Re: Suggesting Improvements in RFC 1990.
Karl Fox <[email protected]>
| Newsgroups | gmane.ietf.pppext |
|---|---|
| Message-ID | <[email protected]> |
If someone were to ask me (which they haven't), I'd say leave RFC 1990 alone unless you fix it. By "fixing it" I mean combining two links, each with probability of failure p, and providing a combined link with probability of failure p^2 rather than 2p as RFC 1990 does. Karl P.S. I may be years out of date on this issue, but the last time I participated in an argument on the topic we couldn't come up with a way of making a single-link-failure not hose up the entire bundle, at least not until LQM shut it down and removed it from the bundle. Cleverer people than I may have solved this since then. _______________________________________________ Pppext mailing list [email protected] https://www1.ietf.org/mailman/listinfo/pppext