sent in response to Status-Server queries MUST be generated using the normal method for calculating the Response Authenticator of the Access-Accept, with the Status-Server Request Authenticator taking the place of the Access-Request Request Authenticator. The Response Authenticator field of Accounting-Response packets sent in response to Status-Server queries MUST be generated using the normal method for calculating the Response Authenticator of the Accounting-Response, with the Status-Server Request Authenticator taking the place of the Accounting-Request Request Authenticator. In more detail, the definition of Response Authenticator is given below. Response Authenticator The value of the Authenticator field in Access-Accept, or Accounting-Response packets is called the Response Authenticator, and contains a one-way MD5 hash calculated over a stream of octets consisting of: the RADIUS packet, beginning with the Code field, including the Identifier, the Length, the Request Authenticator field from the Status-Server packet, and the response Attributes (if any), followed by the shared secret. That is, ResponseAuth = MD5(Code+ID+Length+RequestAuth+Attributes+Secret) where + denotes concatenation. In addition to the above requirements, all Status-Server packets MUST include a Message-Authenticator attribute. Failure to do so would mean that the packets could be trivially spoofed. Status-Server packets MAY include NAS-Identifier, and/or NAS-IP- Address or NAS-IPv6-Address. These attributes are not necessary for the operation of Status-Server, but may be useful information to a server that receives those packets. Other attributes SHOULD NOT be included in a Status-Server packet. User authentication or authorization credentials such as User-Name, User-Password, CHAP-Password, etc. MUST NOT appear in a Status-Server packet sent to a server authentication port. User or NAS accounting attributes such as User-Name, Acct-Session-Id, or Acct-Status-Type MUST NOT appear in a Status-Server packet sent to a server accounting port. DeKok, Alan Informational [Page 9] INTERNET-DRAFT Status-Server Practices 23 February 2007 Similarly, the Access-Accept MAY include a Reply-Message attribute. The Access-Accept or Accounting-Response packets sent in response to a Status-Server query SHOULD NOT contain any attributes. As the intent is to implement a simple "ping" instead of user authentication or accounting, there is no reason to include other attributes in either the query or the corresponding response. Detailed examples are given in Section 7, below. 3.1. Consistent definition for Status-Server When sent to a RADIUS accounting port, Status-Server packets are calculated as described above. That is, even though the packets are being sent to an accounting port, they are not created via the same method as Accounting-Request packets. This difference from the normal Accounting-Request handling has a number of benefits. Having one definition for Status-Server is simpler than defining the packet differently when it is sent to different ports. In addition, if we were to define Status-Server as a similar to Accounting- Request, but containing no attributes, then the packets could be trivially spoofed. We therefore define Status-Server consistently, and vary the response packets depending on the port to which the request is sent. When sent to an authentication port, the response to a Status-Server query is an Access-Accept packet. When sent to an accounting port, the response to a Status-Server query is an Accounting-Response packet. 4. Implementation notes There are a number of considerations to take into account when implementing support for Status-Server. This section describes implementation details and requirements for RADIUS clients and servers that support Status-Server. The following text applies to both authentication and accounting ports. We use the generic term "request packets" to mean "Access- Request packets sent to an authentication port, or Accounting-Request packets sent to an accounting port." We also use the generic term "response packets" to mean "Access-Accept, Access-Challenge, or Access-Reject packets sent from an authentication port, or Accounting-Response packets sent from an accounting port." Using generic terms to describe the Status-Server conversations is simpler than duplicating the text for both authentication and accounting ports. DeKok, Alan Informational [Page 10] INTERNET-DRAFT Status-Server Practices 23 February 2007 4.1. Client Requirements Clients SHOULD permit administrators to globally enable or disable the generation of Status-Server packets. The default SHOULD be to globally disable it. As it is undesirable to send queries to servers that do not support Status-Server, clients SHOULD also have a per- server configuration indicating whether or not the server support Status-Server. The default SHOULD be that the server does not support