I've just tried rp-l2tp without much luck. >From the Linux end I get ============================ 0.000 dgram_take_from_wire() -> type=SCCRQ, tid=0, sid=0, Nr=0, Ns=0 0.000 tunnel_new() -> 33418/0 0.000 33418/30: Peer host name is 'a7800113.lansa.co.uk' 0.002 auth_gen_response(secret=Qwertest1) -> 9c777aeb096f89662565e99dfd965cb8 0.002 tunnel(33418/30) state idle -> wait-ctl-conn 0.002 tunnel_enqueue_dgram(33418/30, SCCRP) rws=8 cwnd=1 ssthresh=8 outstanding=0 0.003 xmit_queued(33418/30): rws=8 cwnd=1 ssthresh=8 outstanding=0 0.003 dgram_send_to_wire() -> type=SCCRP, tid=30, sid=0, Nr=1, Ns=0 0.004 loop in xmit_queued(33418/30): rws=8 cwnd=1 ssthresh=8 outstanding=1 1.014 xmit_queued(33418/30): rws=8 cwnd=1 ssthresh=1 outstanding=0 1.014 dgram_send_to_wire() -> type=SCCRP, tid=30, sid=0, Nr=1, Ns=0 1.015 loop in xmit_queued(33418/30): rws=8 cwnd=1 ssthresh=1 outstanding=1 3.024 xmit_queued(33418/30): rws=8 cwnd=1 ssthresh=1 outstanding=0 3.024 dgram_send_to_wire() -> type=SCCRP, tid=30, sid=0, Nr=1, Ns=0 3.025 loop in xmit_queued(33418/30): rws=8 cwnd=1 ssthresh=1 outstanding=1 7.034 xmit_queued(33418/30): rws=8 cwnd=1 ssthresh=1 outstanding=0 7.034 dgram_send_to_wire() -> type=SCCRP, tid=30, sid=0, Nr=1, Ns=0 7.035 loop in xmit_queued(33418/30): rws=8 cwnd=1 ssthresh=1 outstanding=1 15.044 xmit_queued(33418/30): rws=8 cwnd=1 ssthresh=1 outstanding=0 15.044 dgram_send_to_wire() -> type=SCCRP, tid=30, sid=0, Nr=1, Ns=0 15.045 loop in xmit_queued(33418/30): rws=8 cwnd=1 ssthresh=1 outstanding=1 Error: Too many retransmissions on tunnel (33418/30); closing down 23.054 tunnel_free(33418/30) ======================================= my /etc/l2tp/l2tp.conf looking like: ====================================== # Global section (by default, we start in global mode) global
# Load handlers load-handler "sync-pppd.so" load-handler "cmd.so" # Bind address listen-port 1701 # Configure the sync-pppd handler. You MUST have a "section sync-pppd" line # even if you don't set any options. section sync-pppd lns-pppd-opts "user a7800113.lansa.co.uk name linuxfw2.lansa.co.uk passive -detach auth require-chap 10.45.10.123:10.44.10.12 3 lcp-echo-interval 30 lcp-echo-failure 6" lac-pppd-opts "user example name example noipdefault ipcp-accept-local ipcp-accept-remote lcp-echo-interval 30 lcp-echo-failu re 6" # Peer section section peer peer 195.157.58.111 secret Qwertest1 port 1701 lac-handler sync-pppd lns-handler sync-pppd hide-avps no =============================== and the W2k complaining that the modem reported and error Would you care to give me an example config that works? Rgds Dom -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] Sent: 07 January 2003 14:21 To: [EMAIL PROTECTED] Subject: Re: loopback error Hello, i also had this problem with win 2000. The only real (stable) way to solve this was using the rp-l2tpd-0.3 implementation. I have seen, that l2tpd echoes back the send telegrams and win 2000 dislikes this... Ulrich -----Urspr�ngliche Nachricht----- Von: <[EMAIL PROTECTED]> An: <[EMAIL PROTECTED]> Gesendet: Dienstag, 7. Januar 2003 14:59 Betreff: loopback error > Hi Folks ! > > I'm currently setting up a VPN access server for the organization I'm > working for and my softs of choice are (off course :) L2TPD/Frees/Wan, > on a Linux box. > VPN clients are meant to be Windows 2000/XP and I got some recurrent > connection problem with a "loopback error" detected on the windows side. > After several times (3-5...), the connection succeeds. > I checked the mailing list archive but no good solutions seems to have > araised. I've read on PPP docs that during LCP phase bad magic number > negociation may cause a loopback error. > Connection errors still persist with PPP option "connect-delay" set; > "nomagic" option seems to get things worse than before... > Does anybody have any idea ? Is there any Windows registry value to hack > to get things working better (maybe in > KEY_LOCAL_MACHINE\SYSTEM\CCS\Services\Rasman\PPP) ? > > > Thanx for your help ! > > > -- > BD >
