Re: [Ioam] Internal WG Review: In-situ OAM (ioam)

["Mirja Kuehlewind (IETF)" <[e-mail]>]

Hi all,

as a side notice: marking is define as hybrid measurement in the already cited IPPM RFC 7799:

"Augmentation or modification of the stream of interest, or
      employment of methods that modify the treatment of the stream =>
      Hybrid Type I“

Mirja


> Am 10.02.2017 um 10:32 schrieb Stewart Bryant <[e-mail]>:
> 
> 
> 
> On 09/02/2017 19:30, Carlos Pignataro (cpignata) wrote:
>> Hi, Stewart,
>> 
>> Many thanks for the comments, please see inline.
>> 
>>> On Feb 9, 2017, at 1:50 PM, Stewart Bryant <[e-mail]> wrote:
>>> 
>>> 
>>> 
>>> On 09/02/2017 17:18, Carlos Pignataro (cpignata) wrote:
>>>> Passive means ‘solely by observation and without modification to the packet’ (https://tools.ietf.org/html/draft-ietf-ippm-active-passive-06#section-3.6).
>>> Carlos, that is not quit where we are going with passive. We use packet marking to batch the packets for loss measurement, and we are planning to trigger delay/jitter measurement through marking.
>>> 
>> I’ll follow down this tangent for a bit.
>> 
>> I understand and as you know I’m well aware of the (alternate) packet marking techniques and different methods.
>> 
>> However, the *current* definition is quite unambiguous:
>> 
>> https://tools.ietf.org/html/rfc7799#section-3.6
>> “
>> 3.6.  Passive Methods
>> 
>>    Passive Methods of Measurement are:
>> 
>>    o  based solely on observations of an undisturbed and unmodified
>>       packet stream of interest (in other words, the method of
>>       measurement MUST NOT add, change, or remove packets or fields or
>>       change field values anywhere along the path).
>> “
>> 
>> Since both those datapoints are rooted in IPPM, I’d suggest working through the definitions on IPPM and how marking fits (since it is not only on observation points)
>> 
>> Now, bringing this back to the relevance of the In-situ OAM (ioam) charter, my only point is that In-situ OAM is neither passive nor active.
> 
> ... and the point I make is that packet marking (which I think we have established is the only viable way of making accurate loss measurements in connectionless networking) is also neither active of passive according to these definitions.
> 
>> 
>>> As you know marking is much easier in MPLS that IP.
>>> 
>>> I think the key distinguisher is really that in-situ is about embedding OAM meta-data in user data traffic.
>> This is a good point.
>> 
>> I agree.
>> 
>> I believe this is already clear in the charter, all the way from the very first sentence:
>> 
>> “ It is based on telemetry information which is embedded within live data packets.”
>> 
>> Do you believe this is not clear in the charter? Do you have specific suggestions or concrete recommendations that can improve the charter text?
> 
> I suppose you could say: It is based on telemetry information which is embedded within live data packets and is distinct from packet parking methods being developed elsewhere in the IETF.
> 
> I have not thought it through, but I am wondering what distinguishes the packet types you list (IPv4, IPv6, VXLAN-GPE, LISP, NSH, SRv6, Geneve) from other packet types, the obvious one being MPLS. Not that I am at all keen on trying to get this into the simple fast forwarders we use for MPLS. In other words what is the generic class of packets you are targetting?
> 
> Stewart
> 
>> 
>> Thanks!
>> 
>>> - Stewart
>> 
>> —
>> Carlos Pignataro, [e-mail]
>> 
>> “Sometimes I use big words that I do not fully understand, to make myself sound more photosynthesis."
>> 
>