Status-Server. The client SHOULD also have a configurable global timer (Tw) that is used when sending periodic Status-Server queries during server fail- over. The default value SHOULD be 30 seconds, and the value MUST NOT be permitted to be set below 6 seconds. If a response has not been received within the timeout period, the request is deemed to have no response, and MUST be discarded. When Status-Server packets are sent from a client, they MUST NOT be retransmitted. Instead, the Identity field MUST be changed for every Status-Server packet. The old request should be discarded, and a new Status-Server packet should be generated and sent, with new Identity and Authenticator fields. Clients MUST include the Message-Authenticator attribute in all Status-Server packets. Failure to do so would mean that the packets could be trivially spoofed, leading to potential denial of service (DoS) attacks. Other attributes SHOULD NOT appear in a Status-Server packet. As the intent of the packet is a simple status query, there is no reason for any additional attributes to appear in Status-Server packets. The client MAY increment packet counters as a result of sending a Status-Server, or receiving a response packet. The client MUST NOT perform any other action that is normally performed when it receives a response packet, such as permitting a user login access to a port. When a client sends Status-Server packets, those requests MUST NOT be sent from a source port that is used to send Access-Request or Accounting-Request packets. Clients SHOULD dedicate a source port solely for sending Status-Server requests. Clients MAY send Status- Server requests to both authentication and accounting destination ports from the same source port. The above requirement for a unique source port aids in matching responses to requests. Since the response to a Status-Server packet is an Access-Accept or Accounting-Response packet, those responses are indistinguishable from normal packets sent in response to an Access-Request or Accounting-Request. Therefore, the best way to DeKok, Alan Informational [Page 11] INTERNET-DRAFT Status-Server Practices 23 February 2007 distinguish them from normal traffic is to have a unique port. When the client receives a response to a Status-Server query, the response may be either an Access-Accept packet or an Accounting- Response packet, depending on the behavior of the server, and the port to which the query was sent. It may be difficult for the client to know which response packet to expect, so the client SHOULD accept either packet code as an acceptable response to a Status-Server query, subject to the validation requirements for the Response Authenticator. That is, prior to accepting the response as valid, the client should check that the response code is Access-Accept (2) or Accounting- Response (5). If the code does not match one of those two values, the packet MUST be silently discarded. The client MUST then validate the Response Authenticator via the algorithm given above in Section 3. If the Response Authenticator is not valid, the packet MUST be silently discarded. If, however, the Response Authenticator is valid, then the packet MUST be deemed to be a valid response from the server. If the client instead discarded the response because the packet code did not match what it expected, then it could erroneously discard valid responses from a server, and mark that server as unresponsive. This behavior would affect the stability of a RADIUS network, as "live" servers would not be used by clients. We therefore recommend that clients should be liberal in what they expect as responses to Status-Server queries. 4.2. Server Requirements Servers SHOULD permit administrators to globally enable or disable the acceptance of Status-Server packets. The default SHOULD be to globally enable it. Status-Server packets originating from clients that are not permitted to send the server request packets MUST be silently discarded. If a server does not support Status-Server packets, or is configured to not respond to them, then it MUST silently discard the packet. Servers SHOULD silently discard Status-Server packets if they determine that a client is sending too many Status-Server requests in a particular time period. Note that the server MAY discard the packet without first validating the Message-Authenticator attribute. The method used by a server to make this determination is implementation-specific, and out of scope of this document. If a server supports Status-Server packets, and is configured to DeKok, Alan Informational [Page 12] INTERNET-DRAFT Status-Server Practices 23 February 2007 respond to them, and receives a packet from a known client, it MUST validate the Message-Authenticator attribute as defined in [RFC3579] Section 3.2. Packets failing validation MUST be silently discarded. Servers SHOULD NOT otherwise discard Status-Server packets if they have