DHC Working Group S. Jiang, Ed. Internet-Draft Huawei Technologies Co., Ltd Intended status: Standards Track S. Shen Expires: December 21, 2014 CNNIC D. Zhang Huawei Technologies Co., Ltd T. Jinmei WIDE Project June 19, 2014 Secure DHCPv6 with Public Key draft-ietf-dhc-sedhcpv6-03 Abstract The Dynamic Host Configuration Protocol for IPv6 (DHCPv6) enables DHCPv6 servers to pass configuration parameters. It offers configuration flexibility. If not secured, DHCPv6 is vulnerable to various attacks, particularly spoofing attacks. This document analyzes the security issues of DHCPv6 and specifies a Secure DHCPv6 mechanism for communication between DHCPv6 clients and DHCPv6 servers. This mechanism is based on public/private key pairs. The authority of the sender may depend on either pre-configuration mechanism or Public Key Infrastructure. Status of This Memo This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet- Drafts is at http://datatracker.ietf.org/drafts/current/. 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." This Internet-Draft will expire on December 21, 2014. Copyright Notice Copyright (c) 2014 IETF Trust and the persons identified as the document authors. All rights reserved. Jiang, et al. Expires December 21, 2014 [Page 1] Internet-Draft SeDHCPv6 June 2014 This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 2. Requirements Language and Terminology . . . . . . . . . . . . 3 3. Security Overview of DHCPv6 . . . . . . . . . . . . . . . . . 3 4. Overview of Secure DHCPv6 Mechanism with Public Key . . . . . 4 4.1. New Components . . . . . . . . . . . . . . . . . . . . . 6 4.2. Support for algorithm agility . . . . . . . . . . . . . . 6 4.3. Applicability . . . . . . . . . . . . . . . . . . . . . . 7 5. Extensions for Secure DHCPv6 . . . . . . . . . . . . . . . . 7 5.1. Public Key Option . . . . . . . . . . . . . . . . . . . . 7 5.2. Certificate Option . . . . . . . . . . . . . . . . . . . 8 5.3. Signature Option . . . . . . . . . . . . . . . . . . . . 9 5.4. Status Codes . . . . . . . . . . . . . . . . . . . . . . 10 6. Processing Rules and Behaviors . . . . . . . . . . . . . . . 11 6.1. Processing Rules of Sender . . . . . . . . . . . . . . . 11 6.2. Processing Rules of Recipient . . . . . . . . . . . . . . 12 6.3. Processing Rules of Relay Agent . . . . . . . . . . . . . 14 6.4. Timestamp Check . . . . . . . . . . . . . . . . . . . . . 14 7. Deployment Consideration . . . . . . . . . . . . . . . . . . 16 7.1. Authentication on a client . . . . . . . . . . . . . . . 16 7.2. Authentication on a server . . . . . . . . . . . . . . . 16 8. Security Considerations . . . . . . . . . . . . . . . . . . . 17 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 18 10. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 20 11. Change log [RFC Editor: Please remove] . . . . . . . . . . . 20 12. References . . . . . . . . . . . . . . . . . . . . . . . . . 21 12.1. Normative References . . . . . . . . . . . . . . . . . . 21 12.2. Informative References . . . . . . . . . . . . . . . . . 22 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 22 1. Introduction The Dynamic Host Configuration ProtocoFl for IPv6 (DHCPv6, [RFC3315]) enables DHCPv6 servers to pass configuration parameters. It offers configuration flexibility. If not secured, DHCPv6 is vulnerable to various attacks, particularly spoofing attacks. Jiang, et al. Expires December 21, 2014 [Page 2] Internet-Draft SeDHCPv6 June 2014 This document analyzes the security issues of DHCPv6 in details. This document provides mechanisms for improving the security of DHCPv6 between client and server: o the identity of a DHCPv6 message sender, which can be a DHCPv6 server or a client, can be verified by a recipient. o the integrity of DHCPv6 messages can be checked by the recipient of the message. o anti-replay protection based on timestamp checking. Note: this secure mechanism in this document does not protect the relay-relevant options, either added by a relay agent toward a server or added by a server toward a relay agent, are considered less vulnerable, because they are only transported within operator networks. Communication between a server and a relay agent, and communication between relay agents, may be secured through the use of IPsec, as described in section 21.1 in [RFC3315]. The security mechanisms specified in this document is based on self- generated public/private key pairs. It also integrates timestamps for anti-replay. The authentication procedure defined in this document may depend on either deployed Public Key Infrastructure (PKI, [RFC5280]) or pre-configured sender's public key. However, the deployment of PKI or pre-configuration is out of the scope. Secure DHCPv6 is applicable in environments where physical security on the link is not assured (such as over wireless) and attacks on DHCPv6 are a concern. 2. Requirements Language and Terminology The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC2119] when they appear in ALL CAPS. When these words are not in ALL CAPS (such as "should" or "Should"), they have their usual English meanings, and are not to be interpreted as [RFC2119] key words. 