Re: [tsvwg] draft diffserv-intercon: Handling of a scavenger class / CS1

[Brian E Carpenter <[e-mail]>]

Firstly, note that RFC 3662 is very intentionally not standards-track and
does not recommend a DSCP value. To me, this discussion seems to mix up two
orthogonal concepts:

1. What the PHB for a scavenger class actually is (or in Mikael's
suggestion to AQM, what the PHB set is, since mention of priority
implies several members of a PHB set, like AFxy).

2. What DSCP values happen to be used for this PHB set. Despite
the confusion introduced by RFC 3662, those values don't have to
ones already used for other purposes, especially not CSx. Actually
the text makes it clear that the DSCPs recommended for AFxy should
be considered:

   If a CS PHB is used, note that this configuration will violate the
   "SHOULD" of section 4.2.2.2 of RFC 2474 [RFC2474] since CS1 will have
   a less timely forwarding than CS0.  An operator's goal of providing
   an LE PDB is sufficient cause for violating the SHOULD.  If an AF PHB
   is used, it must be configured and a DSCP assigned such that it does
   not violate the "MUST" of paragraph three of section 2 of RFC 2597
   [RFC2597] which provides for a "minimum amount of forwarding
   resources".

     Brian

On 30/05/2015 00:53, Mikael Abrahamsson wrote:
> On Fri, 29 May 2015, [e-mail] wrote:
> 
>> Gorry,
>>
>> the notes taken in Dallas (and thanks to the note taker!) showed a scavenger-class-at-interconnection related discussion.
>>
>> Summary of Dallas discussion:
>> -       Scavenger is not supported at Interconnection, as CS1 is not a widely deployed standard DSCP for scavenger.
>> -       Taking this into account, there was a request for CS1 not to be re-marked to any other DSCP.
>>
>> Basic questions:
>> -       Is the issue to get Scavenger support in a general DiffServ sense at Interconnection, i.e. the PHB is supported by the
>> receiving domain but the DSCP may be remarked? To me, this would make Scavenger relevant for diffserv-intercon
>> -       Is the issue to avoid remarking of the DSCP, no matter whether the receiving domain supports Scavenger? As I agree to
>> remove the remarking discussion from diffserv-intrcon, I'd regard this to be out of its scope (in both cases, if the receiving
>> domain supports Scavenger by CS1 and also if it transports CS1 unchanged if it isn't aware of the DiffServ PHB indicated by it).
> 
>> From an operational point of view, for me, the only way to get scavenger 
> class incrementally implemented is to create a new DSCP bit pattern for this with the first 3 bits set to 0. CS1 is not well
> suited for scavanger for incremental deployment in todays networks because it means different things in different places.
> 
> I posted about this in another discussion here on AQM:
> 
> http://www.ietf.org/mail-archive/web/aqm/current/msg01240.html
> 
> The proposal is to create a scavanger class which would map to CS0 but have DSCP-style drop probability bits set.
>