Discussion:
LCP DOWN not notified by pppd when LCP Conf request is received during LCP open state
(too old to reply)
s***@gmail.com
2016-04-24 06:20:56 UTC
Permalink
Hi

Have a query regarding pppd behaviour when LCP Conf request is received during LCP open state

pppd version used is 2.4.5

HW : Linux(running pppd 2.4.5) ========== Cisco MWR switch terminating ppp.


pppd Logs
============

LCP negotiation is started.

sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x5a064c51> <pcomp> <accomp> <mrru 1500> <endpoint [IP:14.20.20.1]>]
rcvd [LCP ConfReq id=0xca <magic 0xe0148a35> <pcomp> <accomp> <mrru 1500> <endpoint [local:4d.4c.50.50.50.34]>]
sent [LCP ConfAck id=0xca <magic 0xe0148a35> <pcomp> <accomp> <mrru 1500> <endpoint [local:4d.4c.50.50.50.34]>]
sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x5a064c51> <pcomp> <accomp> <mrru 1500> <endpoint [IP:14.20.20.1]>]
rcvd [LCP ConfAck id=0x1 <asyncmap 0x0> <magic 0x5a064c51> <pcomp> <accomp> <mrru 1500> <endpoint [IP:14.20.20.1]>]

==> LCP is in open state now

sent [LCP EchoReq id=0x0 magic=0x5a064c51]


==> IPCP negotiation is started.

sent [IPCP ConfReq id=0x1 <addr 14.20.20.1>]
rcvd [IPCP ConfAck id=0x1 <addr 14.20.20.1>]

==> During this phase LCP Conf Request is received from the peer side (Cisco MWR)

rcvd [LCP ConfReq id=0xcb <magic 0xe0148a35> <pcomp> <accomp> <mrru 1500> <endpoint [local:4d.4c.50.50.50.34]>]
sent [LCP ConfReq id=0x2 <asyncmap 0x0> <magic 0xdde160a5> <pcomp> <accomp> <mrru 1500> <endpoint [IP:14.20.20.1]>]
sent [LCP ConfAck id=0xcb <magic 0xe0148a35> <pcomp> <accomp> <mrru 1500> <endpoint [local:4d.4c.50.50.50.34]>]
rcvd [LCP ConfAck id=0x2 <asyncmap 0x0> <magic 0xdde160a5> <pcomp> <accomp> <mrru 1500> <endpoint [IP:14.20.20.1]>]

==> LCP is again renegotiated and goes to open state.

But LCP was already in open state and pppd did not notify upper layers about this re negotiation.

Can anybody clarify , Is this behaviour expected OR somehow pppd is notifying upper layers but it is to be verified in some other log ?

Regards,
Sreeraj
Manikandaprabu Kuppannan
2017-11-10 13:48:49 UTC
Permalink
Hi Sreeraj,

In my opinion, this is expected behaviour.
Not needed to inform upper layer about renegotiation.
I assume your question is not about renegotiation in a loop.

Thanks,
ManikandaPrabu

Loading...