3. Security Overview of DHCPv6 DHCPv6 is a client/server protocol that provides managed configuration of devices. It enables DHCPv6 server to automatically configure relevant network parameters on clients. In the basic DHCPv6 specification [RFC3315], security of DHCPv6 message can be improved. Jiang, et al. Expires December 21, 2014 [Page 3] Internet-Draft SeDHCPv6 June 2014 The basic DHCPv6 specifications can optionally authenticate the origin of messages and validate the integrity of messages using an authentication option with a symmetric key pair. [RFC3315] relies on pre-established secret keys. For any kind of meaningful security, each DHCPv6 client would need to be configured with its own secret key; [RFC3315] provides no mechanism for doing this. For the key of the hash function, there are two key management mechanisms. Firstly, the key management is done out of band, usually through some manual process. For example, operators can set up a key database for both servers and clients which the client obtains a key before running DHCPv6. Manual key distribution runs counter to the goal of minimizing the configuration data needed at each host. [RFC3315] provides an additional mechanism for preventing off-network timing attacks using the Reconfigure message: the Reconfigure Key authentication method. However, this method provides no message integrity or source integrity check. This key is transmitted in plaintext. In comparison, the public/private key security mechanism allows the keys to be generated by the sender, and allows the public key database on the recipient to be populated opportunistically or manually, depending on the degree of confidence desired in a specific application. PKI security mechanism is simpler in the local key management respect. 4. Overview of Secure DHCPv6 Mechanism with Public Key In order to enable a DHCPv6 client and a server mutually authenticate each other without previous key deployment, this document introduces the use of public/private key pair mechanism into DHCPv6, also with timestamp. The authority of the sender may depend on either pre- configuration mechanism or PKI. By combining with the signatures, sender identity can be verified and messages protected. This document introduces a Secure DHCPv6 mechanism that uses a public/private key pair to secure the DHCPv6 protocol. In order to enable DHCPv6 clients and DHCPv6 servers to perform mutual authentication, this solution provides two public key based mechanisms with different security strengths. One is stronger and only the certificate signed by a trusted CA or preconfigured public key can be accepted. The other one, called as leap of faith (LoF) mechanism, is relatively weak. It allows a client/server pair that lacks essential trust relationship to build up their trust relationship at run time for subsequent exchanges based on faith. Jiang, et al. Expires December 21, 2014 [Page 4] Internet-Draft SeDHCPv6 June 2014 This design simplifies the precondition of deploying DHCPv6 authentication and provides limited protection of DHCPv6 message. In the proposed solution, either public/private key pairs or certificates can be used in authentication. When using public/ private key pairs directly, the public key of the sender is pre- shared with the recipient, either opportunistically or through a manual process. When using certificates, the sender has a certificate for its public key, signed by a CA that is trusted by the recipient. It is possible for the same public key to be used with different recipients in both modes. In this document, we introduce a public key option, a certificate option and a signature option with a corresponding verification mechanism. Timestamp is integrated into signature options. A DHCPv6 message (from a server or a client), with either a public key or certificate option, and carrying a digital signature, can be verified by the recipient for both the timestamp and authentication, then process the payload of the DHCPv6 message only if the validation is successful. Because the sender can be a DHCPv6 server or a client, the end-to-end security protection can be from DHCPv6 servers to clients or from clients to DHCPv6 servers. The recipient may choose to further process the message from a sender for which no authentication information exists, either non-matched public key or certificate cannot be verified. By recording the public key or unverifiable certificate that was used by the sender, when the first time it is seen, the recipient can make a leap of faith that the sender is trustworthy. If no evidence to the contrary surfaces, the recipient can then validate the sender as trustworthy when it subsequently sees the same public key or certificate used to sign messages from the same sender. In opposite, once the recipient has determined that it is being attacked, it can either forget that sender, or remember that sender in a blacklist and drop further packets associated with that sender. This improves communication security of DHCPv6 messages. Secure DHCPv6 messages are commonly large. IP fragments [RFC2460] are highly possible. Hence, deployment of Secure DHCPv6 should also consider the issues of IP fragment, PMTU, etc. Also, if there are firewalls between secure DHCPv6 clients and secure DHCPv6 servers, it is RECOMMENDED that the firewalls are configureed to pass ICMP Packet Too Big messages [RFC4443]. Jiang, et al. Expires December 21, 2014 [Page 5] Internet-Draft SeDHCPv6 June 2014 4.1. New Components The components of the solution specified in this document are as follows: o The node generates a public/private key pair. A DHCPv6 option is defined that carries the public key. The node may also obtain a certificate from a Certificate Authority that can be used to establish the trustworthiness of the node. Another option is defined to carry the certificate. Because the certificate contains the public key, there is never a need to send both options at the same time. o A signature generated using the private key that protects the integrity of the DHCPv6 messages and authenticates the identity of the sender. o A timestamp, to detect and prevent packet replay. The secure DHCPv6 nodes need to meet some accuracy requirements and be synced to global time, while the timestamp checking mechanism allows a configurable time value for clock drift. The real time provision is out of scope. 