Thu, Jan 03, 2013 at 02:31:16PM -0800, Tony Li: > As I've said before, errors can be reasonably classified into two groups: > syntactic and semantic. > > A syntactic error would be any inconsistency of length fields, > incompatibility between a type and a length, or any other error that would > introduce any doubt about the parsing of the message. Once this type of > error has occurred, then the remainder of the data stream is wholly in doubt. > A session reset seems like the only (conservative) way of handling this, as > it's unclear that further data would be accurate.
Amen. if you've come upon a syntactic error and therefore have zero confidence that any bit in the message is correct, I dont understand how you could have any confidence that treat-as-withdraw is treating the right prefix(es) as withdrawn, or even logging the right prefix(es). you cant even have confidence about the length of the message. The aforementioned sesson re-establishment suppression seems a far better and consistent approach; it also doesnt seem to be the subject for an rfc. _______________________________________________ GROW mailing list GROW@ietf.org https://www.ietf.org/mailman/listinfo/grow
