Re: [Dots] Alissa Cooper's Discuss on draft-ietf-dots-signal-channel-31: (with DISCUSS and COMMENT)

[Alissa Cooper <[e-mail]>]

Hi Med,

> On May 2, 2019, at 3:18 AM, [e-mail] wrote:
> 
> Hi Alissa, 
> 
> Please see inline.
> 
> Cheers,
> Med
> 
>> -----Message d'origine-----
>> De : Alissa Cooper via Datatracker [mailto:[e-mail] <mailto:[e-mail]>]
>> Envoyé : jeudi 2 mai 2019 03:56
>> À : The IESG
>> Cc : [e-mail] <mailto:[e-mail]>; Liang Xia; dots-
>> [e-mail] <mailto:[e-mail]>; [e-mail] <mailto:[e-mail]>; [e-mail] <mailto:[e-mail]>
>> Objet : Alissa Cooper's Discuss on draft-ietf-dots-signal-channel-31: (with
>> DISCUSS and COMMENT)
>> 
>> Alissa Cooper has entered the following ballot position for
>> draft-ietf-dots-signal-channel-31: Discuss
>> 
>> When responding, please keep the subject line intact and reply to all
>> email addresses included in the To and CC lines. (Feel free to cut this
>> introductory paragraph, however.)
>> 
>> 
>> Please refer to https://www.ietf.org/iesg/statement/discuss-criteria.html
>> for more information about IESG DISCUSS and COMMENT positions.
>> 
>> 
>> The document, along with other ballot positions, can be found here:
>> https://datatracker.ietf.org/doc/draft-ietf-dots-signal-channel/
>> 
>> 
>> 
>> ----------------------------------------------------------------------
>> DISCUSS:
>> ----------------------------------------------------------------------
>> 
>> = Section 3 =
>> 
>> "By default, a DOTS signal channel MUST run over port number TBD as
>>   defined in Section 9.1, for both UDP and TCP, unless the DOTS server
>>   has a mutual agreement with its DOTS clients to use a different port
>>   number.  DOTS clients MAY alternatively support means to dynamically
>>   discover the ports used by their DOTS servers (e.g.,
>>   [I-D.boucadair-dots-server-discovery])."
>> 
>> MUST implies an absolute requirement, so "MUST ... unless" is a problematic
>> construction.
> 
> [Med] It seems that you missed "By default, “. 

Even with “by default” this still is problematic. MUST indicates an absolute requirement.

> 
> 
> Furthermore, it doesn't make sense together with "MAY
>> alternatively," which indicates that port number discovery is an alternative
>> to
>> the fixed to-be-assigned port.
>> 
>> I didn't have time to get very far into draft-boucadair-dots-server-
>> discovery,
> 
> [Med] I updated that reference to I-D.ietf-dots-server-discovery. 
> 
>> but it appears that it does not mandate support for any single discovery
>> mechanism for clients and servers to support. If so, that "alternatively"
>> seems
>> like more of a problem, since it allows for there to be no interoperable
>> mechanism for clients to discover server ports. I think maybe what was
>> intended
>> here was:
>> 
>> s/MUST/SHOULD/
>> s/MAY alternatively/MAY additionally/
> 
> [Med] I implemented the second change. 
> 
>> 
>> = Section 4.4.1 =
>> 
>> (1)
>> "In deployments where server-domain DOTS gateways are enabled,
>>   identity information about the origin source client domain SHOULD be
>>   propagated to the DOTS server.  That information is meant to assist
>>   the DOTS server to enforce some policies such as grouping DOTS
>>   clients that belong to the same DOTS domain, limiting the number of
>>   DOTS requests, and identifying the mitigation scope.  These policies
>>   can be enforced per-client, per-client domain, or both.  Also, the
>>   identity information may be used for auditing and debugging purposes."
>> 
>> Does "identity information" just refer to cdid, or something else?
> 
> [Med] It refers to the information conveyed in cdid. 

I think it would be helpful to clarify that.

> 
>> 
>> (2) The constructions "MUST ... (absent explicit policy/configuration
>> otherwise)" are problematic. I'm assuming these are meant to be SHOULDs.
>> 
> 
> [Med] I checked this wording with Ben.  

Ok, perhaps he can comment then.

> 
>> = Section 13.1 =
>> 
>> I don't understand why RFC 7951 is a normative reference but
>> draft-ietf-core-yang-cbor is an informative reference.
> 
> [Med] We used to have both as informative references, but unless I'm mistaken 7951 was moved to normative so that at least one method is supported.

This is being discussed in another thread, but if that is the case the normative requirement text needs to change too.

Thanks,
Alissa

> 
>> 
>> 
>> ----------------------------------------------------------------------
>> COMMENT:
>> ----------------------------------------------------------------------
>> 
>> = Section 4.4.1 =
>> 
>> "The 'cuid' is intended to be stable when communicating with a
>>      given DOTS server, i.e., the 'cuid' used by a DOTS client SHOULD
>>      NOT change over time. "
>> 
>> Why is this the recommended behavior?
>> 
> 
> [Med] because all resources/state of a DOTS client are bound to this identifier.