4.2. Support for algorithm agility Hash functions are used to provide message integrity checks. In order to provide a means of addressing problems that may emerge in the future with existing hash algorithms, as recommended in [RFC4270], this document provides a mechanism for negotiating the use of more secure hashes in the future. In addition to hash algorithm agility, this document also provides a mechanism for signature algorithm agility. The support for algorithm agility in this document is mainly a unilateral notification mechanism from sender to recipient. A recipient MAY support various algorithms simultaneously, and the differenet senders in a same administrative domain may be allowed to use various algorithms simultaneously. If the recipient does not support the algorithm used by the sender, it cannot authenticate the message. In the client-to-server case, the server SHOULD reply with a AlgorithmNotSupported status code (defined in Section 5.4). Upon receiving this status code, the client MAY resend the message protected with the mandatory algorithm (defined in Section 5.3). Jiang, et al. Expires December 21, 2014 [Page 6] Internet-Draft SeDHCPv6 June 2014 4.3. Applicability By default, a secure DHCPv6 enabled client SHOULD start with secure mode by sending secure DHCPv6 messages. If the recipient is secure DHCPv6 enabled server, their communication would be in secure mode. In the scenario where the secure DHCPv6 enabled client and server fail to build up secure communication between them, the secure DHCPv6 enabled client MAY choose to send unsecured DHCPv6 message towards the server. A secure DHCPv6 enabled server MAY also provide services for unsecured clients. In such case, the resources allocated for unsecured clients SHOULD be separated and restricted, in order to protect against bidding down attacks. In the scenario where the recipient is a legacy DHCPv6 server that does not support secure mechanism, the DHCPv6 server (for all of known DHCPv6 implementations) would just omit or disregard unknown options (secure options defined in this document) and still process the known options. The reply message would be unsecured, of course. It is up to the local policy of the client whether to accept the messages. If the client accepts the unsecured messages from the DHCPv6 server, the subsequent exchanges will be in the unsecured mode. In the scenario where a legacy client sends an unsecured message to a secure DHCPv6 enabled server, there are two possibilities depending on the server policy. If the server's policy requires the authentication, an UnspecFail (value 1, [RFC3315]) error status code, SHOULD be returned. In such case, the client cannot build up the connection with the server. If the server has been configured to support unsecured clients, the server would fall back to the unsecured DHCPv6 mode, and reply unsecured messages toward the client. The resources allocated for unsecured clients SHOULD be separated and restricted. 5. Extensions for Secure DHCPv6 This section extends DHCPv6. Three new options have been defined. The new options MUST be supported in the Secure DHCPv6 message exchange. 5.1. Public Key Option The Public Key option carries the public key of the sender. The format of the Public Key option is described as follows: Jiang, et al. Expires December 21, 2014 [Page 7] Internet-Draft SeDHCPv6 June 2014 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | OPTION_PK_PARAMETER | option-len | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | . Public Key (variable length) . . . +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ option-code OPTION_PK_PARAMETER (TBA1). option-len Length of public key in octets. Public Key A variable-length field containing public key and identify the algorithm with which the key is used (e.g., RSA, DSA, or Diffie-Hellman). The algorithm is identified using the AlgorithmIdentifier structure specified in section 4.1.1.2, [RFC5280]. The object identifiers for the supported algorithms and the methods for encoding the public key materials (public key and parameters) are specified in [RFC3279], [RFC4055], and [RFC4491]. 5.2. Certificate Option The Certificate option carries the certificate of the sender. The format of the Certificate option is described as follows: 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | OPTION_CERT_PARAMETER | option-len | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | . Certificate (variable length) . . . +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ option-code OPTION_CERT_PARAMETER (TBA2). option-len Length of certificate in octets. Certificate A variable-length field containing certificate. The encoding of certificate and certificate data MUST be in format as defined in Section 3.6, [RFC5996]. The support of X.509 certificate is mandatory. The length of a certificate is various. Jiang, et al. Expires December 21, 2014 [Page 8] Internet-Draft SeDHCPv6 June 2014 5.3. Signature Option The Signature option allows public key-based signatures to be attached to a DHCPv6 message. The Signature option could be any place within the DHCPv6 message. It protects the entire DHCPv6 header and options, including itself, except for the Authentication Option. The format of the Signature option is described as follows: 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | OPTION_SIGNATURE | option-len | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | HA-id | SA-id | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | Timestamp (64-bit) | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | | . Signature (variable length) . . . +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ option-code OPTION_SIGNATURE (TBA3). option-len 10 + Length of Signature field in octets. HA-id Hash Algorithm id. The hash