Hi Med,

NEWER (this is what I’m asking RFC Editor to do for the suite of client-server 
drafts)

This section is modeled after the template described in Section 3.7 of 
[RFCAAAA].

The "<module-name>" YANG module defines a data model that is designed to be 
accessed via YANG-based management protocols, such as NETCONF [RFC6241] and 
RESTCONF [RFC8040].  These protocols have mandatory-to-implement secure 
transport layers (e.g., SSH [RFC4252], TLS [RFC8446], QUIC [RFC9000]) and 
mandatory-to-implement mutual authentication.

Kent // contributor



> On Sep 4, 2024, at 10:28 AM, Kent Watsen <[email protected]> wrote:
> 
> Hi Med,
> 
> A number of client-server drafts (in the NETCONF WG) are in AUTH48.   One 
> AD-level issue raised on all the drafts regards the Security Considerations 
> section not following the template in RFC 8407 or the template in rfc8407bis. 
>  You can see the discussion below.
> 
> My suggestion follows.
> 
> OLD:
> This section uses the template described in Section 3.7 of [RFCAAAA].
> 
> The YANG module specified in this document defines a schema for data
> that is designed to be accessed via network management protocols such
> as NETCONF [RFC6241] or RESTCONF [RFC8040]. These network management
> protocols are required to use a secure transport layer and mutual
> authentication, e.g., SSH [RFC6242] without the "none" authentication
> option, Transport Layer Security (TLS) [RFC8446] with mutual X.509
> authentication, and HTTPS with HTTP authentication (Section 11 of
> [RFC9110]).
> 
> 
> NEW:
> This section is modeled after the template described in Section 3.7 of 
> [RFCAAAA].
> 
> The <module-name> YANG module defines "grouping” and “container” statements 
> that are designed to be accessed via YANG-based management protocols, such as 
> NETCONF [RFC6241] and RESTCONF [RFC8040].  These protocols have 
> mandatory-to-implement secure transport layers (e.g., SSH [RFC4252], TLS 
> [RFC8446], QUIC [RFC9000]) and mandatory-to-implement mutual authentication.
> 
> 
> ALSO NEW?  (Add after the above paragraph?)
> 
> For the NETCONF protocol [RFC6241] , transport mapping documents include:
>   - Using the NETCONF Protocol over Secure Shell (SSH) [RFC6242]
>   - Using the NETCONF Protocol over Transport Layer Security (TLS) with 
> Mutual X.509 Authentication [RFC 7589]
>   - Using NETCONF over QUIC connection [NETCONF-QUIC]
> 
> For the RESTCONF protocol [RFC8040], the mandatory-to-implement transport is 
> HTTPS, as defined in:
>   - HTTP/1.1 [RFC9112]
>   - HTTP/2 [RFC9112]
>   - HTTP/3 [RFC9112]
> 
> 
> Thoughts?
> 
> 
> Kent  // contributor
> 
> 
> 
>> Begin forwarded message:
>> 
>> From: Mahesh Jethanandani <[email protected]>
>> Subject: Re: AD - Re: AUTH48: RFC-to-be 9644 
>> <draft-ietf-netconf-ssh-client-server-40> for your review
>> Date: September 3, 2024 at 5:41:58 PM EDT
>> To: Kent Watsen <[email protected]>, Alanna Paloma <[email protected]>
>> Cc: Alice Russo <[email protected]>, [email protected], NETCONF WG Chairs 
>> <[email protected]>, Per Andersson <[email protected]>, RFC Editor 
>> <[email protected]>, auth48archive <[email protected]>
>> Resent-From: <[email protected]>
>> Resent-To: [email protected], [email protected], [email protected]
>> 
>> Hi Kent,
>> 
>>> On Sep 3, 2024, at 12:18 PM, Kent Watsen <[email protected] 
>>> <mailto:[email protected]>> wrote:
>>> 
>>> Hi Mahesh,
>>> 
>>>> On Aug 29, 2024, at 6:27 PM, Mahesh Jethanandani <[email protected] 
>>>> <mailto:[email protected]>> wrote:
>>>> 
>>>> Hi Kent,
>>>> 
>>>>> On Aug 29, 2024, at 1:59 PM, Kent Watsen <[email protected] 
>>>>> <mailto:[email protected]>> wrote:
>>>>> 
>>>>> 
>>>>> Hi Mahesh (and Alice/Alanna)
>>>>> 
>>>>> 
>>>>>>>> 10) <!--[rfced] *AD - We note that the text in Sections 5.1 - 5.7 does 
>>>>>>>> not 
>>>>>>>> exactly match the YANG security considerations boilerplate text at
>>>>>>>> <https://wiki.ietf.org/group/ops/yang-security-guidelines>, including
>>>>>>>> missing citations to RFC 8446. Please review and let us know if the 
>>>>>>>> text
>>>>>>>> requires any updates.
>>>>>>>> -->
>>>>>> 
>>>>>> We have two options. There is the YANG security guidelines template that 
>>>>>> Alice points to, and then there is rfc8407bis template, that has been 
>>>>>> fairly stable for sometime. Even if you do not want to use the latter, 
>>>>>> because it is still a draft, we should adopt the guidelines that Alice 
>>>>>> points to. Do you disagree?
>>>>> 
>>>>> Yes, I sort of do disagree.
>>>>> 
>>>>> 1) They are “guidelines”, which suggests variability MAY occur.
>>>> 
>>>> I will agree that they are guidelines, and as such can have variability in 
