On Oct 21, 2010, at 16:48, Margaret Wasserman wrote:
[in her RPL review about RPL's various forms of encapsulation:]

> (1) 6lowpan,
>            which is the only link type I know of that RPL has been
>            run over to date, has an MTU of 1280, not 1280+++ as
>            required by the 6lowpan spec,

(I think you meant 6man/RPL spec at the end.)

Indeed.

However, the number 1280 is somewhat arbitrary for 6LoWPAN, as we have to piece 
L3 packets together out of ~100 byte L2 fragments anyway.  The actual 
adaptation layer format is fine with any MTU up to 2047; we just chose the 
lowest MTU possible*) in IPv6 as a concession for constrained nodes (and 1280 
still is quite heavy for some of them).

So it is not unconceivable that 6LoWPAN could help RPL routers by adding a mode 
that extends the MTU somewhat.
Of course, the details of such an extension would need to be defined, 
preferably in conjunction with the rest of the 6man/RPL stack.

Gruesse, Carsten

*) The assumption was that, as a stub network, a 6LoWPAN would not need to 
support tunneling.
That assumption may have been wrong.

_______________________________________________
6lowpan mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lowpan

Reply via email to