algorithm is used for computing the signature result. This design is adopted in order to provide hash algorithm agility. The value is from the Hash Algorithm for Secure DHCPv6 registry in IANA. The support of SHA-256 is mandatory. A registry of the initial assigned values is defined in Section 8. SA-id Signature Algorithm id. The signature algorithm is used for computing the signature result. This design is adopted in order to provide signature algorithm agility. The value is from the Signature Algorithm for Secure DHCPv6 registry in IANA. The support of RSASSA-PKCS1-v1_5 is mandatory. A registry of the initial assigned values is defined in Section 8. Timestamp The current time of day (NTP-format timestamp [RFC5905] in UTC (Coordinated Universal Time), a 64-bit unsigned fixed-point number, in seconds Jiang, et al. Expires December 21, 2014 [Page 9] Internet-Draft SeDHCPv6 June 2014 relative to 0h on 1 January 1900.). It can reduce the danger of replay attacks. Signature A variable-length field containing a digital signature. The signature value is computed with the hash algorithm and the signature algorithm, as described in HA-id and SA-id. The signature constructed by using the sender's private key protects the following sequence of octets: 1. The DHCPv6 message header. 2. All DHCPv6 options including the Signature option (fill the signature field with zeroes) except for the Authentication Option. The signature filed MUST be padded, with all 0, to the next octet boundary if its size is not an even multiple of 8 bits. The padding length depends on the signature algorithm, which is indicated in the SA-id field. Note: if both signature and authentication option are presented, signature option does not protect the Authentication Option. It allows to be created after signature has been calculated and filled with the valid signature. It is because both needs to apply hash algorithm to whole message, so there must be a clear order and there could be only one last-created option. In order to avoid update [RFC3315] because of changing auth option, the authors chose not include authentication option in the signature. 5.4. Status Codes o AlgorithmNotSupported (TBD4): indicates that the DHCPv6 server does not support algorithms that sender used. o AuthFailNotSupportLoF (TBD5): indicates that the DHCPv6 client fails authentication check and the DHCPv6 server does not support the leaf of faith mode o AuthFailSupportLoF (TBD6): indicates that the DHCPv6 client fails authentication check. Although the DHCPv6 server does support the leaf of faith, its list that stores public keys or unverifiable certificates in the leap of faith mode currently exceeds. o TimestampFail (TBD7): indicates the message from DHCPv6 client fails the timstamp check. Jiang, et al. Expires December 21, 2014 [Page 10] Internet-Draft SeDHCPv6 June 2014 o SignatureFail (TBD8): indicates the message from DHCPv6 client fails the signature check. 6. Processing Rules and Behaviors This section only covers the scenario where both DHCPv6 client and DHCPv6 server are secure enabled. 6.1. Processing Rules of Sender The sender of a Secure DHCPv6 message could be a DHCPv6 server or a DHCPv6 client. The node must have a public/private key pair in order to create Secure DHCPv6 messages. The node may have a certificate which is signed by a CA trusted by both sender and recipient. To support secure DHCPv6, the secure DHCPv6 enabled sender MUST construct the DHCPv6 message following the rules defined in [RFC3315]. A Secure DHCPv6 message, except for Relay-forward and Relay-reply messages, MUST contain either a Public Key or a Certificate option, which MUST be constructed as explained in Section 5.1 or Section 5.2. A Secure DHCPv6 message, except for Relay-forward and Relay-reply messages, MUST contain one and only one Signature option, which MUST be constructed as explained in Section 5.3. It protects the message header and all DHCPv6 options except for the Authentication Option. Within the Signature option the Timestamp field SHOULD be set to the current time, according to sender's real time clock. A Relay-forward and relay-reply message MUST NOT contain any additional Public Key or Certificate option or Signature Option, aside from those present in the innermost encapsulated message from the client or server. If the sender is a DHCPv6 client, in the failure cases, it receives a Reply message with an error status code. The error status code indicates the failure reason on the server side. According to the received status code, the client MAY take follow-up action: o Upon receiving a AlgorithmNotSupported error status code, the client MAY resend the message protected with the mandatory algorithms. o Upon receiving an AuthFailNotSupportLoF error status code, the client is not able to build up the secure communication with the Jiang, et al. Expires December 21, 2014 [Page 11] Internet-Draft SeDHCPv6 June 2014 recipient. The client MAY switch to other certificate or public key if it has. But it SHOULD NOT retry with the same certificate/ public-key. It MAY retry with the same certificate/public-key following normal retransmission routines defined in [RFC3315]. o Upon receiving an AuthFailSupportLoF error status code, the client is not able to build up the secure communication with the recipient. The client MAY switch to other certificate or public key if it has. The client MAY retry with the same certificate/ public-key following normal retransmission routines defined in [RFC3315]. o Upon receiving a TimestampFail error status code, the client MAY fall back to unsecured mode. o Upon receiving a SignatureFail error status code, the client MAY resend the message following normal retransmission routines defined in [RFC3315]. 