rp-pppoe-3.5-r5 gives me this message if i try to stop connection with adsl-stop: adsl-stop: No ADSL connection appears to be running this is so strange because if I use rp-pppoe-3.5-r2, the connection stops well (but I have other problems). I use the service /etc/init.d/rp-pppoe on the runlevel default so when i boot up the computer my adsl connection starts and when i boot off my computer rp-pppoe gives the same message. Reproducible: Always Steps to Reproduce: 1.adsl-stop 2. 3.
do you have a PIDFILE variable in /etc/ppp/pppoe.conf? do that file exist after you run adsl-start?
in /etc/ppp/pppoe.conf i have CF_BASE=`basename $CONFIG` PIDFILE="/var/run/$CF_BASE-adsl.pid" and in /var/run i have 63-adsl.pid (the PID of this process doesn't exist at all!!!) 63-adsl.pid.pppd (with the right PID of the process pppd) 63-adsl.pid.pppoe (with the right PID of the process pppoe) I doesn't have the process with the PID listed in 63-adsl.pid (adsl-connect i think)! Using rp-pppoe-3.5-r2 thi process exist for all the time of connection.
are you familiar with bash? can you put some info messages in adsl-connect script to see why adsl-connect process gets killed? if not, please put here your configuration. without username and password of course.
I'm seeing the same problem here. It looks like the named pipe patches are the culprit for causing this. On my system, adsl-connect would create a file /tmp/sh-np-... everytime it connects. When bringing the connection down, it would look in the old places to find the pid which isn't there (adsl-status does the same). Please note that I'm not using net.ppp0 but adsl-start directly.
The /tmp/sh-np-... files are in fact named pipes, by the way.
so, if /tmp/sh-np* are named pipes then they ain't pid files. do you still have adsl-connect process after you start the connection? if not, please insert some log messages in adsl-connect script to see when it gets killed. first of all, be sure it gets started. hint: if you have RETRY_ON_FAILURE=no, it will end without any log messages.
I didn't say it would try using /tmp/sh-np-* for storing the PID. It stores them fine under /var/run/sh-np-* (the referenced processes do exist), but when invoking adsl-status, it tries to read /var/run/pppoe.conf-adsl.pid.pppoe, which isn't there. As for adsl-connect, there is no process with that name here. I'll try to find out tomorrow what's going on there.
This is my configuration file: ETH='eth0' USER='*****' DEMAND=no DNSTYPE=SPECIFY PEERDNS=no DNS1=130.244.127.161 DNS2=130.244.127.169 DEFAULTROUTE=yes CONNECT_TIMEOUT=30 CONNECT_POLL=2 ACNAME= SERVICENAME= PING="." CF_BASE=`basename $CONFIG` PIDFILE="/var/run/$CF_BASE-adsl.pid" SYNCHRONOUS=no CLAMPMSS=1412 LCP_INTERVAL=20 LCP_FAILURE=3 PPPOE_TIMEOUT=80 FIREWALL=STANDALONE LINUX_PLUGIN= PPPOE_EXTRA="" PPPD_EXTRA="" FIREWALL=STANDALONE
This problem is a consequence of gentoo-netscripts.patch added by vapier in order to solve bug #68934. If you rename /etc/ppp/pppoe.conf and re-emerge rp-pppoe you'll see that we install a pppoe.conf with no CF_BASE variable. Instead, a proper pppoe.conf file should have: # NB: Gentoo overrides PIDFILE when adsl-start is run from the # networking scripts. This setting has no effect in that case. PIDFILE="/var/run/adsl.pid"
*** Bug 84394 has been marked as a duplicate of this bug. ***