recently sent the client a response packet. The query may have originated from an administrator who does not have access to the response packet stream, or who is interested in obtaining additional information about the server. The server SHOULD prioritize the handling Status-Server queries over normal request handling, subject to the rate limiting described above. As the intent of Status-Server is to query the responsiveness of the server, it is unhelpful for the server to queue Status-Server packets for later handling. Such queuing may lead a client to believe that a server is unresponsive, when it is merely slow. Some server implementations require that Access-Request packets are accepted only on "authentication" ports, (e.g. 1812/udp), and that Accounting-Request packets are accepted only on "accounting" ports (e.g. 1813/udp). Those implementations SHOULD reply to Status-Server packets sent to an "authentication" port with an Access-Accept packet. Those implementations SHOULD reply to Status-Server packets sent to an "accounting" port with an Accounting-Response packet. Some server implementations accept both Access-Request and Accounting-Request packets on one port, and do not distinguish between "authentication" only ports, and "accounting" only ports. Those implements SHOULD reply to Status-Server packets with an Access-Accept packet. The server MAY increment packet counters as a result of receiving a Status-Server, or sending a response packet. The server MUST NOT perform any other action that is normally performed when it receives a request, other than sending a response packet. 4.3. More Robust Fail-over with Status-Server A common problem in RADIUS client implementations is the implementation of a robust fail-over mechanism. A client may have multiple servers configured for example, with one server marked as primary and another marked as secondary. If the client determines that the primary is unresponsive, it "fails over" to the secondary, and begins to send requests to the secondary instead of the primary. However, it is difficult to know when the client should start sending requests to the primary again. Sending test Access-Requests or Accounting-Requests to see if the server is alive has all of the DeKok, Alan Informational [Page 13] INTERNET-DRAFT Status-Server Practices 23 February 2007 issues outlined above. Clients could alternately send real traffic to the primary, on the hope that it is responsive. If the server is still unresponsive, however, the result may be lost accounting data or user login failures. This problem is one that the Status-Server solution can address. When a client fails over from one server to another because of a lack of responsiveness, it SHOULD send periodic Status-Server packets to the unresponsive server, using the timer (Tw) defined above. Once three time periods have passed where Status-Server messages have been sent and responded to, the server should be deemed responsive and RADIUS requests may sent to it again. This determination should be made separately for each server that the client has a relationship with, though the same algorithm applies to both authentication and accounting servers. The above behavior is modelled after [RFC3539] Section 3.4.1. We note that if a reliable transport is used for RADIUS, then the algorithms specified in [RFC3539] MUST be used in preference to the ones given here. 4.4. Proxy Server handling of Status-Server Many RADIUS servers can act as proxy servers, and forward requests to home servers. Such servers MUST NOT proxy Status-Server packets. The purpose of Status-Server as specified here is to permit the client to query the responsiveness of a server. Proxying Status- Server queries negates any usefulness that may be gained by implementing support for them. Proxy servers MAY be configured to respond to Status-Server queries from clients, and MAY act as clients sending Status-Server queries to other servers. However, those operations MUST be independent of one another. 4.5. MIB Considerations 4.5.1. Interaction with RADIUS Server MIBs Since Status-Server packets are sent to the normal RADIUS ports, the question arises of how these packets affect the [RFC4669] and [RFC4671] RADIUS server MIBs. [RFC4669] defines a counter named radiusAuthServTotalUnknownTypes, that counts "The number of RADIUS packets of unknown type that were received". [RFC4671] defines a similar counter named radiusAcctServTotalUnknownTypes. Implementations not supporting Status-Server, or implementations that DeKok, Alan Informational [Page 14] INTERNET-DRAFT Status-Server Practices 23 February 2007 are configured to not respond to Status-Server packets MUST use these counters to track received Status-Server packets. If, however, Status-Server is supported and the server is configured to respond as described above, then the counters defined in [RFC4669] and [RFC4671] MUST NOT be used to track Status-Server requests or responses to those requests. That is, the [RFC4669] and [RFC4671] counters MUST be unaffected by the transmission or reception of packets relating to Status-Server. If an implementation supports Status-Server and the [RFC4669] or [RFC4671] MIBs, then it SHOULD also support vendor-specific MIBs containing similar information as those MIBs, but which are instead dedicated solely to tracking Status-Server requests and responses. The exact definition of the server MIBs for Status-Server is outside of the scope of this document. 