6.2. Processing Rules of Recipient The recipient of a Secure DHCPv6 message could be a DHCPv6 server or a DHCPv6 client. In the failure cases, both DHCPv6 server and client SHOULD NOT process received message, and the server SHOULD reply a correspondent error status code, while the client does nothing. The specific behavior depends on the configured local policy. When receiving a DHCPv6 message, except for Relay-Forward and Relay- Reply messages, a secure DHCPv6 enabled recipient SHOULD discard the DHCPv6 message if the Signature option is absent, or multiple Signature option is presented, or both the Public Key and Certificate options are absent, or both the Public Key and Certificate option are presented. In such failure, the DHCPv6 server SHOULD reply an UnspecFail (value 1, [RFC3315]) error status code. If all three options are absent, the sender MAY be legacy node or in unsecured mode, then, the recipient MAY fall back to the unsecured DHCPv6 mode if its local policy allows. The recipient SHOULD first check the support of algorithms that sender used. If all algorithms are supported, the recipient then checks the authority of this sender. If not, the message is dropped. In such failure, the DHCPv6 server SHOULD reply a AlgorithmNotSupported error status code, defined in Section 5.4, back to the client. If the sender uses certificate, the recipient SHOULD validate the sender's certificate following the rules defined in [RFC5280]. An implementation may create a local trust certificate record for a Jiang, et al. Expires December 21, 2014 [Page 12] Internet-Draft SeDHCPv6 June 2014 verified certificate in order to avoid repeated verification procedure in the future. A sender certificate that finds a match in the local trust certificate list is treated as verified. A fast search index may be created for this list. If the sender uses a public key, the recipient SHOULD validate it by finding a matching public key from the local trust public key list, which is pre-configured or recorded from previous communications. A local trust public key list is a data table maintained by the recipient. It restores public keys from all trustworthy senders. A fast search index may be created for this list. The recipient may choose to further process the message from a sender for which no authentication information exists, either non-matched public key or certificate cannot be verified. By recording the public key or unverifiable certificate that was used by the sender, when the first time it is seen, the recipient can make a leap of faith (LoF) that the sender is trustworthy. If no evidence to the contrary surfaces, the recipient can then validate the sender as trustworthy for subsequent message exchanges. In opposite, once the recipient has determined that it is being attacked, it can either forget that key, or remember that key in a blacklist and drop further packets associated with that key. If recipient does not support the leap of faith mode, the message that fails authentication check MUST be dropped. In such failure, the DHCPv6 server SHOULD reply an AuthFailNotSupportLoF error status code, defined in Section 5.4, back to the client. On the recipient that supports the leap of faith mode, the number of cached public keys or unverifiable certificates MAY be limited in order to protect against resource exhaustion attacks. If the recipient's list that stores public keys or unverifiable certificates in the leap of faith mode exceeds, the message that fails authentication check MUST be dropped. In such failure, the DHCPv6 server SHOULD reply an AuthFailNotSupportLoF error status code, defined in Section 5.4, back to the client. The resource releasing policy against exceeding situations is out of scope. Giving the complexity, the key rollover mechanism is out of scope of this document. At this point, the recipient has either recognized the authentication of the sender, or decided to attempt a leap of faith. The recipient MUST now authenticate the sender by verifying the Signature and checking timestamp (see details in Section 6.4). The order of two procedures is left as an implementation decision. It is RECOMMENDED to check timestamp first, because signature verification is much more computationally expensive. Jiang, et al. Expires December 21, 2014 [Page 13] Internet-Draft SeDHCPv6 June 2014 The signature field verification MUST show that the signature has been calculated as specified in Section 5.3. Only the messages that get through both the signature verifications and timestamp check are accepted as secured DHCPv6 messages and continue to be handled for their contained DHCPv6 options as defined in [RFC3315]. Messages that do not pass the above tests MUST be discarded or treated as unsecured messages. In the case the recipient is DHCPv6 server, the DHCPv6 server SHOULD reply a SignatureFail error status code, defined in Section 5.4, for the signature verification failure, or a TimestampFail error status code, defined in Section 5.4, for the timestamp check failure, back to the client. Furthermore, the node that supports the verification of the Secure DHCPv6 messages MAY record the following information: Minbits The minimum acceptable key length for public keys. An upper limit MAY also be set for the amount of computation needed when verifying packets that use these security associations. The appropriate lengths SHOULD be set according to the signature algorithm and also following prudent cryptographic practice. For example, minimum length 1024 and upper limit 2048 may be used for RSA [RSA]. A Relay-forward or Relay-reply message with any Public Key, Certificate or the Signature option is invalid. The message MUST be discarded silently. 