>>>> how they are used, specially as it applies to modules that are only 
>>>> defining enumerations or identities. However, comparing the guidelines in 
>>>> rfc8407bis to what is in the draft for ietf-ssh-server and ietf-ssh-client 
>>>> modules, I notice that
>>>> 
>>>> - There is no reference to RFC6242 for NETCONF over SSH in this draft.
>>>> - There is no reference to RFC8446 for RESTCONF over TLS in this draft.
>>>> 
>>>> Was there a reason to not have those references?
>>> 
>>> A few comments:
>>>   - RFC8446 is TLS 1.3 (not RC over TLS)
>>>   - In rfc8407bis, I think RFC6242 was supposed to be RFC4252 (for SSH 
>>> transport).  That is, it’s not suppose to be “NC over SSH”.
>>>   - Also in rfc8407bis, I don’t think the HTTPS w/ authentication part is 
>>> critical.
>>>   - Also in rfc8407bis, it would be great to add QUIC to the list of 
>>> transports…
>>> 
>>> Following addresses the above, and makes the existing text more like what 
>>> rfc8407bis should be (IMO).  Can we can work with Med to update the text in 
>>> rfc8407bis to match the following?
>>> 
>>> OLD:
>>> The "ietf-ssh-client" YANG module defines "grouping" statements that are 
>>> designed to be accessed via YANG-based management protocols, such as 
>>> NETCONF [RFC6241] and RESTCONF [RFC8040]. Both of these protocols have 
>>> mandatory-to-implement secure transport layers (e.g., SSH, TLS) with mutual 
>>> authentication.
>>> 
>>> NEW:
>>> The "ietf-ssh-client" YANG module defines "grouping" statements that are 
>>> designed to be accessed via YANG-based management protocols, such as 
>>> NETCONF [RFC6241] and RESTCONF [RFC8040].  These protocols have 
>>> mandatory-to-implement secure transport layers (e.g., SSH [RFC4252], TLS 
>>> [RFC8446], QUIC [RFC9000]) and mandatory-to-implement mutual authentication.
>>> 
>>> A similar update would be in the "ietf-ssh-server" YANG module section too.
>>> 
>>> Background: from a Security-analysis perspective, what is important to know 
>>> is that a secure transport and mutual authentication is used.  That’s why 
>>> the text above focuses on just saying that.  
>> 
>> Thanks for the clarification. 
>> 
>> Alanna, I am good with the updated text.
>> 
>>> 
>>> If we want rfc8407bis to enumerate all known NC/RC-to-transport RFCs, I 
>>> suggest putting that text into its own paragraph, following the NEW 
>>> paragraph above.  We can work with Med to update rfc8407bis to do this also 
>>> - yes?
>> 
>> Please do. Do you want to spawn a separate thread with Med on the netmod 
>> mailing list?
>> 
>> Thanks.
>> 
>>> 
>>> 
>>> 
>>>>> 2) They contain incorrect/outdated statements, which SHOULD be fixed.
>>>>> 3) the updates to the rfc8407bis template are, by and large, due to my 
>>>>> attempts to fix the above.
>>>> 
>>>> Has this been brought up with the authors of rfc8407bis? If the statements 
>>>> are indeed outdated or incorrect, they should be addressed before the WG 
>>>> decides to publish rfc8407bis.
>>> 
>>> Yes, the text currently in rfc8407bis is, in large part, my attempt to fix 
>>> the text in RFC 8407.   
>>> 
>>> I do suggest we work with Med to tweak this section accordingly.
>>> 
>>> 
>>> 
>>> 
>>>>> IMO, the biggest issue is that the current text says that it "follows the 
>>>>> template defined in 
>>>>> https://datatracker.ietf.org/doc/html/rfc8407#section-3.7.1”.  But it 
>>>>> doesn’t follow that template.
>>>>> 
>>>>> Option-1: Simply tweak that sentence
>>>>> 
>>>>> OLD: This section follows the template defined in...
>>>>> NEW: This section is modeled after the template defined in...
>>>> 
>>>> I am ok with this option.
>>> 
>>> 
>>> Thanks
>>> 
>>> Kent // author
>>> 
>>> 
>>>> 
>>>> Thanks.
>>>> 
>>>>> 
>>>>> Option-2: Rewrite to match old/legacy/wrong RFC 8407 template
>>>>>   - honestly, I’m unsure if it’s possible (due to there being only 
>>>>> groupings)
>>>>> 
>>>>> Option-2: Rewrite to match new/better rfc8407bis template, and point to 
>>>>> it instead.
>>>>>   - It’s an Informative reference, so no MISREF would occur.
>>>>> 
>>>>> 
>>>>> Kent  // as author, and progenitor of Section 3.7.1 in RFC 8407
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>> 
>>>> 
>>>> Mahesh Jethanandani
>>>> [email protected] <mailto:[email protected]>
>> 
>> Mahesh Jethanandani
>> [email protected] <mailto:[email protected]>
>> 
>> 
>> 
>> 
>> 
>> 
> 
> _______________________________________________
> netmod mailing list -- [email protected]
> To unsubscribe send an email to [email protected]

_______________________________________________
netmod mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to