Re: Networkhood file system.

Gregory Maxwell (linker@z.ml.org)
Wed, 26 Aug 1998 22:25:00 -0400 (EDT)


On Wed, 26 Aug 1998, Richard B. Johnson wrote:

> You know how Micro$oft gets this information? They steal it. They
> rob CPU time from everyone on the local network. These garbage
> machines trash the LAN with continuous "chatter", which turns into
> a nearly 100% load on the network when you have a room full of
> these abortions. No thanks.

Wow, colorful.

We could use wins, but it would need to be expanded to support more then
smb. We could also create our own, doing something like this:

A netlookd watches for microsoft garbage packets and accepts 'who's on'
messages from a wins type server. Services providing services can be
listed there (either by netlookd at boot on unix boxen, or by it watching
M$ brodcasts, takling to wins servers, watching for IPX 'hellos', and
finally from a config file)..

This would reduce traffic, as the 'who's online' can by updated on the
client by sending a unicast UDP query to the server and the server can
decide to send a unicast responce or a broadcast responce. (Broadcasts
wouldn't be sent out with greater then X frequency). Clients would never
directly generate broadcast traffic. Multicast should be supported too for
sites nice enough to support it. The server could also send out
unicast push updates to listening clients and finally, the client could
connect to the server with a TCP connection to keep a constant view of the
network state.

Sounds nice enough, and much less of a bandwidth waste. MS allows you to
use WINS which really lowers traffic, and stops the brodcasts and samba
acts as good wins server, if it could be expanded to support non smb stuff
it would be fine.

ObKernel: Would the new BSD packet filters be a good way of listening for
winlose brodcasts to update a list?

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html