4.5.2. Interaction with RADIUS Client MIBs Clients implementing Status-Server MUST NOT increment [RFC4668] or [RFC4670] counters upon reception of response packets to Status- Server queries. That is, the [RFC4668] and [RFC4670] counters MUST be unaffected by the transmission or reception of packets relating to Status-Server. If an implementation supports Status-Server and the [RFC4668] or [RFC4670] MIBs, then it SHOULD also support vendor-specific MIBs containing similar information as those MIBs, but which are instead dedicated solely to tracking Status-Server requests and responses. The exact definition of the clients MIBs for Status-Server is outside of the scope of this document. 5. Additional considerations There are additional topics related to the use of Status-Server that may be covered. As those topics do not fit well into the preceding sections, they are covered in this section, below. 5.1. Local site testing There is at least one situation where using Access-Request or Accounting-Request packets may be useful, despite the recommendations above in Section 2.1.1 and Section 2.2.1. That situation is local site testing, where the RADIUS client, server, and user store are under the control of a single administrator or administrative entity. In that situation, administrators MAY configure a well-known "test" user to enable local site testing. DeKok, Alan Informational [Page 15] INTERNET-DRAFT Status-Server Practices 23 February 2007 The advantage to creating such a local user is that it is now possible for the administrator to send a RADIUS request that performs end-to-end testing of the RADIUS server. As above with Status- Server, this tesing includes RADIUS server responsiveness. It may also include querying databases of user authentication credentials, or storing accounting data to a billing database. The information obtained from performing those queries is that the entire RADIUS server infrastructure, including all dependencies, is functioning as expected. These queries are most useful in deployments where an administrator has internal RADIUS server that proxy to other internal RADIUS servers, such as for load balancing or fail over. If used, the names used for these test users SHOULD be difficult to guess by an attacker. An Access-Request packet for a test user otherwise should be treated as follows, depending on its origin: o Packets from localhost (127.0.0.1 or ::1). RADIUS servers SHOULD respond with an Access-Accept packet, subject to the limitations outlined in the Table of Attributes in Section 6, below. o Packets from NASes that normally originate Access-Request packets (i.e. not proxy servers). RADIUS servers SHOULD respond with an Access-Reject packet, as the use of Status-Server is preferred. o Packets from other machines controlled by the administrator. RADIUS servers MAY respond with an Access-Accept packet, subject to the limitations outlined in the Table of Attributes in Section 6, below. o Packets originating from machines not controlled by the administrator. RADIUS servers MUST respond with an Access-Reject packet. If a RADIUS server is configured to support test users for Accounting-Request packets, it MAY respond with an Accounting- Response packet, independent of the origin of the request. However, any subsequent analysis of the accounting data such as billing or usage MUST NOT include the data for the test user. If these recommendations are implemented, it may be possible in some situations to safely query a RADIUS server for responsiveness using Access-Request or Accounting-Request packets. DeKok, Alan Informational [Page 16] INTERNET-DRAFT Status-Server Practices 23 February 2007 5.2. RADIUS over reliable transports Although RADIUS has been assigned two TCP ports (1812/tcp and 1813/tcp) in addition to the normally used UDP ports, there has been as yet no implementations using TCP as a reliable transport for RADIUS. If an implementation were to be created, then the transport issues discussed in [RFC3539] would apply. Further, when RADIUS is run over reliable transports, the watchdog algorithm described in [RFC3539] Section 3.4 MUST be used rather than the algorithm described above. Status-Server SHOULD be the packet used as the watchdog request, in preference to Access-Request or Accounting-Request. Clients sending Status-Server over reliable transport MUST ensure that the Identifier field is unique for all requests on a particular connection, independent of the packet code. That is, if a Status- Server with a particular value in the Identifier field is sent to a server, the client MUST NOT simultaneously send an Access-Request or Accounting-Request packet with that same Identifier value, on that connection. Once the client has either received a response to the Status-Server packet, or has determined that the Status-Server packet has timed out, it may re-use that Identifier in an Access-Request or Accounting-Request. 