Linux rp-pppoe plugin re-connect problem

Hi All,

I am facing a problem in re-connecting the pppoe session in my target.
When the link goes down inorder to make ppp reconnect i have enabled
the "persist" option. I'm gettting the error "write: Bad file
descriptor (9)". The re-connect happens successfuly sometimes. And
eventhough i have specified the holdoff as 60, the re-connection
happens around 2 hours after the connection goes down..Will appreciate
your inputs on this.

Here is my PPP options
======================
asyncmap 0
debug #PPPD will output connection information
defaultroute #This interface will be a default gateway.
nodeflate #Do not use compression
nobsdcomp #Do not use bsd compression
noccp #Disable CCP (Compression Control Protocol)
noaccomp #Disable Address/Control compression
nopcomp #Disable protocol field compres.
novj #Disable Van Jacobson TCP/IP header compres.
novjccomp #Disable the connection-ID compression
nopredictor1 #Do not accept or agree to Predictor-1 compres.
name new_miniv_1
plugin rp-pppoe.so
persist #re-open connection if ppp is down
holdoff 60
lcp-echo-interval 20
lcp-echo-failure 15


___________________________________________________________
<31>Jan 2 01:59:57 pppd[1490]: PADS: Service-Name: ''

<30>Jan 2 01:59:57 pppd[1490]: PPP session is 531

<31>Jan 2 01:59:57 pppd[1490]: using channel 21

<30>Jan 2 01:59:57 pppd[1490]: Using interface ppp0

<29>Jan 2 01:59:57 pppd[1490]: Connect: ppp0 <--> tap0

<28>Jan 2 01:59:57 pppd[1490]: Couldn't increase MTU to 1500

<28>Jan 2 01:59:57 pppd[1490]: Couldn't increase MRU to 1500

<31>Jan 2 01:59:57 pppd[1490]: sent [LCP ConfReq id=0xc <magic
0x4cdbe8b4>]

<29>Jan 2 01:59:57 pppd[1490]: Connection terminated.

<31>Jan 2 02:00:00 pppd[1490]: sent [LCP ConfReq id=0xc <magic
0x4cdbe8b4>]

<27>Jan 2 02:00:00 pppd[1490]: write: Bad file descriptor (9)

<31>Jan 2 02:00:03 pppd[1490]: sent [LCP ConfReq id=0xc <magic
0x4cdbe8b4>]

<27>Jan 2 02:00:03 pppd[1490]: write: Bad file descriptor (9)

<31>Jan 2 02:00:06 pppd[1490]: sent [LCP ConfReq id=0xc <magic
0x4cdbe8b4>]

<27>Jan 2 02:00:06 pppd[1490]: write: Bad file descriptor (9)

I think there is some problem with the timers. The problem occurs
whenever there is a LCP Term from the peer for the pppoe connection and
PPPoe client trying to re-establish. The pppoe negotiation is
successful, but the ppp connection is getting terminated. Appreciate
any help on this

<30>Jan 1 00:08:02 pppd[75]: LCP terminated by peer (User request)
<30>Jan 1 00:08:02 pppd[75]: Connect time 0.5 minutes.
<30>Jan 1 00:08:02 pppd[75]: Sent 0 bytes, received 0 bytes.
<31>Jan 1 00:08:02 pppd[75]: Script /etc/ppp/ip-down started (pid 89)
<28>Jan 1 00:08:02 pppd[75]: Couldn't increase MTU to 1500
<28>Jan 1 00:08:02 pppd[75]: Couldn't increase MRU to 1500
<31>Jan 1 00:08:02 pppd[75]: sent [LCP TermAck id=0x2]
<31>Jan 1 00:08:02 pppd[75]: Script /etc/ppp/ip-down finished (pid
89), status = 0x0
<29>Jan 1 00:08:05 pppd[75]: Connection terminated.
<31>Jan 1 00:08:11 pppd[75]: PADS: Service-Name: ''
<30>Jan 1 00:08:11 pppd[75]: PPP session is 1
<31>Jan 1 00:08:11 pppd[75]: using channel 2
<30>Jan 1 00:08:11 pppd[75]: Using interface ppp0
<29>Jan 1 00:08:11 pppd[75]: Connect: ppp0 <--> tap0
<28>Jan 1 00:08:11 pppd[75]: Couldn't increase MTU to 1500
<28>Jan 1 00:08:11 pppd[75]: Couldn't increase MRU to 1500
<31>Jan 1 00:08:11 pppd[75]: sent [LCP ConfReq id=0x2 <mru 1492>
<magic 0x8b152f1b>]
<29>Jan 1 00:08:11 pppd[75]: Modem hangup
<29>Jan 1 00:08:41 pppd[75]: Modem hangup
<29>Jan 1 00:08:41 pppd[75]: Connection terminated.


Thanks
Vanitha
[email]vanitha@agilis.st.com.sg[/email] wrote:[color=blue]
> Hi All,
>
> I am facing a problem in re-connecting the pppoe session in my target.
> When the link goes down inorder to make ppp reconnect i have enabled
> the "persist" option. I'm gettting the error "write: Bad file
> descriptor (9)". The re-connect happens successfuly sometimes. And
> eventhough i have specified the holdoff as 60, the re-connection
> happens around 2 hours after the connection goes down..Will appreciate
> your inputs on this.
>
> Here is my PPP options
> ======================
> asyncmap 0
> debug #PPPD will output connection information
> defaultroute #This interface will be a default gateway.
> nodeflate #Do not use compression
> nobsdcomp #Do not use bsd compression
> noccp #Disable CCP (Compression Control Protocol)
> noaccomp #Disable Address/Control compression
> nopcomp #Disable protocol field compres.
> novj #Disable Van Jacobson TCP/IP header compres.
> novjccomp #Disable the connection-ID compression
> nopredictor1 #Do not accept or agree to Predictor-1 compres.
> name new_miniv_1
> plugin rp-pppoe.so
> persist #re-open connection if ppp is down
> holdoff 60
> lcp-echo-interval 20
> lcp-echo-failure 15
>
>
> ___________________________________________________________
> <31>Jan 2 01:59:57 pppd[1490]: PADS: Service-Name: ''
>
> <30>Jan 2 01:59:57 pppd[1490]: PPP session is 531
>
> <31>Jan 2 01:59:57 pppd[1490]: using channel 21
>
> <30>Jan 2 01:59:57 pppd[1490]: Using interface ppp0
>
> <29>Jan 2 01:59:57 pppd[1490]: Connect: ppp0 <--> tap0
>
> <28>Jan 2 01:59:57 pppd[1490]: Couldn't increase MTU to 1500
>
> <28>Jan 2 01:59:57 pppd[1490]: Couldn't increase MRU to 1500
>
> <31>Jan 2 01:59:57 pppd[1490]: sent [LCP ConfReq id=0xc <magic
> 0x4cdbe8b4>]
>
> <29>Jan 2 01:59:57 pppd[1490]: Connection terminated.
>
> <31>Jan 2 02:00:00 pppd[1490]: sent [LCP ConfReq id=0xc <magic
> 0x4cdbe8b4>]
>
> <27>Jan 2 02:00:00 pppd[1490]: write: Bad file descriptor (9)
>
> <31>Jan 2 02:00:03 pppd[1490]: sent [LCP ConfReq id=0xc <magic
> 0x4cdbe8b4>]
>
> <27>Jan 2 02:00:03 pppd[1490]: write: Bad file descriptor (9)
>
> <31>Jan 2 02:00:06 pppd[1490]: sent [LCP ConfReq id=0xc <magic
> 0x4cdbe8b4>]
>
> <27>Jan 2 02:00:06 pppd[1490]: write: Bad file descriptor (9)[/color]

I think there is some problem with the timers. The problem occurs
whenever there is a LCP Term from the peer for the pppoe connection and
PPPoe client trying to re-establish. The pppoe negotiation is
successful, but the ppp connection is getting terminated. Appreciate
any help on this

<30>Jan 1 00:08:02 pppd[75]: LCP terminated by peer (User request)
<30>Jan 1 00:08:02 pppd[75]: Connect time 0.5 minutes.
<30>Jan 1 00:08:02 pppd[75]: Sent 0 bytes, received 0 bytes.
<31>Jan 1 00:08:02 pppd[75]: Script /etc/ppp/ip-down started (pid 89)
<28>Jan 1 00:08:02 pppd[75]: Couldn't increase MTU to 1500
<28>Jan 1 00:08:02 pppd[75]: Couldn't increase MRU to 1500
<31>Jan 1 00:08:02 pppd[75]: sent [LCP TermAck id=0x2]
<31>Jan 1 00:08:02 pppd[75]: Script /etc/ppp/ip-down finished (pid
89), status = 0x0
<29>Jan 1 00:08:05 pppd[75]: Connection terminated.
<31>Jan 1 00:08:11 pppd[75]: PADS: Service-Name: ''
<30>Jan 1 00:08:11 pppd[75]: PPP session is 1
<31>Jan 1 00:08:11 pppd[75]: using channel 2
<30>Jan 1 00:08:11 pppd[75]: Using interface ppp0
<29>Jan 1 00:08:11 pppd[75]: Connect: ppp0 <--> tap0
<28>Jan 1 00:08:11 pppd[75]: Couldn't increase MTU to 1500
<28>Jan 1 00:08:11 pppd[75]: Couldn't increase MRU to 1500
<31>Jan 1 00:08:11 pppd[75]: sent [LCP ConfReq id=0x2 <mru 1492>
<magic 0x8b152f1b>]
<29>Jan 1 00:08:11 pppd[75]: Modem hangup
<29>Jan 1 00:08:41 pppd[75]: Modem hangup
<29>Jan 1 00:08:41 pppd[75]: Connection terminated.


