| Independent Submission | M. Bäuerle |
| Internet-Draft | STZ Elektronik |
| Intended status: Standards Track | March 10, 2017 |
| Expires: September 11, 2017 |
Cancel-Locks in Netnews articles
draft-baeuerle-netnews-cancel-lock-02
This document defines an extension to the Netnews Article Format that may be used to authenticate the cancelling and superseding of existing articles.
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 September 11, 2017.
Copyright (c) 2017 IETF Trust and the persons identified as the document authors. All rights reserved.
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.
This document may contain material from IETF Documents or IETF Contributions published or made publicly available before November 10, 2008. The person(s) controlling the copyright in some of this material may not have granted the IETF Trust the right to allow modifications of such material outside the IETF Standards Process. Without obtaining an adequate license from the person(s) controlling the copyright in such materials, this document may not be modified outside the IETF Standards Process, and derivative works of it may not be created outside the IETF Standards Process, except to format it for publication as an RFC or to translate it into languages other than English.
The authentication system defined in this document is intended to be used as a simple method to verify that the author of an article which cancels ([RFC5537] Section 5.3) or supersedes ([RFC5537] Section 5.4) another one is either the poster, posting agent, moderator or injecting agent that processed the original article when it was in its proto-article form.
One property of this system is that it prevents tracking of individual users.
There are other authentication systems available with different properties. When everybody should be able to verify who the originator is, e.g. for control messages to add or remove newsgroups ([RFC5537] Section 5.2), an OpenPGP [RFC4880] signature is suited.
Any term not defined in this document has the same meaning as it does in [RFC5536] or [RFC5537].
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].
Please write the letters "ae" in "Baeuerle" as an a-umlaut (U+00E4, "ä" in XML), the first letter in "Elie" with an acute accent (U+00C9, "É" in XML), the letters "ss" in Janssen as an eszett (U+00DF, "ß" in XML) and the letters "ue" in Baden-Wuerttemberg as an u-umlaut (U+00FC, "ü" in XML) wherever this is possible.
This section describes the formal syntax of the new header fields using ABNF [RFC5234][RFC7405]. It extends the syntax in Section 3 of [RFC5536] and non-terminals not defined in this document are defined there. The [RFC5536] ABNF should be imported first before attempting to validate these rules.
The new header fields Cancel-Lock and Cancel-Key are defined by this document, they follow the rules described in [RFC5536] Section 2.2:
fields =/ *( cancel-lock / cancel-key )
Each of these header fields MUST NOT occur more than once in an article.
Both new header fields contain lists of encoded values. Every entry is based on a <scheme>:
scheme = %s"sha-256" / 1*scheme-char / obs-scheme scheme-char = LOWER / DIGIT / "-" LOWER = %x61-7A ; lowercase characters [a-z]
The hash algorithms for <scheme> are defined in [SHA], see also [RFC1321] and [RFC6151] for MD5, [RFC3174] for SHA1 and [RFC6234] for the SHA2 family. The Base64 encoding used is defined in Section 6.8 of [RFC2045].
This document defines one value for <scheme>: "sha-256". This scheme is mandatory to implement.
Note that the obsolete syntax <obs-scheme> was defined case-insensitive. This is changed in this document and the scheme MUST now be generated with lowercase letters.
The case sensitivity of <scheme> is defined to simplify the checks.
cancel-lock = "Cancel-Lock:" SP c-lock-list CRLF c-lock-list = c-lock *(CFWS c-lock) [CFWS] c-lock = scheme ":" c-lock-string c-lock-string = *(4base64-char) [base64-terminal] base64-char = ALPHA / DIGIT / "+" / "/" base64-terminal = 2base64-char "==" / 3base64-char "="
If <scheme> is not supported by an implementation, the corresponding <c-lock> element MUST be skipped and potential following <c-lock> elements MUST NOT be ignored.