5.3. Other uses for Status-Server While other uses of Status-Server are possible, uses beyond those specified here are beyond the scope of this document. It may be tempting to increase the utility of Status-Server by having the responses carry additional information, implementors are warned that such used have not been analyzed for potential security issues or network problems. 5.4. Potential Uses for Status-Client RADIUS currently defines an experimental Status-Client packet type, in addition to Status-Server. It could be possible to define Status- Client similar to Status-Server, except that it would be applicable to Change of Authorization, and Disconnect Request packets, currently sent to a NAS on port 3799 [RFC3576]. We do no more than mention the possibility here. Any definition of Status-Client is outside of the scope of this document. DeKok, Alan Informational [Page 17] INTERNET-DRAFT Status-Server Practices 23 February 2007 6. Table of Attributes The following table provide a guide to which attributes may be found in Status-Server packets, and in what quantity. No attributes other than the ones listed below should be found in Status-Server packets. Status- Access- Accounting- Server Accept Response # Attribute 0-1 0 0 4 NAS-IP-Address 0 0+ 0 18 Reply-Message 0+ 0+ 0+ 26 Vendor-Specific 0-1 0 0 32 NAS-Identifier 1 0-1 0-1 80 Message-Authenticator 0-1 0 0 95 NAS-IPv6-Address The following table defines the meaning of the above table entries. 0 This attribute MUST NOT be present in packet. 0+ Zero or more instances of this attribute MAY be present in packet. 0-1 Zero or one instance of this attribute MAY be present in packet. 1 Exactly one instance of this attribute MUST be present in packet. 7. Examples A few examples are presented to illustrate the flow of packets to both the authentication and accounting ports. These examples are not intended to be exhaustive, many others are possible. Hexadecimal dumps of the example packets are given in network byte order, using the shared secret "xyzzy5461". 7.1. Minimal Query to Authentication Port The NAS sends a Status-Server UDP packet with minimal content to a RADIUS server on port 1812. The Request Authenticator is a 16 octet random number generated by the NAS. Message-Authenticator is included in order to authenticate that the request came from a known client. 0c da 00 26 8a 54 f4 68 6f b3 94 c5 28 66 e3 02 18 5d 06 23 50 12 5a 66 5e 2e 1e 84 11 f3 e2 43 82 20 97 c8 4f a3 1 Code = Status-Server (12) 1 ID = 218 2 Length = 38 DeKok, Alan Informational [Page 18] INTERNET-DRAFT Status-Server Practices 23 February 2007 16 Request Authenticator Attributes: 18 Message-Authenticator (80) = 5a665e2e1e8411f3e243822097c84fa3 The Response Authenticator is a 16-octet MD5 checksum of the code (2), id (218), Length (20), the Request Authenticator from above, and the shared secret. 02 da 00 14 ef 0d 55 2a 4b f2 d6 93 ec 2b 6f e8 b5 41 1d 66 1 Code = Access-Accept (2) 1 ID = 218 2 Length = 20 16 Request Authenticator Attributes: None. 7.2. Minimal Query to Accounting Port The NAS sends a Status-Server UDP packet with minimal content to a RADIUS server on port 1813. The Request Authenticator is a 16 octet random number generated by the NAS. Message-Authenticator is included in order to authenticate that the request came from a known client. 0c b3 00 26 92 5f 6b 66 dd 5f ed 57 1f cb 1d b7 ad 38 82 60 80 12 e8 d6 ea bd a9 10 87 5c d9 1f da de 26 36 78 58 1 Code = Status-Server (12) 1 ID = 179 2 Length = 38 16 Request Authenticator Attributes: 18 Message-Authenticator (80) = e8d6eabda910875cd91fdade26367858 The Response Authenticator is a 16-octet MD5 checksum of the code (5), id (179), Length (20), the Request Authenticator from above, and the shared secret. 