6.3. Processing Rules of Relay Agent To support Secure DHCPv6, relay agents just need to follow the same processing rules defined in [RFC3315]. There is nothing more the relay agents have to do, either verify the messages from client or server, or add any secure DHCPv6 options. Actually, by definition in this document, relay agents SHOULD NOT add any secure DHCPv6 options. 6.4. Timestamp Check Recipients SHOULD be configured with an allowed timestamp Delta value, a "fuzz factor" for comparisons, and an allowed clock drift parameter. The recommended default value for the allowed Delta is 300 seconds (5 minutes); for fuzz factor 1 second; and for clock drift, 0.01 second. Note: the Timestamp mechanism is based on the assumption that communication peers have roughly synchronized clocks, with certain allowed clock drift. So, accurate clock is not necessary. If one has a clock too far from the current time, the timestamp mechanism would not work. Jiang, et al. Expires December 21, 2014 [Page 14] Internet-Draft SeDHCPv6 June 2014 To facilitate timestamp checking, each recipient SHOULD store the following information for each sender, from which at least one accepted secure DHCPv6 message is successfully verified (for both timestamp check and signature verification): o The receive time of the last received and accepted DHCPv6 message. This is called RDlast. o The timestamp in the last received and accepted DHCPv6 message. This is called TSlast. A verified (for both timestamp check and signature verification) secure DHCPv6 message initiates the update of the above variables in the recipient's record. Recipients MUST check the Timestamp field as follows: o When a message is received from a new peer (i.e., one that is not stored in the cache), the received timestamp, TSnew, is checked, and the message is accepted if the timestamp is recent enough to the reception time of the packet, RDnew: -Delta < (RDnew - TSnew) < +Delta After the signature verification also successes, the RDnew and TSnew values SHOULD be stored in the cache as RDlast and TSlast. o When a message is received from a known peer (i.e., one that already has an entry in the cache), the timestamp is checked against the previously received Secure DHCPv6 message: TSnew + fuzz > TSlast + (RDnew - RDlast) x (1 - drift) - fuzz If this inequality does not hold or RDnew < RDlast, the recipient SHOULD silently discard the message. If, on the other hand, the inequality holds, the recipient SHOULD process the message. Moreover, if the above inequality holds and TSnew > TSlast, the recipient SHOULD update RDlast and TSlast after the signature verification also successes. Otherwise, the recipient MUST NOT update RDlast or TSlast. An implementation MAY use some mechanism such as a timestamp cache to strengthen resistance to replay attacks. When there is a very large number of nodes on the same link, or when a cache filling attack is in progress, it is possible that the cache holding the most recent timestamp per sender will become full. In this case, the node MUST remove some entries from the cache or refuse some new requested Jiang, et al. Expires December 21, 2014 [Page 15] Internet-Draft SeDHCPv6 June 2014 entries. The specific policy as to which entries are preferred over others is left as an implementation decision. 7. Deployment Consideration This document defines two levels of authentication: full authentication based on certificate or pre-shared key verification and weaker authentication based on leap-of-faith (LoF). As a mechanism, both levels can be applied on servers and clients. Depending on the details of expected threats and other constraints, some cases may have limited applicability. This section discusses such details. 7.1. Authentication on a client For clients, DHCP authentication generally means authenticating the server (the sender of DHCP messages) and verifying message integrity. This is satisfied with full authentication. Due to the configuration overhead, however, full authentication may not always be feasible. It would still be viable in a controlled environment with skilled staff, such as a corporate intranet. If LoF is used, message integrity is provided but there is a chance for the client to incorrectly trust a malicious server at the beginning of the first session with the server (and therefore keep trusting it thereafter). But LoF guarantees the subsequent messages are sent by the same server that sent the public key, and therefore narrows the attack scope. This may make sense if the network can be reasonably considered secure and requesting pre-configuration is deemed to be infeasible. A small home network would be an example of such cases. For environments that are neither controlled nor really trustworthy, such as a network cafe, full authentication wouldn't be feasible due to configuration overhead, while pure LoF, i.e. silently trusting the server at the first time, would be too insecure. But some middleground might be justified, such as requiring human intervention at the point of LoF. 7.2. Authentication on a server As for authentication on a server, there are several different scenarios to consider, each of which has different applicability issues. A server may have to selectively serve a specific client or deny specific clients depending on the identify of the client. This will Jiang, et al. Expires December 21, 2014 [Page 16] Internet-Draft SeDHCPv6 June 2014 require full authentication, since if the server allows LoF any malicious user can pretend to be a new legitimate client. Also, the use of certification wouldn't be feasible in this case, since it's less likely for all such clients to have valid (and generally different) certificates. So the applicable case may be limited, but a controlled environment with skilled staff and a specifically expected set of clients such as a corporate intranet