Network Working Group E. Lear Request for Comments: 4744 Cisco Systems Category: Standards Track K. Crozier December 2006
Using the NETCONF Protocol over the Blocks Extensible Exchange Protocol (BEEP)
Status of This Memo
This document specifies an Internet standards track protocol for the Internet community, and requests discussion and suggestions for improvements. Please refer to the current edition of the "Internet Official Protocol Standards" (STD 1) for the standardization state and status of this protocol. Distribution of this memo is unlimited.
Copyright Notice
Copyright (C) The IETF Trust (2006).
Abstract
This document specifies an application protocol mapping for the Network Configuration Protocol (NETCONF) over the Blocks Extensible Exchange Protocol (BEEP).
The NETCONF protocol [1] defines a simple mechanism through which a network device can be managed. NETCONF is designed to be usable over a variety of application protocols. This document specifies an application protocol mapping for NETCONF over the Blocks Extensible Exchange Protocol (BEEP) [7].
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 RFC 2119 [2].
Use of BEEP is natural as an application protocol for transport of XML. As a peer-to-peer protocol, BEEP provides an easy way to implement NETCONF, no matter which side of the connection was the initiator. This "bidirectionality" allows for either manager or agent to initiate a connection. This is particularly important to support large numbers of intermittently connected devices, as well as those devices that must reverse the management connection in the face of firewalls and network address translators (NATs).
BEEP makes use of the Simple Authentication and Security Layer (SASL) [3]. The SASL profile used by BEEP allows for a simple and direct mapping to the existing security model for command line interface (CLI), while Transport Layer Security (TLS) [4] provides a strong, well-tested encryption mechanism with either server or server and client-side authentication.
All NETCONF over BEEP implementations MUST implement the profile and functional mapping between NETCONF and BEEP as described below.
For purposes of this document, a manager is a NETCONF client, and an agent is a NETCONF server. Use of client/server language in BEEP is avoided because of the common notion that in networking clients connect to servers.
Managers may be either BEEP listeners or initiators. Similarly, agents may be either listeners or initiators. To establish a connection, the initiator connects to the listener on TCP port 831. Thus, the initial exchange takes place without regard to whether a manager or the agent is the initiator. After the transport connection is established, as greetings are exchanged, they SHOULD
Lear & Crozier Standards Track [Page 2]
RFC 4744 NETCONF over BEEP December 2006
each announce their support for TLS and optionally SASL. Once BEEP greeting messages are exchanged, if TLS is to be used and available by both parties, the listener STARTs a channel with the TLS profile.
Once TLS has been started, a new BEEP greeting message is sent by both initiator and listener, as required by the BEEP RFC.
After all BEEP greeting messages are exchanged in order for roles to be clear, the agent MUST advertise the NETCONF profile. The manager MUST NOT advertise the NETCONF profile. If the agent side of the communication (either initiator or listener) receives a BEEP <greeting> element that contains the NETCONF profile, it MUST close the connection. Similarly, if neither side issues a NETCONF profile it is equally an error, and the listener MUST close the connection.
At this point, if SASL is desired, the initiator starts a BEEP channel to perform a SASL exchange to authenticate itself. Upon completion of authentication the channel is closed. That is, the channel is exclusively used to authenticate.
Examples of both TLS and SASL profiles can be found in [7].
It is anticipated that the SASL PLAIN mechanism will be heavily used in conjunction with TLS [5]. In such cases, in accordance with RFC 2595 the PLAIN mechanism MUST NOT be advertised in the first BEEP <greeting>, but only in the one following a successful TLS negotiation. This applies only if TLS and SASL PLAIN mechanisms are both to be used. To avoid risk of eavesdropping, the SASL PLAIN mechanism MUST NOT be used over unencrypted channels. More specifics about the use of SASL and TLS are mentioned in Security Considerations below.
Once authentication has occurred, there is no need to distinguish between initiator and listener. We now distinguish between manager and agent, and it is assumed that each knows its role in the conversation.
At this point, we are ready to proceed on BEEP channel 1 with NETCONF operations.
NETCONF messages are transmitted with a Content-type header set to "text/xml".
Next the manager and the agent exchange NETCONF <hello> elements on the new channel so that each side learns the other's capabilities. This occurs through a MSG. Each side will then respond positively. The following example is adapted from [1] Section 8.1:
Nearly all NETCONF operations are executed through the <rpc> element. To issue a remote procedure call (RPC), the manager transmits on the operational channel a BEEP MSG containing the RPC and its arguments. In accordance with the BEEP standard, RPC requests may be split across multiple BEEP frames.
