Hello again,
Well, last night I finally got my cablemodem working with diald :-)
Here's the story and some suggestions for changes to diald and its docs.
My problem was that diald never made the transition from START_LINK to UP.
This was my fault, since I had carefully taken my eth0 device down prior to running
diald, expecting that diald would take it up like it does with ppp devices.
Oh well, my fault, I guess - but I would like to suggest enhancing the diald docs in
the parts where they (should) mention "mode dev" setups. In short:
"device" statements must include a bogus /dev/ part (/dev/eth0 in my case).
devices must be up (unlike ppp) and configured to the same IP# as the "local"
statement in diald.conf
With respect to the /dev/ part, it would of course be better if diald wouldn't
coredump if you just put 'eth0' in the statement, which seems more natural...
Contrary to what I wrote earlier, I've decided to do my signon/off in the
"addroute"/"delroute" scripts, which immediately worked fine for "addroute". The
"delroute" part was another story, though :-/
To do the signoff, I have to telnet to a host on my ISP's network, which requires a
route to it via the eth0 interface.
The problem is that diald deletes these routes before calling "delroute", so I have to
add the routes (gateway/default) again, do the signoff and delete the routes once more.
I suggest that the route setups in diald be made symmetrical, i.e.:
diald adds routes
"addroute" is called
diald waits...
"delroute" is called
diald deletes routes
The two last steps are reversed in the current diald.
My suggestion would mean 4 less calls to /sbin/route for me...
There's probably some very good reason for it being the way it is(?), but at least
I've made my suggestion now :-)
BTW, if anyone has recommendations about the filter setup for "fast connect" metered
lines (e.g. ISDN, cablemodems), I'd be quite interested in hearing about it ;-)
Best regards,
Torben Thellefsen
-
To unsubscribe from this list: send the line "unsubscribe linux-diald" in
the body of a message to [EMAIL PROTECTED]
