RFC Errata


Errata Search

 
Source of RFC  
Summary Table Full Records

Found 3 records.

Status: Verified (2)

RFC 6546, "Transport of Real-time Inter-network Defense (RID) Messages over HTTP/TLS", April 2012

Source of RFC: mile (sec)

Errata ID: 3267
Status: Verified
Type: Technical
Publication Format(s) : TEXT

Reported By: John Field
Date Reported: 2012-06-26
Verifier Name: Sean Turner
Date Verified: 2012-06-28

Section 3 says:

In paragraph 4 of section 3, fourth sentence:

   As RID messages MUST be
   sent using the POST method, the GET and HEAD methods have no
   particular meaning on a RID system; a RID system SHOULD answer
   'GET /' or 'HEAD /' with 204 No Content.

It should say:

Consistent with RFC 2616 section 10.4.6, a RID system MUST answer 
any HTTP request to Request-URI of '/' which uses an HTTP method 
other than 'POST' by producing an HTTP response with a status code 
of 405 Method Not Allowed.  The RID system HTTP response MUST also 
include an Allow header indicating that only the 'POST' method is 
supported.

Notes:

There has been a brief discussion of this errata on the MILE list, with the first message in the thread having been posted on June 5, 2012.

The corrected text that I have suggested above has been written as narrowly as possible, and remains consistent with the original functionality described in 6546.

Lacking support for 'GET' means that there is no way to verify if a RID endpoint is active, other than by doing a real request, i.e. a Report, or Query, etc. Thus, one might also consider supporting HEAD, e.g. for RID testing purposes, though that option has not been discussed yet. Note, however, that supporting HEAD potentially raises further issues since according to RFC 2616 the response headers to a HEAD request SHOULD be consistent with a GET, which is specifically not supported.

Errata ID: 3455
Status: Verified
Type: Technical
Publication Format(s) : TEXT

Reported By: John Field
Date Reported: 2013-01-14
Verifier Name: Sean Turner
Date Verified: 2013-03-16

Section 3 says:

If a RID system receives an improper RID message in an HTTP Request,
it MUST return an appropriate 4xx Client Error result code to the 
requesting RID system.

It should say:

If a RID system receives an improper HTTP Request, it MUST return 
an appropriate 4xx Client Error result code to the requesting RID 
system.

Notes:

There has been some discussion of this issue on the MILE mailing list. Another possible option for the corrected text is to say nothing at all. That is, by changing the specification to focus on an improper HTTP request, rather than an improper RID message, the corrected text is simply a restatement of existing HTTP behavior. (Either way, this still does constitute a technical change since we would no longer be requiring the 400 status code when the error is with the *RID* content). On this technical point, we had consensus on the MILE mailing list: we SHOULD NOT require an HTTP 4xx status code when there is an error with the RID content itself (as opposed to the HTTP layer). HTTP 4xx status is reserved for errors occurring in the HTTP protocol layer. Errors in the RID content will be reported via the RID Acknowledgement message type, with appropriate choices for the RequestStatus element, and Justification attribute.

Status: Held for Document Update (1)

RFC 6546, "Transport of Real-time Inter-network Defense (RID) Messages over HTTP/TLS", April 2012

Source of RFC: mile (sec)

Errata ID: 3277
Status: Held for Document Update
Type: Editorial
Publication Format(s) : TEXT

Reported By: John Field
Date Reported: 2012-07-03
Held for Document Update by: Sean Turner

Section 3 says:

Table 1 lists the allowable RID message types in an HTTP Response for a given RID message type in the Request.  A RID system MUST be prepared to handle an HTTP Response of the given type(s) when sending the corresponding HTTP Request.  A RID system MUST NOT send an HTTP Response containing any RID message other than the one corresponding to the one sent in the HTTP Request.

(table 1 appears here)

The use of stable DNS names to address RID systems is RECOMMENDED; in addition to facilitating connection to RID systems within a consortium, these are to be used as reference identifiers for a RID system's peers.  For security purposes, RID systems SHOULD NOT return 3xx Redirection response codes, and SHOULD NOT follow any 3xx Redirection.  The protocol provides no in-band method for handling a change of address of a RID system.

It should say:

Insert new text just before table 1:

"An X appearing in the Callback column of Table 1 means that the exchange itself IS a callback.  In these cases the HTTP request contains a RID message that is intended to conclude an earlier RID exchange which initially returned 202.   Note that RID Acknowledgment and RID Result messages can only ever appear in an HTTP request when the message is being generated as a Callback.  However, a RID Report message that appears in an HTTP request may represent either a unsolicited Report, or a delayed Callback.  It is important to note that any RID message that is sent as a Callback must be answered with a 200, and so cannot itself generate yet another Callback."

Notes:

This is a request to insert some additional text to help clarify the meaning of the "X" that appears in the Callback column of table 1. I believe this will be of benefit to implementers who must understand the message exchange patterns described in table 1 in order to properly implement the RID protocol.

Report New Errata



Advanced Search