OK, I have a very strange problem which I've been trying to solve for a few days but just couldn't, so maybe you could help me on this one...
I have an ECI Ethernet modem, configured my ADSL connection using drakconf, all is OK, it does connect, but on failure it will not reconnect - and I really don't understand what's going on.
The logs show:
Code:
Jan 23 10:48:53 phenix pppd[31408]: pppd 2.4.3 started by root, uid 0
Jan 23 10:48:53 phenix pppd[31408]: Using interface ppp0
Jan 23 10:48:53 phenix pppd[31408]: Connect: ppp0 <--> /dev/pts/1
Jan 23 10:49:24 phenix pppd[31408]: LCP: timeout sending Config-Requests
Jan 23 10:49:24 phenix pppd[31408]: Connection terminated.
Jan 23 10:49:24 phenix pppd[31408]: Using interface ppp0
Jan 23 10:49:24 phenix pppd[31408]: Connect: ppp0 <--> /dev/pts/2
Jan 23 10:49:28 phenix pppoe[31443]: Timeout waiting for PADO packets
Jan 23 10:49:30 phenix pppoe[31451]: PPP session is 5281
Jan 23 10:49:30 phenix pppd[31408]: Connection terminated.
Jan 23 10:49:30 phenix pppd[31408]: tcflush failed: Bad file descriptor
Jan 23 10:49:30 phenix pppd[31408]: Using interface ppp0
Jan 23 10:49:30 phenix pppd[31408]: Connect: ppp0 <--> /dev/pts/3
Jan 23 10:49:40 phenix pppd[31408]: Connection terminated.
Jan 23 10:49:40 phenix pppd[31408]: tcflush failed: Bad file descriptor
And it goes on and on and on... What's more, all spawned pppoe processes take some time to disappear, and the original one never gets killed. Stranger yet, they run in their own session, unlike the first one, which is a child of pppd - as expected. That's this tcflush error which mostly puzzles me. Never seen that before.
I changed one thing in /etc/ppp/peers/ppp0, add a -D option to pppoe in order to see what happens. But for now I don't have the faintest idea.
Noteworthy is that autoreconnection DOES work with Mandrake 10.1... And the /etc/ppp/peers/ppp0 file is quite different as well. Right now I'm into figuring out if any of the added options in 2006 can trigger this.
02-12-2006
