From: Mas Kato (loomis_towcar@xxxxxxxxxxxxxx)
Date: Wed Apr 03 2002 - 16:51:50 GMT-3
[demime could not interpret encoding binary - treating as plain text]
John,
Once again, your logic is impeccable. It's uncanny, but I can remember drawing
similar conclusions about this issue, so I can certainly relate to your frustra
tion.
I'm not trying to justify their choice of keywords, but perhaps this may help:
In the same "Designing DLSw+ Internetworks" document under "DLSw+ Transport Fle
xibility," the last two bullets refer to the Direct and DLSw Lite transport met
hods. I would submit that these are the two modes the 'pass-thru' keyword and a
ssociated 'frame-relay map' commands are specifying.
I think what throws a wrench into the works is the fact that DLSw Lite still us
es LLC2 encapsulation per RFC 1490 for reliable transport across the FR cloud,
even though the "LLC2 sessions" on each end are terminated locally.
What can also be confusing is, according to the standard, RFC 1795, DLSw always
terminates the LLC2 sessions. There is no notion of 'pass-thru.' But then agai
n, in the standard, there is also no notion of any transport method other than
TCP/IP. The FST, Direct and DLSw Lite methods are all non-standard (although I
think they submitted DLSw Lite to the standards bodies).
I hope this helps.
Best of luck on your exam,
Mas Kato
https://ecardfile.com/id/mkato
>Date: Tue, 2 Apr 2002 16:55:21 -0500
> ccielab@groupstudy.com John Neiberger <neiby@ureach.com> DLSw+ over FR with p
ass thru PART 2Reply-To: John Neiberger <neiby@ureach.com>
>
>Well, I thought that we used 'frame map
>llc2' with pass thru (meaning that we're
>passing thru LLC2 frames) and 'frame map
>dlsw' without pass through, meaning that
>we're locally acknowledging LLC2 and not
>'passing it thru' over the link. I then
>checked the 12.1 Configuration Guides on CCO
>and there is an example that agreed with me.
>
>However, I was notified in private that I
>was wrong so I checked CCO again. If you
>look at the Designing DLSw+ Internetworks
>document, it shows that I was wrong.
>According to that document:
>
>WITH pass-thru:
>
>frame map dlsw
>
>WITHOUT pass-thru:
>
>frame map llc2
>
>This is counterintuitive to me, but since
>we're talking about Cisco that's nothing
>new. :-) I still say it makes NO sense to
>map llc2 frames to a DLCI if we're not
>passing LLC2 frames!! If we're not passing
>thru the llc2 frames, then we're passing
>DLSw-encapsulated frames and should map dlsw
>to the PVC, *not* llc2. If we *are* passing
>thru LLC2 frames, then it makes more sense
>to then map llc2 to the PVC. Apparently,
>Cisco completely disagrees with this logic.
>
>Having inconsistent documentation on CCO
>makes it even more confusing. I've been
>looking at the configuration guides all
>along thinking I was right. Now I
>definitely need to brush up on my dlsw
>configs over the next couple of days so I
>can keep this straight!
>
>John
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:57:52 GMT-3