Diald

Diald

Post by Richar » Fri, 11 Feb 2000 04:00:00



Hello everybody.

I am using SuSE 6.3 with a ethernet card and a ISDN modem.

I have the following problem with my linux gateway router. When I start up
the Windhoosbox I can connect to Internet through linux but when the
connection goes down the sencond time it doesn't go up. I have a script
which sets the connection up (ip-up) It connects ones and it's silent again.

Is there something wrong with my diald ? I have search the internet for
configuration settings but no show. Please help.

Richard

 
 
 

1. 2.2.1, diald, diald-control, pipe, wish

I planned to post this in a greman news group, but then I found some
postings about this subject in this news group. I made some remarks at
the end.

Hallo

Ich konnte dem reissen nicht wiederstehen, kernel 2.2.1 musste her.

Ich habe jetzt ein problem mit dctrl. Oder auch nicht.

Dctrl shickt diald ein request seine monitoring daten durch eine pipe zu
schicken.

Mit kernel 2.2.x erhalte ich folgende fehler in /var/log/messages:

Feb 12 18:04:35 localhost diald[307]: FIFO: full monitor connection to
monitor /tmp/dctrl requested
Feb 12 18:04:35 localhost diald[307]: FIFO: could not open pipe
/tmp/dctrl: Device not configured.

Diald control startet zwar, l?sst sich aber nicht bedienen.
Wenn ich nach einiger zeit den monitor request von hand an diald
schicke,
gehts wieder.

Mein workaround:

-----------------------------------------
#!/bin/bash
#diald-control starten

# -fifo option benutzt?
FIFO=/etc/diald/diald.ctl
while  [ -n "$1" ] ;do
   if [ $1 = "-fifo" ];then
      FIFO=$2
   fi
   shift
done

sleep 5s
# $! ist der zuletzt gestartete hintergrundprozess (PID)
echo monitor /tmp/dctrl.$! > "$FIFO"
-----------------------------------------

Entsprechnde Zeilen aus /usr/bin/dctrl (wish script):

-----------------------------------------
#!/bin/sh
# the next line restarts using wish \

............

proc openFifo {} {
    global fifofd monfifo monfd default_FIFO

    # Turn off any previous monitoring
    if {$monfd!=""} {close $monfd}
    if {$monfifo!=""} {catch {exec rm -f $monfifo}}

    # get new monitoring fifo
    set fifofd [open $default_FIFO w]

    set monfifo /tmp/dctrl.[pid]
    catch {exec mkfifo -m 0600 $monfifo}
    fifoCmd "monitor $monfifo"
    set monfd [open $monfifo r]
    fileevent $monfd readable {stateChange}
----------------------------------

???
catch {exec mkfifo -m 0600 $monfifo}
???
was ist da los?

Versionen:
suse 6.0

diald-0.16.5

tk-4.2p2-35

und wo ist nun dieses 'feature' versteckt?
kernel, wish ,diald oder diald-control?

I had troubles using diald-control (dctrl).
I found a way to overcome this problem. (see my work around).
I think I found the line in the dctrl script that causes the hanging.
    catch {exec mkfifo -m 0600 $monfifo}
    of the  "proc openFifo {}" fragment above.
I don't know howto track it down further.

Kolja

2. Routing issue on three legged firewall

3. diald problem - pppd won't start from diald

4. simple cursor question

5. ip_masq and diald - diald won't bring up link for masq-ed packets

6. .forward file

7. Diald

8. NT system authentication with a Linux radius????

9. diald dying!

10. prob: diald/pppd

11. diald

12. DialD Problem

13. diald & smail(SMTP+UUCP)