may still find it useful and viable. A server can prevent an attack on the DHCP session with an existing client from a malicious client, e.g., by sending a bogus Release message: the server would remember the original client's public key at the beginning of the DHCP session and authenticate subsequent messages (and their sender). Neither full authentication nor LoF is needed for this purpose, since the server does not have to trust the public key itself. So this can be generally used for any usage of DHCP. A server can prevent an attack by a malicious client that pretends to be a valid past client and tries to establish a new DHCP session (whether this is a real security threat may be a subject of debate, but this is probably at least annoying). This is similar to the first scenario, but full authentication may not necessarily be required; since the purpose is to confirm a returning client has the same identify as a valid past client, the server only has to remember the client's public key at the first time. So LoF can be used at the risk of allowing a malicious client to mount this attack before the initial session with a valid client. An uncontrolled, but reasonably reliable network like a home network may use this defense with LoF. 8. Security Considerations This document provides new security features to the DHCPv6 protocol. Using public key based security mechanism and its verification mechanism in DHCPv6 message exchanging provides the authentication and data integrity protection. Timestamp mechanism provides anti- replay function. The Secure DHCPv6 mechanism is based on the pre-condition that the recipient knows the public key of senders or the sender's certificate can be verified through a trust CA. It prevents DHCPv6 server spoofing. The clients may discard the DHCPv6 messages from unknown/ unverified servers, which may be fake servers; or may prefer DHCPv6 messages from known/verified servers over unsigned messages or messages from unknown/unverified servers. The pre-configuration operation also needs to be protected, which is out of scope. The deployment of PKI is also out of scope. Jiang, et al. Expires December 21, 2014 [Page 17] Internet-Draft SeDHCPv6 June 2014 However, when a DHCPv6 client first encounters a new public key or a new unverifiable certificate, it can make a leap of faith. If the DHCPv6 server that used that public key or unverifiable certificate is in fact legitimate, then all future communication with that DHCPv6 server can be protected by storing the public key or unverifiable certificate. This does not provide complete security, but it limits the opportunity to mount an attack on a specific DHCPv6 client to the first time it communicates with a new DHCPv6 server. The number of cached public keys or unverifiable certificates MUST be limited in order to protect the DHCPv6 server against resource exhaustion attacks. Downgrade attacks cannot be avoided if nodes are configured to accept both secured and unsecured messages. A future specification may provide a mechanism on how to treat unsecured DHCPv6 messages. [RFC6273] has analyzed possible threats to the hash algorithms used in SEND. Since the Secure DHCPv6 defined in this document uses the same hash algorithms in similar way to SEND, analysis results could be applied as well: current attacks on hash functions do not constitute any practical threat to the digital signatures used in the signature algorithm in the Secure DHCPv6. A window of vulnerability for replay attacks exists until the timestamp expires. Secure DHCPv6 nodes are protected against replay attacks as long as they cache the state created by the message containing the timestamp. The cached state allows the node to protect itself against replayed messages. However, once the node flushes the state for whatever reason, an attacker can re-create the state by replaying an old message while the timestamp is still valid. In addition, the effectiveness of timestamps is largely dependent upon the accuracy of synchronization between communicating nodes. However, how the two communicating nodes can be synchronized is out of scope of this work. Attacks against time synchronization protocols such as NTP [RFC5905] may cause Secure DHCPv6 nodes to have an incorrect timestamp value. This can be used to launch replay attacks, even outside the normal window of vulnerability. To protect against these attacks, it is recommended that Secure DHCPv6 nodes keep independently maintained clocks or apply suitable security measures for the time synchronization protocols. 9. IANA Considerations This document defines three new DHCPv6 [RFC3315] options. The IANA is requested to assign values for these three options from the DHCPv6 Option Codes table of the DHCPv6 Parameters registry maintained in Jiang, et al. Expires December 21, 2014 [Page 18] Internet-Draft SeDHCPv6 June 2014 http://www.iana.org/assignments/dhcpv6-parameters. The three options are: The Public Key Option (TBA1), described in Section 5.1. The Certificate Option (TBA2), described in Section 5.2. The Signature Option (TBA3), described in Section 5.3. The IANA is also requested to add two new registry tables to the DHCPv6 Parameters registry maintained in http://www.iana.org/assignments/dhcpv6-parameters. The two tables are the Hash Algorithm for Secure DHCPv6 table and the Signature Algorithm for Secure DHCPv6 table. Initial values for these registries are given below. Future assignments are to be made through Standards Action [RFC5226]. Assignments for each registry consist of a name, a value and a RFC number where the registry is defined. Hash Algorithm for Secure DHCPv6. The values in this table are 8-bit unsigned integers. The following initial values are assigned for Hash Algorithm for Secure DHCPv6 