Iljitsch van Beijnum wrote:
On 28-jan-04, at 22:00, Randall R. Stewart (home) wrote:
In other words, when there is a serious solution to
multihoming -- ie, being able to preserve a connection when
using more than one IP Address -- it will likely work for IPv4.
Yes.. SCTP solves the problem for V4 and V6 (missed that bit last time).
Let me go through these one by one...
That remains to be seen. The list of issues that SCTP has or at least
seems to have is long. To name a few:
- increased overhead compared to TCP
Ok lets see. SCTP takes on average 4 more bytes per data packet then
TCP. However, if the TCP implementation enables timestamps then
that is not true and TCP takes more overhead by about 4 bytes...
Unless you are discussing another type of overhead...
- requires significant changes from applications
Ok let see.. for Mozilla we converted two lines of code
sd = socket(AF_INET6, SOCK_STREAM, IPPROTO_TCP);
became --------------> sd = socket(AF_INET, SOCK_STREAM, IPPROTO_SCTP);
and
setsockopt(sd, IPPROTO_TCP, TCP_NODELAY, &on_off, sizeof(on_off));
became ---------------> setsockopt(sd, IPPROTO_SCTP, SCTP_NODELAY,
&on_off, sizeof(on_off));
Now to take advantage of the stream feature you would need to do
more.. but for pure multi-homing one or two lines of change does
not seem that big of deal to me ...
- no backward compatibility of any kind
I am not sure what you mean by backward compatable? You definetly
can't have TCP and SCTP talk.. they are after all different protocols...
But if an application needs the redundancy move to it.. its there today
with about 2 lines of coding change...
- source address selection problem isn't addressed fully, if at all
I don't think I understand this issue either.. We have fully
addressed source address selection in the KAME implementation. It
is not a difficult problem.. it does require some code.. but any good
implementation must address this issue... And since the site scope
went away in IPv6 (at least for now) its easier to do then it was).
I can't be sure right now, but I also suspect SCTP could very well be
vulnerable to some of the threats identified lately in the multi6 wg.
I would have to look at your threats... if you add the dynamic address
feature.. sure
thats why the document has not progressed.. but that will change when the
Purpose Built Keys instantation in SCTP happens.. hopefully soon.. just
another
draft to write :-D
I think you may want to have a little closer look at SCTP.. you might
want to get the KAME BSD implementation and play with it for
a bit.. I think you would be amazed at how simple it is to
convert an application and you end up with multi-homing for
free with those 2 lines of code :->
R
--
Randall R. Stewart
815-477-2127 (office)
815-342-5222 (cell phone)