On 21-Aug-2000 Lindsay McPhee wrote:
> Hi all,
> Please forgive if I have sent a message that is too long, but I thought
> in order to get the assistance required I needed to supply all available
> information.
Quite right too, IMHO.
> I have installed Red Hat 6.1 and daild 0.99.4.[...] The
> problem: I can dial - with error messages - and login but I cant start
> PPP successfully!! If I use /sbin/ifup ppp0 I get all the way. So I
> figure it's to do with the connect script of the diald configuration.
First off, how did you install diald? Did you build it from source yourself or
install from an RPM somewhere?
The reason I ask is the first collection of log messages:
> Aug 21 12:59:42 eric modprobe: can't locate module tap0
> Aug 21 12:59:42 eric modprobe: can't locate module tap1
etc.
Diald can be built to use ethertap or SLIP for monitoring traffic. However,
ethertap isn't on the list of modules available with the stock kernel on this
RH6.2 machine.
> Aug 21 12:59:46 eric diald[1513]: start sl0: SIOCSIFMETRIC: Operation
> not supported
On the other hand, it does look as if something - and that SIOCSIFMETRIC
messages is a classic diald hallmark - is using sl0 (SLIP). So maybe diald is
falling back to SLIP after ethertap won't work, and all is OK after all. (Must
check sources and see if it does that).
> Aug 21 12:59:46 eric diald[1513]: Calling site 192.168.1.9
> Aug 21 12:59:47 eric connect: Initializing Modem
> Aug 21 12:59:48 eric connect: Dialing system
> Aug 21 13:00:15 eric connect: Connected
> Aug 21 13:00:15 eric connect: Loggin in
> Aug 21 13:00:16 eric connect: Protocol started
> Aug 21 13:00:16 eric diald[1513]: Connected to site 192.168.1.9
> Aug 21 13:00:16 eric diald[1513]: Running pppd (pid = 1544).
> Aug 21 13:00:16 eric modprobe: can't locate module char-major-108
> Aug 21 13:00:17 eric kernel: PPP: version 2.3.7 (demand dialling)
> Aug 21 13:00:17 eric kernel: PPP line discipline registered.
> Aug 21 13:00:17 eric kernel: registered device ppp0
> Aug 21 13:00:17 eric pppd[1544]: pppd 2.3.10 started by root, uid 0
> Aug 21 13:00:17 eric pppd[1544]: Using interface ppp0
> Aug 21 13:00:17 eric pppd[1544]: Connect: ppp0 <--> /dev/ttyS1
> Aug 21 13:00:23 eric kernel: PPP BSD Compression module registered
> Aug 21 13:00:23 eric kernel: PPP Deflate Compression module registered
> Aug 21 13:00:24 eric pppd[1544]: local IP address 203.57.114.143
> Aug 21 13:00:24 eric pppd[1544]: remote IP address 203.57.114.129
> Aug 21 13:00:24 eric diald[1513]: New addresses: local 203.57.114.143,
> remote 203.57.114.129, broadcast 0.0.0.0
> Aug 21 13:00:24 eric diald[1513]: start ppp0: SIOCSIFMETRIC: Operation
> not supported
That all looks 100% OK. You can safely ignore the SIOCSIFMETRIC whingeing -
this is stuff necessary for a 2.0 kernel but not required or supported by 2.2.
My working diald produces them exactly as above.
To be honest, it looks from the above as if diald is working fine. Diald
doesn't start pppd until after the chat script has completed, so I don't think
there is much wrong with that. And pppd is, from the above, connecting and
starting the link OK.
What does your routing table look like after starting diald and at that point?
And if you're running X does dctrl think you're online, and can it see traffic
passing?
Oh, one more thought. What's in your /etc/ppp/options? Remember diald requires
this to be pretty much empty. Though I have to confess that /sbin/ifup could run
into similar trouble. On which subject, what's in
/etc/sysconfig/network-scripts/chat-ppp (or wherever the pppd chat script is)?
FWIW, I have found that authentication via PPP is a lot quicker than scripted
login - my diald chat script exits just before
># We're connected try to log in.
> message "Loggin in"
--
Jim Hague - [EMAIL PROTECTED] (Work), [EMAIL PROTECTED] (Play)
Never trust a computer you can't lift.
--
SLUG - Sydney Linux User Group Mailing List - http://slug.org.au/
More Info: http://slug.org.au/lists/listinfo/slug
