Re: ForCES TLV size
Jamal Hadi Salim <[email protected]> Sun, 19 Aug 2007 07:58:23 -0400
| Newsgroups | gmane.ietf.forces |
|---|---|
| Organization | ZNYX Networks |
| Message-ID | <1187524703.5165.40.camel@localhost> |
Weiming, You dont need to break down any other TLV. Only the portion that contains the data as described in the proposal. Any outer TLV contains this part and its length is appropriately set. Does that make sense? In any case, is this issue worth time we are spending? It is a corner case; and rules of optimization are simple: you optimize only for the common case not a corner case. Lets please close this issue. Should we just put a note somewhere in the appendix and move on? cheers, jamal On Sun, 2007-19-08 at 19:49 +0800, Wang,Weiming wrote: > To break down a Oper-TLV into 64k fragments, then we have to break down a LFB-select TLV > which includes even one 64k Oper-TLV into more fragments. We then also need to define a > Redirect-TLV, Redirect-data TLV on how they are broken down into less than 64k fragments. > > Things are getting mess more or less. > > thanks, > Weiming