RE: Q: is framing-capabilities value=0 legal
"Ricky Charlet" <[email protected]>
| Newsgroups | gmane.ietf.l2tpext |
|---|---|
| Message-ID | <65399618110DC3429471909DC04562C70589409E@diamond.tasmannetworks.com> |
Ah, Thanks! I have been inclined toward accepting the tunnel. But L2TP is not my normal gig and I just wanted to 'ask the list' to get a sence of the matter from other folks. --- Ricky Charlet W: 408.754.1733 [email protected] --- _ ( ) ASCII ribbon campaign X - against HTML email / \ > -----Original Message----- > From: Ignacio Goyret [mailto:[email protected]] > Sent: Tuesday, November 01, 2005 10:45 AM > To: Ricky Charlet > Cc: Ignacio Goyret; [email protected] > Subject: RE: [L2tpext] Q: is framing-capabilities value=0 legal > > At 09:22 11/1/2005 -0800, Ricky Charlet wrote: > > I hope I paraphrase your response correctly to say that > Fr-cap=0 is > >allowable and its meaning is "I will not initiate", and it > may even be > >common from an LNS. > > > > > > But even if I see your meaning correctly, that still > does not help me > >in my situation. I'm the LNS, and I am receiving Fr-Cap=0 > from the LAC. > >The LAC I am dealing with has an obvious bug in that it's Fr-Cap > >setting is configurable but it ignores the configuration and always > >sends 0. Yet still, this is a popular LAC and I want to interoperate > >with it. > > > > From my LNS's perspective... should I be generous in > what I accept, or > >is there a reason I must abort the connection? > > > The question you have to ask yourself is what do you gain by > being picky on something inconsequential. > > Remember that L2TP (even 2661) is symmetrical: the role of > LNS or LAC is a per-call role, not a per-tunnel role. By > sending framing capabilities == 0, the L2TP peer is promising > that it won't send an OCRQ. From your point of view, why is > that a problem to you? > > If the remote's promise of never send an OCRQ is not a > problem to you (and, if your product is only an LNS, it > shouldn't be), ask yourself why reject the tunnel? > > By all means, I strongly suggest that you should follow Jon > Postel's principle of robustness (RFC793, section 2.10): be > conservative in what you send and liberal in what you accept > from others. > > Unless something really affects your functionality, what's > the harm in accepting it? > >