[hipl-users] Re: LSI functionality in HIPL
- From: Miika Komu <miika@xxxxxx>
- To: hipl-users@xxxxxxxxxxxxx
- Date: Mon, 22 May 2006 15:44:39 +0300 (EEST)
On Mon, 22 May 2006, Stephen Herborn wrote:
Currently when hipd starts, an LSI is generated for each HIT. The LSIs are
assigned to the dummy0 interface, and additionally routes are inserted into
the hosts ipv4 routing table corresponding to each LSI.
My question is are these LSIs that are functional in any way?
No. It was supposed to part of Abhinav's bachelor thesis work but we had
to limit the topic to get it done.
If so are there any sample applications available that use LSI, or any
other way I can see examples of how LSIs may used by ipv4-only
applications?
Currently not. Would you be interested in making the LSIs work in the HIP
daemon and contributing the code?
By the way I did see this post from Miika in Feb '06 saying LSI is not yet
implemented, so perhaps its still unlikely that its available by now (May
'06)
//www.freelists.org/archives/hipl-users/02-2006/msg00002.html
No, it is not done yet. It requires a substantial amount of modifications
and testing of the HIP daemon. Also, there has been two other reasons why
we haven't hurried towards this approach:
* Most commonly used applications already seem to support IPv6
addressing
* We have been investigating if we can do some socket wrapping in the
userspace, similarly as in TESLA [1]
But of course, counterarguments are welcome :)
[1] http://nms.lcs.mit.edu/tesla/
--
Miika Komu miika@xxxxxx http://www.iki.fi/miika/
Other related posts:
