[dapper] pppd does not always honor the unit option
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
ppp (Ubuntu) |
Confirmed
|
Undecided
|
Unassigned |
Bug Description
Binary package hint: ppp
To make the ADSL link interface distinguishable from VPN interfaces, I use the "unit" option to set the ADSL ppp interface to "ppp200" (while my VPN links start with ppp100). But sometimes, I see that ppp does not create ppp200, but the standard "ppp0" instead, breaking my firewall configuration:
Apr 7 17:30:07 sauron pppd[15784]: Terminating on signal 15
Apr 7 17:30:07 sauron pppd[15784]: Connect time 1440.3 minutes.
Here are the 24 hours...
Apr 7 17:30:07 sauron pppd[15784]: Script /etc/ppp/ip-down started (pid 10302)
Apr 7 17:30:07 sauron pppd[15784]: sent [LCP TermReq id=0x7 "User request"]
Apr 7 17:30:07 sauron pppd[15784]: Script /etc/ppp/ip-down finished (pid 10302), status = 0x0
Apr 7 17:30:10 sauron pppd[15784]: sent [LCP TermReq id=0x8 "User request"]
Apr 7 17:30:12 sauron pppd[10340]: Plugin rp-pppoe.so loaded.
Apr 7 17:30:12 sauron pppd[10342]: pppd 2.4.4b1 started by root, uid 0
Apr 7 17:30:12 sauron pppd[10342]: PADS: Service-Name: ''
Apr 7 17:30:12 sauron pppd[10342]: PPP session is 11778
Apr 7 17:30:12 sauron pppd[10342]: using channel 14
Apr 7 17:30:12 sauron pppd[10342]: Couldn't allocate PPP unit 200 as it is already in use
Here the trouble starts...
Apr 7 17:30:12 sauron pppd[10342]: Using interface ppp0
Apr 7 17:30:12 sauron pppd[10342]: Connect: ppp0 <--> eth1
Apr 7 17:30:12 sauron pppd[10342]: sent [LCP ConfReq id=0x1 <mru 1492> <magic 0xda5fc088>]
Apr 7 17:30:12 sauron pppd[10342]: rcvd [LCP ConfReq id=0x1 <mru 1492> <auth pap> <magic 0x55e0c262>]
Apr 7 17:30:12 sauron pppd[10342]: sent [LCP ConfAck id=0x1 <mru 1492> <auth pap> <magic 0x55e0c262>]
Apr 7 17:30:12 sauron pppd[10342]: rcvd [LCP ConfAck id=0x1 <mru 1492> <magic 0xda5fc088>]
Apr 7 17:30:12 sauron pppd[10342]: sent [LCP EchoReq id=0x0 magic=0xda5fc088]
(...PPPoE negotiations...)
Apr 7 17:30:12 sauron pppd[10342]: Cannot determine ethernet address for proxy ARP
Apr 7 17:30:12 sauron pppd[10342]: local IP address 91.21.83.167
Apr 7 17:30:12 sauron pppd[10342]: remote IP address 217.0.118.51
Apr 7 17:30:12 sauron pppd[10342]: Script /etc/ppp/ip-up started (pid 10352)
Apr 7 17:30:13 sauron pppd[15784]: Connection terminated.
Here finally the old ppp finishes, only now freeing ppp200...
Apr 7 17:30:13 sauron pppd[15784]: Modem hangup
Apr 7 17:30:13 sauron pppd[15784]: Exit.
This usually happens (I'm not yet sure if it happens always or just most of the time) when my ISP disconnects the line after 24 hours (this is a usual customer-nagging practice in Germany).
Instead of silently ignoring the "unit 200" option, IMHO ppp should abort with an error if ppp200 is already in use, or retry until the requested device name is no longer in use. For now I "fixed" it with a cronjob which checks my pppd status and restarts it if necessary.
Ciao
Martin
Is this symptom still reproducible in 8.10 beta?