Re: [trill] TRILL OAM Requirements: Sections

[Thomas Narten <[e-mail]>]

Hi Tissa.

"Tissa Senevirathne (tsenevir)" <[e-mail]> writes:

> "Section" is a general terminology used to identify a partial segment of
> the path between two end points. Please refer to RFC 5860 "Requirements
> for Operations, Administration, and Maintenance (OAM) in MPLS Transport
> Networks". 5860 is listed as a reference in the document.
> draft-tissa-trill-oam-req assumes readers are familiar with various OAM
> concept in the industry. Relevant documents are also listed as
> references.

You didn't really respond to my core point though.

Why should TRILL OAM support this notion? What value does it have in a
TRILL network?

I cannot agree to say TRILL OAM needs to support this notion because I
don't understand the implications in terms of how it will impact the
OAM mechanisms and protocols that need to be defined.

Without understanding that (in a rough, architectural sense), I cannot
support it.

I also just don't see the value. Do operators define sections? How is
that useful in a TRILL network?

Thomas


> -----Original Message-----
> From: [e-mail] [mailto:[e-mail]] On Behalf
> Of Thomas Narten
> Sent: Thursday, April 26, 2012 11:13 AM
> To: [e-mail]
> Subject: [trill] TRILL OAM Requirements: Sections

> draft-tissa-trill-oam-req-00.txt says;

> >    Section: A section is a partial segment of a path between any two
> >    given RBridges. As an example, consider the case where RB1 is
> >    connected to RBx via RB2 and RB3. The segment between RB3 to RBx is
> >    referred to as a Section of the path RB1 to RBx.

> Where does this notion of a section come from and why do we need it in
> TRILL?  AFAIK, TRILL doesn't have this notion, so this is something that
> OAM is creating on its own.

> Then later:

> >    An RBridge SHOULD have the ability to verify the above connectivity
> >    tests on sections. As an example, assume RB1 is connected to RB5
> via
> >    RB2, RB3 and RB4. An operator SHOULD be able to verify the RB1 to
> >    RB5 connectivity on the section from RB3 to RB5. The difference is
> >    that the ingress and egress TRILL nicknames in this case are RB1
> and
> >    RB5 as opposed to RB3 and RB5, even though the message itself may
> >    originate at RB3.

> I am not comfortable with this requirement. This would seem to
> complicate the requirements and I'm not sure why it is needed. I'm also
> a bit wary about what the implication is in terms of what sort of
> functionality or mechanism would be required to be able to achieve this
> requirement.

> (The notion of "sections" comes up in several places in the document,
> and the above applies to all of them.)

> Do we really want or need sections?

> Thomas

> _______________________________________________
> trill mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/trill