in this document: Name | Value | RFCs -------------------+---------+-------------- SHA-1 | 0x01 | this document SHA-256 | 0x02 | this document SHA-512 | 0x03 | this document Signature Algorithm for Secure DHCPv6. The values in this table are 8-bit unsigned integers. The following initial values are assigned for Signature Algorithm for Secure DHCPv6 in this document: Name | Value | RFCs -------------------+---------+-------------- RSASSA-PKCS1-v1_5 | 0x01 | this document IANA is requested to assign the following new DHCPv6 Status Codes, defined in Section 5.4, in the DHCPv6 Parameters registry maintained in http://www.iana.org/assignments/dhcpv6-parameters: Jiang, et al. Expires December 21, 2014 [Page 19] Internet-Draft SeDHCPv6 June 2014 Code | Name | Reference ---------+-----------------------+-------------- TBD4 | AlgorithmNotSupported | this document TBD5 | AuthFailNotSupportLoF | this document TBD6 | AuthFailSupportLoF | this document TBD7 | TimestampFail | this document TBD8 | SignatureFail | this document 10. Acknowledgments The authors would like to thank Bernie Volz, Ted Lemon, Ralph Droms, Jari Arkko, Sean Turner, Stephen Kent, Thomas Huth, David Schumacher, Francis Dupont, Tomek Mrugalski, Gang Chen, Qi Sun, Suresh Krishnan, Tatuya Jinmei and other members of the IETF DHC working groups for their valuable comments. This document was produced using the xml2rfc tool [RFC2629]. 11. Change log [RFC Editor: Please remove] draft-ietf-dhc-sedhcpv6-03: addressed comments from WGLC. Added a new section "Deployment Consideration". Corrected the Public Key Field in the Public Key Option. Added considation for large DHCPv6 message transmission. Added TimestampFail error code. Refined the retransmission rules. 2014-06-18. draft-ietf-dhc-sedhcpv6-02: addressed comments (applicability statement, redesign the error codes and their logic) from IETF89 DHC WG meeting and volunteer reviewers. 2014-04-14. draft-ietf-dhc-sedhcpv6-01: addressed comments from IETF88 DHC WG meeting. Moved Dacheng Zhang from acknowledgement to be co-author. 2014-02-14. draft-ietf-dhc-sedhcpv6-00: adopted by DHC WG. 2013-11-19. draft-jiang-dhc-sedhcpv6-02: removed protection between relay agent and server due to complexity, following the comments from Ted Lemon, Bernie Volz. 2013-10-16. draft-jiang-dhc-sedhcpv6-01: update according to review comments from Ted Lemon, Bernie Volz, Ralph Droms. Separated Public Key/ Certificate option into two options. Refined many detailed processes. 2013-10-08. draft-jiang-dhc-sedhcpv6-00: original version, this draft is a replacement of draft-ietf-dhc-secure-dhcpv6, which reached IESG and dead because of consideration regarding to CGA. The authors followed Jiang, et al. Expires December 21, 2014 [Page 20] Internet-Draft SeDHCPv6 June 2014 the suggestion from IESG making a general public key based mechanism. 2013-06-29. 12. References 12.1. Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 (IPv6) Specification", RFC 2460, December 1998. [RFC3279] Bassham, L., Polk, W., and R. Housley, "Algorithms and Identifiers for the Internet X.509 Public Key Infrastructure Certificate and Certificate Revocation List (CRL) Profile", RFC 3279, April 2002. [RFC3315] Droms, R., Bound, J., Volz, B., Lemon, T., Perkins, C., and M. Carney, "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)", RFC 3315, July 2003. [RFC4055] Schaad, J., Kaliski, B., and R. Housley, "Additional Algorithms and Identifiers for RSA Cryptography for use in the Internet X.509 Public Key Infrastructure Certificate and Certificate Revocation List (CRL) Profile", RFC 4055, June 2005. [RFC4443] Conta, A., Deering, S., and M. Gupta, "Internet Control Message Protocol (ICMPv6) for the Internet Protocol Version 6 (IPv6) Specification", RFC 4443, March 2006. [RFC4491] Leontiev, S. and D. Shefanovski, "Using the GOST R 34.10-94, GOST R 34.10-2001, and GOST R 34.11-94 Algorithms with the Internet X.509 Public Key Infrastructure Certificate and CRL Profile", RFC 4491, May 2006. [RFC5280] Cooper, D., Santesson, S., Farrell, S., Boeyen, S., Housley, R., and W. Polk, "Internet X.509 Public Key Infrastructure Certificate and Certificate Revocation List (CRL) Profile", RFC 5280, May 2008. [RFC5905] Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network Time Protocol Version 4: Protocol and Algorithms Specification", RFC 5905, June 2010. Jiang, et al. Expires December 21, 2014 [Page 21] Internet-Draft SeDHCPv6 June 2014 [RFC5996] Kaufman, C., Hoffman, P., Nir, Y., and P. Eronen, "Internet Key Exchange Protocol Version 2 (IKEv2)", RFC 5996, September 2010. 12.2. Informative References [RFC2629] Rose, M., "Writing I-Ds and RFCs using XML", RFC 2629, June 1999. [RFC4270] Hoffman, P. and B. Schneier, "Attacks on Cryptographic Hashes in Internet Protocols", RFC 4270, November 2005. [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an IANA Considerations Section in RFCs", BCP 26, RFC 5226, May 2008. [RFC6273] Kukec, A., Krishnan, S., and S. Jiang, "The Secure Neighbor Discovery (SEND) Hash Threat Analysis", RFC 6273, June 2011. [RSA] RSA Laboratories, "RSA Encryption Standard, Version 2.1, PKCS 1", November 2002. Authors' Addresses Sheng Jiang (editor) Huawei Technologies Co., Ltd Q14, Huawei Campus, No.156 Beiqing Road Hai-Dian District, Beijing, 100095 P.R. China Email: jiangsheng@huawei.com Sean Shen CNNIC 4, South 4th Street, Zhongguancun Beijing 100190 P.R. China Email: shenshuo@cnnic.cn Jiang, et al. Expires December 21, 2014 [Page 22] Internet-Draft SeDHCPv6 June 2014 Dacheng Zhang Huawei Technologies Co., Ltd Q14, Huawei Campus, No.156 Beiqing Road Hai-Dian District, Beijing, 100095 P.R. China Email: zhangdacheng@huawei.com Tatuya Jinmei WIDE Project Japan Email: jinmei@wide.ad.jp Jiang, et al. Expires December 21, 2014 [Page 23]