Once received and processed, the agent responds with BEEP RPY messages on the same channel with the response to the RPC. In accordance with the BEEP standard, responses may be split across multiple BEEP frames.
Upon receipt of <close-session> from the manager, once the agent has completed all RPCs, it will close BEEP channel 0. When an agent needs to initiate a close, it will do so by closing BEEP channel 0. Although not required to do so, the agent should allow for a reasonable period for a manager to release an existing lock prior to initiating a close. Once the agent has closed channel 0, all locks are released, and each side follows teardown procedures as specified in [8]. Having received a BEEP close or having sent <close-session>, a manager MUST NOT send further requests. If there are additional activities due to expanded capabilities, they MUST cease in an orderly manner and should be properly described in the capability mapping.
Configuration information is by its very nature sensitive. Its transmission in the clear and without integrity checking leaves devices open to classic so-called "person-in-the-middle" attacks. Configuration information often times contains passwords, user names, service descriptions, and topological information, all of which are sensitive. A NETCONF application protocol, therefore, must minimally support options for both confidentiality and authentication.
The BEEP mapping described in this document addresses both confidentiality and authentication in a flexible manner through the use of TLS and SASL profiles. Confidentiality is provided via the TLS profile and is used as discussed above. In addition, the server certificate shall serve as the server's authentication to the client. The client MUST be prepared to recognize and validate a server certificate and SHOULD by default reject invalid certificates.
In order to validate a certificate, the client must be able to access a trust anchor. While such validation methods are beyond the scope of this document, they will depend on the type of device and circumstance. Both the implementor and the administrator are cautioned to be aware of any circular dependencies that various methods may introduce. For instance, Online Certificate Status Protocol (OCSP) servers may not be available in a network cold-start scenario and would be ill-advised for core routers to depend on to receive configuration at boot.
Lear & Crozier Standards Track [Page 6]
RFC 4744 NETCONF over BEEP December 2006
For client-side authentication, there are several options. The client MAY provide a certificate during the initiation phase of TLS, in which case the subject of that certificate shall be considered principle for authentication purposes. Once again, server implementors should be aware of any interdependencies that could be created through protocols used to validate trust anchors.
TLS endpoints may be authorized based on subject name or certificate authority (CA), depending on circumstances. For instance, it would be unwise for a core Internet router to allow a netconf agent connection simply based on a valid certificate signed by a common CA, but not unreasonable to allow a connection from an agent with a particular distinguished name. On the other hand, it might be desirable for enterprises to trust certificates signed by CAs of their network operations team.
In the case where the client has not authenticated through TLS, the server SHOULD advertise one or more SASL profiles, from which the client will choose. In the singular case where TLS is established, the minimum profile MAY be PLAIN. Otherwise, implementations MUST support the DIGEST-MD5 profile as described in [6], and they MAY support other profiles such as the One-Time Password (OTP) mechanism [10].
Different environments may well allow different rights prior to and then after authentication. An authorization model is not specified in this document. When an operation is not properly authorized, then a simple rpc-error containing "permission denied" is sufficient. Note that authorization information may be exchanged in the form of configuration information, which is all the more reason to ensure the security of the connection.
This work is the product of the NETCONF IETF working group, and many people have contributed to the NETCONF discussion. Most notably, Rob Ens, Phil Schafer, Andy Bierman, Wes Hardiger, Ted Goddard, and Margaret Wasserman all contributed in some fashion to this work, which was originally to be found in the NETCONF base protocol specification. Thanks also to Weijing Chen, Keith Allen, Juergen Schoenwaelder, Marshall Rose, and Eamon O'Tuathail for their very constructive participation. The authors would also like to thank Elwyn Davies for his constructive review.
[9] Sperberg-McQueen, C., Paoli, J., Maler, E., and T. Bray, "Extensible Markup Language (XML) 1.0 (Second Edition)", World Wide Web Consortium, First Edition, http://www.w3.org/TR/2000/REC-xml-20001006, October 2000.
[10] Newman, C., "The One-Time-Password SASL Mechanism", RFC 2444, October 1998.
Lear & Crozier Standards Track [Page 8]
RFC 4744 NETCONF over BEEP December 2006
Authors' Addresses
Eliot Lear Cisco Systems Glatt-com CH-8301 Glattzentrum, Zurich CH
EMail: lear@cisco.com
Ken Crozier
EMail: ken.crozier@gmail.com
Lear & Crozier Standards Track [Page 9]
RFC 4744 NETCONF over BEEP December 2006
Full Copyright Statement
Copyright (C) The IETF Trust (2006).
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.
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.
Intellectual Property
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 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.
Acknowledgement
Funding for the RFC Editor function is currently provided by the Internet Society.