02 b3 00 1a 0f 6f 92 14 5f 10 7e 2f 50 4e 86 0a 48 60 66 9c DeKok, Alan Informational [Page 19] INTERNET-DRAFT Status-Server Practices 23 February 2007 1 Code = Accounting-Response (5) 1 ID = 179 2 Length = 20 16 Request Authenticator Attributes: None. 7.3. Verbose Query and Response The NAS at 192.0.2.16 sends a Status-Server UDP packet to the RADIUS server on port 1812. The Request Authenticator is a 16 octet random number generated by the NAS. 0c 47 00 2c bf 58 de 56 ae 40 8a d3 b7 0c 85 13 f9 b0 3f be 04 06 c0 00 02 10 50 12 85 2d 6f ec 61 e7 ed 74 b8 e3 2d ac 2f 2a 5f b2 1 Code = Status-Server (12) 1 ID = 71 2 Length = 44 16 Request Authenticator Attributes: 6 NAS-IP-Address (4) = 192.0.2.16 18 Message-Authenticator (80) = 852d6fec61e7ed74b8e32dac2f2a5fb2 The Response Authenticator is a 16-octet MD5 checksum of the code (2), id (71), Length (52), the Request Authenticator from above, the attributes in this reply, and the shared secret. The Reply-Message is "RADIUS Server up 2 days, 18:40" 02 47 00 34 46 f4 3e 62 fd 03 54 42 4c bb eb fd 6d 21 4e 06 12 20 52 41 44 49 55 53 20 53 65 72 76 65 72 20 75 70 20 32 20 64 61 79 73 2c 20 31 38 3a 34 30 1 Code = Access-Accept (2) 1 ID = 71 2 Length = 52 16 Request Authenticator Attributes: 32 Reply-Message (18) DeKok, Alan Informational [Page 20] INTERNET-DRAFT Status-Server Practices 23 February 2007 8. IANA Considerations This specification does not create any new registries, nor does it require assignment of any protocol parameters. 9. Security Considerations This document defines the Status-Server packet as being similar in treatment to the Access-Request packet, and is therefore subject to the same security considerations as described in [RFC2865], Section 8. Status-Server packets also use the Message-Authenticator attribute, and are therefore subject to the same security considerations as [RFC3579], Section 4. We reiterate that Status-Server packets MUST contain a Message- Authenticator attribute. Early implementations supporting Status- Server may not have enforced this requirement, and may have been subject to DoS attacks as a result. Where this document differs from [RFC2865] is that it defines a new request/response method in RADIUS; the Status-Server "ping". As this use is based on previously described and implemented standards, we know of no additional security considerations that arise from the use of Status-Server as defined herein. 10. References 10.1. Normative references [RFC2865] Rigney, C., Willens, S., Rubens, A. and W. Simpson, "Remote Authentication Dial In User Service (RADIUS)", RFC 2865, June 2000. [RFC3579] Aboba, B., Calhoun, P., "RADIUS (Remote Authentication Dial In User Service) Support For Extensible Authentication Protocol (EAP)", RFC 3579, September 2003. 10.2. Informative references [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", RFC 2119, March, 1997. [RFC3539] Aboba, B., Wood, J., "Authentication, Authorization, and Accounting (AAA) Transport Profile", RFC 3539, June 2003. [RFC3576] Chiba, M., Dommety, G., Eklund, M., Mitton, D., Aboba, B., "Dynamic Authorization Extensions to Remote Authentication DeKok, Alan Informational [Page 21] INTERNET-DRAFT Status-Server Practices 23 February 2007 Dial In User Service (RADIUS)", RFC 3576, July 2003. [RFC4668] Nelson, D., "RADIUS Authentication Client MIB for IPv6", RFC 4668, August 2006. [RFC4669] Nelson, D., "RADIUS Authentication Server MIB for IPv6", RFC 4669, August 2006. [RFC4670] Nelson, D., "RADIUS Accounting Client MIB for IPv6", RFC 4670, August 2006. [RFC4671] Nelson, D., "RADIUS Accounting Server MIB for IPv6", RFC 4671, August 2006. Acknowledgments Parts of the text in Section 3 defining the Request and Response Authenticators were taken with minor edits from [RFC2865] Section 3. The author would like to thank Mike McCauley of Open Systems Consultants for making a Radiator server available for inter- operability testing. Authors' Addresses Alan DeKok The FreeRADIUS Server Project http://freeradius.org Email: aland@freeradius.org Intellectual Property Statement The IETF takes no position regarding the validity or scope of any Intellectual Property Rights or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; nor does it represent that it has made any independent effort to identify any such rights. Information on the procedures with respect to rights in RFC documents can be found in BCP 78 and BCP 79. Copies of IPR disclosures made to the IETF Secretariat and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementers or users of this specification can be obtained from the IETF on-line IPR repository at DeKok, Alan Informational [Page 22] INTERNET-DRAFT Status-Server Practices 23 February 2007 http://www.ietf.org/ipr. The IETF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights that may cover technology that may be required to implement this standard. Please address the information to the IETF at ietf- ipr@ietf.org. DeKok, Alan Informational [Page 23] INTERNET-DRAFT Status-Server Practices 23 February 2007 Disclaimer of Validity This document and the information contained herein are provided on an "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Full Copyright Statement Copyright (C) The IETF Trust (2007). This document is subject to the rights, licenses and restrictions contained in BCP 78, and except as set forth therein, the authors retain all their rights. Acknowledgment Funding for the RFC Editor function is currently provided by the Internet Society. DeKok, Alan Informational [Page 24]