<c-lock-string> is the Base64 encoded output of a hash operation (defined by <scheme>) of the Base64 encoded key "K" that is intended to authenticate the person or agent that created or processed respectively the article up to injection (inclusively):
base64(hash(base64(K)))
Because of the one-way nature of the hash operation the key "K" is not revealed.
cancel-key = "Cancel-Key:" SP c-key-list CRLF c-key-list = c-lock *(CFWS c-lock) [CFWS] c-key = scheme ":" c-key-string c-key-string = 1*base64-octet base64-octet = ALPHA / DIGIT / "+" / "/" / "="
If <scheme> is not supported by an implementation, the corresponding <c-key> element MUST be skipped and potential following <c-key> elements MUST NOT be ignored.
<c-key-string> is the Base64 encoded key "K" that was used to create the Cancel-Lock header field (as defined in Section 2.1 of this document) of the original article:
base64(K)
The relaxed syntax definition of <c-key-string> above is required for backward compatibility. New implementations always SHOULD generate valid Base64.
A Cancel-Lock header field MAY be added to a proto-article by the poster or posting agent which will include one or more <c-lock> elements.
If the poster or posting agent doesn't add a Cancel-Lock header field to an article, then an injecting agent (or moderator) MAY add one or more provided that it positively authenticates the author. The injecting agent (or moderator) MUST NOT add this header field to an article unless it is able to authenticate all cancelling or superseding attempts from the poster and automatically add working Cancel-Key header fields for such articles.
Other agents MUST NOT add this header to articles or proto-articles that they process.
If a Cancel-Lock header field has already been added to a proto-article then any agent further processing the proto-article up to the injecting agent (inclusively) MAY append additional <c-lock> elements to those already in the header.
Use cases for extending the Cancel-Lock header field:
Once an article is injected then this header MUST NOT be altered. In particular, relaying agents beyond the injecting agent MUST NOT alter it.
A Cancel-Key header field MAY be added to a proto-article containing a Control or Supersedes header field by the poster or posting agent which will include one or more <c-key> elements. They will correspond to some or all of the <c-lock> elements in the article referenced by the Control (with a "cancel" command as defined in [RFC5537]) or Supersedes header field.
If, as mentioned in Section 3.2 an injecting agent (or moderator) has added a Cancel-Lock header field to an article listed in the Control (with "cancel" command as defined in [RFC5537]) or Supersedes header field then (given that it authenticates the poster as being the same as the poster of the original article) it MUST add (or extend, if already present) the Cancel-Key header field with at least one <c-key> element that correspond to that article.
Other agents MUST NOT alter this header.
When a serving agent receives an article that attempts to cancel or supersede a previous article via Control (with a "cancel" command as defined in [RFC5537]) or Supersedes header field, the system defined in this document can be used for authentication. The general handling of articles containing such attempts as defined in [RFC5537] is not changed by this document.
To process the authentication, the received article must contain a Cancel-Key header field and the original article a Cancel-Lock header field. If this is not the case, the authentication is not possible (failed).
For the authentication check, every supported <c-key> element from the received article is processed as follows:
This section is informative, not normative.
It is suggested to use the function K = HMAC(mid+sec) to create the key "K" for an article with Message-ID <mid>, where HMAC is outlined in [RFC2104]. <sec> is a secret held locally that can be used for multiple articles. This method removes the need for a per-article database containing the keys used for every article.
The local secret <sec> should have a length of at least the output size of the hash function that is used by HMAC (32 octets for SHA-256). If the secret is not a random value, but e.g. some sort of human readable password, it should be much longer. In any case it is important that this secret can not be guessed.
Note that the hash algorithm used as base for the HMAC operation is not required to be the same as specified by <scheme>. An agent that verifies a Cancel-Key header field simply checks whether it matches one of the <c-lock> elements with the same <scheme> in the Cancel-Lock header field of the original article.
Common libraries like OpenSSL can be used for the cryptographic operations.
Example data for creation of a <c-lock> element with HMAC-SHA256 (as suggested in Section 4):
Message-ID: <12345@mid.example>
mid: <12345@mid.example> sec: ExampleSecret K : HMAC-SHA256(mid+sec) ;"mid" used as HMAC message, "sec" used as HMAC key
Calculation of Base64(K) using the OpenSSL command line tools in a POSIX shell:
$ printf "%s" "<12345@mid.example>" \
| openssl dgst -sha256 -hmac "ExampleSecret" -binary \
| openssl enc -base64
qv1VXHYiCGjkX/N1nhfYKcAeUn8bCVhrWhoKuBSnpMA=
This can be used as <c-key-string> for canceling or superseding the article <12345@mid.example>.