Thanks
Vanitha
[email]vanitha@agilis.st.com.sg[/email] wrote:[color=blue]
> Hi All,
>
> I am facing a problem in re-connecting the pppoe session in my target.
> When the link goes down inorder to make ppp reconnect i have enabled
> the "persist" option. I'm gettting the error "write: Bad file
> descriptor (9)". The re-connect happens successfuly sometimes. And
> eventhough i have specified the holdoff as 60, the re-connection
> happens around 2 hours after the connection goes down..Will appreciate
> your inputs on this.
>
> Here is my PPP options
> ======================
> asyncmap 0
> debug #PPPD will output connection information
> defaultroute #This interface will be a default gateway.
> nodeflate #Do not use compression
> nobsdcomp #Do not use bsd compression
> noccp #Disable CCP (Compression Control Protocol)
> noaccomp #Disable Address/Control compression
> nopcomp #Disable protocol field compres.
> novj #Disable Van Jacobson TCP/IP header compres.
> novjccomp #Disable the connection-ID compression
> nopredictor1 #Do not accept or agree to Predictor-1 compres.
> name new_miniv_1
> plugin rp-pppoe.so
> persist #re-open connection if ppp is down
> holdoff 60
> lcp-echo-interval 20
> lcp-echo-failure 15
>
>
> ___________________________________________________________
> <31>Jan 2 01:59:57 pppd[1490]: PADS: Service-Name: ''
>
> <30>Jan 2 01:59:57 pppd[1490]: PPP session is 531
>
> <31>Jan 2 01:59:57 pppd[1490]: using channel 21
>
> <30>Jan 2 01:59:57 pppd[1490]: Using interface ppp0
>
> <29>Jan 2 01:59:57 pppd[1490]: Connect: ppp0 <--> tap0
>
> <28>Jan 2 01:59:57 pppd[1490]: Couldn't increase MTU to 1500
>
> <28>Jan 2 01:59:57 pppd[1490]: Couldn't increase MRU to 1500
>
> <31>Jan 2 01:59:57 pppd[1490]: sent [LCP ConfReq id=0xc <magic
> 0x4cdbe8b4>]
>
> <29>Jan 2 01:59:57 pppd[1490]: Connection terminated.
>
> <31>Jan 2 02:00:00 pppd[1490]: sent [LCP ConfReq id=0xc <magic
> 0x4cdbe8b4>]
>
> <27>Jan 2 02:00:00 pppd[1490]: write: Bad file descriptor (9)
>
> <31>Jan 2 02:00:03 pppd[1490]: sent [LCP ConfReq id=0xc <magic
> 0x4cdbe8b4>]
>
> <27>Jan 2 02:00:03 pppd[1490]: write: Bad file descriptor (9)
>
> <31>Jan 2 02:00:06 pppd[1490]: sent [LCP ConfReq id=0xc <magic
> 0x4cdbe8b4>]
>
> <27>Jan 2 02:00:06 pppd[1490]: write: Bad file descriptor (9)[/color]

[email]vanitha@agilis.st.com.sg[/email] writes:[color=blue]
> <31>Jan 1 00:08:11 pppd[75]: sent [LCP ConfReq id=0x2 <mru 1492>
> <magic 0x8b152f1b>]
> <29>Jan 1 00:08:11 pppd[75]: Modem hangup[/color]

"Modem hangup" here means that the lower layer (PPPoE) went down.
[color=blue][color=green]
> > <31>Jan 2 02:00:00 pppd[1490]: sent [LCP ConfReq id=0xc <magic
> > 0x4cdbe8b4>]
> >
> > <27>Jan 2 02:00:00 pppd[1490]: write: Bad file descriptor (9)[/color][/color]

This also seems to mean that the lower layer has been disconnected.
I'd start by investigating what's wrong with PPPoE.

--
James Carlson, KISS Network <james.d.carlson@sun.com>
Sun Microsystems / 1 Network Drive 71.234W Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757 42.497N Fax +1 781 442 1677

I raised a bug request to the ppp.samba.org. And Paul has provided a
fix for this to the cvs version of pppd. Now its working fine for me.

Thanks