[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

n-hurd networking

From: Hisham Kotry
Subject: n-hurd networking
Date: Thu, 8 Aug 2002 12:20:38 -0700 (PDT)

Ugh, my mistake, I made an assumption that all n-hurds
(neighbour hurds) would share the same IP address with
the first hurd started on the machin. 

Well, as long as every n-hurd has its own IP addy,
most of the stuff will be fine.

The general concept so far is that there will be a
Simple Interface Manager (SIM, it isn't called an
abstract/virtual interface because it has more duties
than provide a common interface to the underlying
hardware) that does some Layer2 classification, I
guess adding an EXCEPTION target to the packet
classifier is needed for multi-cast addresses
(otheriwse, it wont allow two n-hurds to bind to the
same multi-cast address).

I wont go into any hurd-net design details
<sarcasm>and I know that all of you are waiting for it
:p </sarcasm> until Jeff's a bit more free to discuss
some issues.

Simon, I think Thomas' idea of a collective is more
like a distributed hurd where (possibly) every server
lies on a diffrent host, a problem that appeared was
who would have control of "/"? in monolothic kernels,
if the host controling "/" crashed, the system would
loose all file system access, on the Hurd, you'd just
loose everything (the FS acts as a rendez-vous for
translators). So, probably someone will have to come
up with a distributed FS, or just come up with another
concept for rendez-vous'ing.


Do You Yahoo!?
HotJobs - Search Thousands of New Jobs

reply via email to

[Prev in Thread] Current Thread [Next in Thread]