Calculation of Base64(SHA256(Base64(K))) required for <c-lock-string> using the OpenSSL command line tools in a POSIX shell:
$ printf "%s" "qv1VXHYiCGjkX/N1nhfYKcAeUn8bCVhrWhoKuBSnpMA=" \
| openssl dgst -sha256 -binary \
| openssl enc -base64
s/pmK/3grrz++29ce2/mQydzJuc7iqHn1nqcJiQTPMc=
Inserted into the header of article <12345@mid.example> it looks like this:
Cancel-Lock: sha-256:s/pmK/3grrz++29ce2/mQydzJuc7iqHn1nqcJiQTPMc=
Inserted into the header of an article that should cancel or supersede article <12345@mid.example> it looks like this:
Cancel-Key: sha-256:qv1VXHYiCGjkX/N1nhfYKcAeUn8bCVhrWhoKuBSnpMA=
Other matching pair of Cancel-Lock and Cancel-Key header fields:
Cancel-Lock: sha-256:RrKLp7YCQc9T8HmgSbxwIDlnCDWsgy1awqtiDuhedRo= Cancel-Key: sha-256:sSkDke97Dh78/d+Diu1i3dQ2Fp/EMK3xE2GfEqZlvK8=
With obsolete syntax (requires case-insensitive parsing of <scheme> and uses a <c-key-string> with invalid/missing Base64 padding):
Cancel-Lock: sha1:bNXHc6ohSmeHaRHHW56BIWZJt+4= Cancel-Key: ShA1:aaaBBBcccDDDeeeFFF
Let's assume that all the examples above are associated to the same article (e.g. created by different agents):
Cancel-Lock: sha-256:s/pmK/3grrz++29ce2/mQydzJuc7iqHn1nqcJiQTPMc=
sha-256:RrKLp7YCQc9T8HmgSbxwIDlnCDWsgy1awqtiDuhedRo=
sha1:bNXHc6ohSmeHaRHHW56BIWZJt+4=
Cancel-Key: sha-256:qv1VXHYiCGjkX/N1nhfYKcAeUn8bCVhrWhoKuBSnpMA=
sha-256:sSkDke97Dh78/d+Diu1i3dQ2Fp/EMK3xE2GfEqZlvK8=
ShA1:aaaBBBcccDDDeeeFFF
Manual checks using the OpenSSL command line tools in a POSIX shell:
$ printf "%s" "qv1VXHYiCGjkX/N1nhfYKcAeUn8bCVhrWhoKuBSnpMA=" \
| openssl dgst -sha256 -binary \
| openssl enc -base64
s/pmK/3grrz++29ce2/mQydzJuc7iqHn1nqcJiQTPMc=
$ printf "%s" "sSkDke97Dh78/d+Diu1i3dQ2Fp/EMK3xE2GfEqZlvK8=" \
| openssl dgst -sha256 -binary \
| openssl enc -base64
RrKLp7YCQc9T8HmgSbxwIDlnCDWsgy1awqtiDuhedRo=
$ printf "%s" "aaaBBBcccDDDeeeFFF" \
| openssl dgst -sha1 -binary \
| openssl enc -base64
bNXHc6ohSmeHaRHHW56BIWZJt+4=
Implementations of earlier drafts of this specification allowed more liberal (case insensitive) syntax and defined a different value for <scheme> than this version. The following value for <scheme> is now deprecated and SHOULD NOT be generated anymore. Serving agents SHOULD still accept it for a transition period as long as the corresponding hash function is not considered unsafe. See Section 7 for details.
obs-scheme = "sha1"
<obs-scheme> MUST be parsed case-insensitive.
It is important for backward compatibility that the deprecated value for <scheme> is not phased out too early. Security and compatibility concerns should be carefully weighed before choosing to remove <obs-scheme> from existing implementations (or not implementing it in new ones).
The important properties of the hash function used for <scheme> are the preimage and second preimage resistance. A successful preimage attack would reveal the real Cancel-Key that was used to create the Cancel-Lock of the original article. A successful second preimage attack would allow to create a new, different Cancel-Key that matches a Cancel-Lock too. Both cases would break the authentication system defined in this document.
Collision resistance of the hash function used for <scheme> is less important. Finding two Cancel-Keys that match an arbitary Cancel-Lock is not helpful to break the authentication system defined in this document (if a specific article is defined as target). Only collateral damage like arbitrary deletion or spam is possible.
Currently there is no known practicable preimage and second preimage attack against the hash function SHA1. Therefore there is no hurry to replace it. The reasons why this document specifies SHA-256 (aka SHA2-256) are:
The operation HMAC(mid+sec) as suggested in Section 4 must be able to protect the local secret <sec>. The Message-ID <mid> is public (in the article header). An attacker who wants to steal/use a local secret only need to break this algorithm (regardless of <scheme>), because Cancel-Keys are explicitly published for every request to modify or delete existing articles.
Even if HMAC-MD5 and HMAC-SHA1 are not considered broken today, it is desired to have some more security margin here. Breaking <scheme> only allows to authenticate a single forged modify or delete request. With <sec> in hand it is possible to forge such requests for all articles that contain Cancel-Locks based on Cancel-Keys generated with this <sec> in the past.
IANA has registered the following header fields in the Permanent Message Header Field Repository, in accordance with the procedures set out in [RFC3864]:
The Netnews Cancel-Lock hash algorithm registry will be maintained by IANA.
The registry will be available at <https://www.iana.org/assignments/netnews-cancel-lock-parameters/>.
IANA will register new Cancel-Lock hash algorithm names on a First Come First Served basis, as defined in BCP 26 [RFC5226]. IANA has the right to reject obviously bogus registration requests, but will perform no review of claims made in the registration form.
Registration of a Netnews Cancel-Lock hash algorithm is requested by filling in the following template and sending it via electronic mail to IANA at <iana@iana.org>:
Authors may seek community review by posting a specification of their proposed algorithm as an Internet-Draft. Netnews Cancel-Lock hash algorithms intended for widespread use should be standardized through the normal IETF process, when appropriate.
This section gives a formal definition of the Netnews Cancel-Lock hash algorithms as required by Section 8.1 for the IANA registry.
| [RFC1321] | Rivest, R., "The MD5 Message-Digest Algorithm", RFC 1321, DOI 10.17487/RFC1321, April 1992. |
| [RFC2104] | Krawczyk, H., Bellare, M. and R. Canetti, "HMAC: Keyed-Hashing for Message Authentication", RFC 2104, DOI 10.17487/RFC2104, February 1997. |
| [RFC3174] | Eastlake 3rd, D. and P. Jones, "US Secure Hash Algorithm 1 (SHA1)", RFC 3174, DOI 10.17487/RFC3174, September 2001. |
| [RFC4880] | Callas, J., Donnerhacke, L., Finney, H., Shaw, D. and R. Thayer, "OpenPGP Message Format", RFC 4880, DOI 10.17487/RFC4880, November 2007. |
| [RFC6151] | Turner, S. and L. Chen, "Updated Security Considerations for the MD5 Message-Digest and the HMAC-MD5 Algorithms", RFC 6151, DOI 10.17487/RFC6151, March 2011. |
| [RFC6234] | Eastlake 3rd, D. and T. Hansen, "US Secure Hash Algorithms (SHA and SHA-based HMAC and HKDF)", RFC 6234, DOI 10.17487/RFC6234, May 2011. |
The author acknowledges the original author of the Cancel-Lock authentication system as documented in draft-ietf-usefor-cancel-lock: Simon Lyall. He has written the original draft and former version <https://tools.ietf.org/html/draft-ietf-usefor-cancel-lock-01> and approved the usage of his work for this document. This document is mostly based on his work and was originally intended as revision 02. It must be renamed because the USEFOR IETF WG is now closed.
The author would like to thank the following individuals for contributing their ideas and reviewing this specification: Julien Élie, Richard Kettlewell, Holger Marzen. And Peter Faust, Urs Janßen and Alfred Peters for providing statistic data about the algorithms currently in use.