Re: AD evaluation comments on draft-ietf-nsis-rmd-15

Georgios,
Hannes,
 
Once again, the QSPEC specifies in Section 3.1.1. (http://tools.ietf.org/html/draft-ietf-nsis-qspec-24#section-3.3.1) 
"   The <Traffic Model> (TMOD) parameter is mandatory for the QNI to
   include in the initiator QSPEC and mandatory for downstream QNEs to
   interpret. "
and
"   All 5 of the sub-parameters MUST be included in the TMOD parameter.
   The TMOD parameter can be set to describe the traffic source.  If,
   for example, TMOD is set to specify bandwidth only, then set r = peak
   rate = p, b = large, m = large. "
 
Hence the RMD-QOSM MUST send the TMOD-1 parameter in the initiator QSPEC and downstream QNEs MUST interpret TMOD-1.  As per the second statement, the way to use TMOD-1 to specify "bandwidth" is by setting sub-parameters as follows:
 
r = peak rate
p = peak rate
b = large
m = large
 
Since the TMOD-1 parameter must be used, there is no point in defining a separate <bandwidth> parameter; it will be redundant with TMOD-1.
 
Thanks,
Jerry


--- On Sat, 2/27/10, Georgios Karagiannis <[email protected]> wrote:


From: Georgios Karagiannis <[email protected]>
Subject: Re: [NSIS] AD evaluation comments on draft-ietf-nsis-rmd-15
To: "Gerald Ash" <[email protected]>
Cc: "[email protected]" <[email protected]>, "'NSIS'" <[email protected]>
Date: Saturday, February 27, 2010, 4:13 PM


Hi Hannes,

Thank you very much!
I think that it is good to define the Bandwidth parameter in the same way
as in RFC 5624, which also defined the TMOD-1 AVP and the Bandwidth AVP,
where the description of the Bandwdith AVP, see RFC 5624, is:

3.3.  Bandwidth AVP (from RFC 5624)

   The Bandwidth AVP (AVP Code 502) is of type Float32 and is measured
   in octets of IP datagrams per second.  The Bandwidth AVP represents a
   simplified description of the following TMOD setting whereby the
   token rate (r) = peak traffic rate (p), the bucket depth (b) = large,
   and the minimum policed unit (m) = large when only bandwidth has to
   be expressed.

My new proposal is to define the Bandwidth parameter in RMD-QOSM in a
very similar way as defined in RFC 5624 and as it was defined since 5
years now in:
http://www.ietf.org/id/draft-ietf-nsis-rmd-15.txt:

---------------

   The <Bandwidth> parameter (ID=17) is of type Float32 and is measured
   in octets of IP datagrams per second.  The <Bandwidth> parameter
   represents a simplified description of the following <TMOD-1> parameter
   setting, specified in [QSP-T], whereby the token rate (r) = peak
traffic
   rate (p), the bucket depth (b) = large, and the minimum policed unit
   (m)  = large when only bandwidth has to be expressed.

   The bit format used for the <Bandwidth> parameter is shown below and it
   is a translation of the <TMOD-1> parameter, see [QSP-T], to specify
   bandwidth only, by setting, see [QSP-T]:
    o  TMOD Rate-1 (r)= Peak Data Rate-1 (p)
    o  TMOD Size-1 (b) = large
    o  Minimum Policed Unit-1 (m) = large


     0                   1                   2                   3
     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    |1|E|0|r|       17              |r|r|r|r|          1            |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    | Peak Data Rate-1 (p)  (32-bit IEEE floating point number)     |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+


--------------------------

Best regards,
Georgios


On 2/27/2010, "Georgios Karagiannis" <[email protected]> wrote:

>Hi Jerry
>
>See below!
>
>On 2/27/2010, "Gerald Ash" <[email protected]> wrote:
>
>>Georgios,
>> 
>>See below.
>> 
>>> The bandwidth parameter is just the translation of the T-MOD parameter to
>>> the bandwidth paramter. I will include the following information in the
>>> draft:
>>>
>>>   The bit format used for the <Bandwidth> parameter is shown below and
>>>   it is a translation of the <TMOD-1> parameter, see [QSP-T], to
>>>   specify bandwidth only, by setting, see [QSP-T]:
>>>     o  TMOD Rate-1 (r)= Peak Data Rate-1 (p)
>>>     o  TMOD Size-1 (b) = large
>>>     o  Minimum Policed Unit-1 (m) = large
>>>
>>>    0                   1                   2                   3
>>>    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |1|E|0|r|  Bandwidth ID         |r|r|r|r|          5            |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |  Peak Data Rate-1 (p) (32-bit IEEE floating point number)     |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |  TMOD Size-1 (b) (32-bit IEEE floating point number)          |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |  Peak Data Rate-1 (p) (32-bit IEEE floating point number)     |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |  Minimum Policed Unit-1 (m) (32-bit unsigned integer)         |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |  Maximum Packet Size-1 (MPS) (32-bit unsigned integer)        |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>
>>Just to be clear, the "Bandwidth ID" = 1, which is the same value as the TMOD-1 parameter ID (= 1), as already assigned by IANA.  It is not necessary to define another QSPEC parameter ID.  What you are calling the "<bandwidth> parameter" is actually the TMOD-1 parameter with particular values.  Once again, it is required that all NSIS/NSLP/QOSM implementations send the TMOD-1 parameter in their QSPEC's. 
>
>The Bandwidth ID = 17.
>
>This parameter is different then the definition of TMOD, since it is
>specifying bandwidth only, see the particular paragraph in the QSPEC
>draft:
>
>   "The TMOD parameter can be set to describe the traffic source.  If,
>   for example, TMOD is set to specify bandwidth only, then set r = peak
>   rate = p, b = large, m = large.  As another example if TMOD is set
>   for TCP traffic, then set r = average rate, b = large, p = large."
>
>The bit format of Bandwidth is representing the semantics of the above
>paragraph. As you can see here the difference in the bit format is that
>the  TMOD Rate-1 (r) field is replaced by the  Peak Data Rate-1 (p) field.
>
>
>0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |1|E|0|r|  Bandwidth ID         |r|r|r|r|          5            |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |  Peak Data Rate-1 (p) (32-bit IEEE floating point number)     |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |  TMOD Size-1 (b) (32-bit IEEE floating point number)          |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |  Peak Data Rate-1 (p) (32-bit IEEE floating point number)     |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |  Minimum Policed Unit-1 (m) (32-bit unsigned integer)         |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>>>   |  Maximum Packet Size-1 (MPS) (32-bit unsigned integer)        |
>>>   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>
>
>Best regards,
>Georgios
>
>>
>>> Regarding the "large" values, I will specify them in Section 4.6.
>> 
>>OK.
>> 
>>Thanks,
>>Jerry
>>
>>
>>      )

_______________________________________________
nsis mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/nsis