I have installed RedHat 6.1 and all the latest rpms and updates, but PPP refuses to work for me. Below is a copy of what I get when I "tail - f /var/log/messages". I 'x'ed' out sensitive information. I need to know if this bug can be fixed. Nov 25 15:59:26 wuntang gnome-name-server[529]: name server starting Nov 25 16:00:05 wuntang ifup-ppp: pppd started for ppp0 on /dev/ttyS2 at 57600 Nov 25 16:00:05 wuntang modprobe: can't locate module char-major-108 Nov 25 16:00:05 wuntang kernel: CSLIP: code copyright 1989 Regents of the University of California Nov 25 16:00:05 wuntang kernel: PPP: version 2.3.7 (demand dialling) Nov 25 16:00:05 wuntang kernel: PPP line discipline registered. Nov 25 16:00:05 wuntang kernel: registered device ppp0 Nov 25 16:00:05 wuntang pppd[587]: pppd 2.3.10 started by root, uid 0 Nov 25 16:00:07 wuntang WvDial: WvDial: Internet dialer version 1.40 Nov 25 16:00:07 wuntang WvDial: Initializing modem. Nov 25 16:00:07 wuntang WvDial: Sending: ATZ Nov 25 16:00:07 wuntang WvDial: ATZ Nov 25 16:00:07 wuntang WvDial: OK Nov 25 16:00:07 wuntang WvDial: Sending: ATM1L3 Nov 25 16:00:07 wuntang WvDial: ATM1L3 Nov 25 16:00:07 wuntang WvDial: OK Nov 25 16:00:07 wuntang WvDial: Modem initialized. Nov 25 16:00:07 wuntang WvDial: Sending: ATDT xxx-xxxx Nov 25 16:00:07 wuntang WvDial: Waiting for carrier. Nov 25 16:00:07 wuntang WvDial: ATDT xxx-xxxx Nov 25 16:00:35 wuntang WvDial: CONNECT 57600 Nov 25 16:00:35 wuntang WvDial: Carrier detected. Waiting for prompt. Nov 25 16:00:36 wuntang WvDial: ** i48-25.pdx.du ** Nov 25 16:00:36 wuntang WvDial: i48-25.pdx.du Login: Nov 25 16:00:36 wuntang WvDial: Looks like a login prompt. Nov 25 16:00:36 wuntang WvDial: Sending: xxxxxxx Nov 25 16:00:36 wuntang WvDial: xxxxxxx Nov 25 16:00:36 wuntang WvDial: Password: Nov 25 16:00:36 wuntang WvDial: Looks like a password prompt. Nov 25 16:00:36 wuntang WvDial: Sending: (password) Nov 25 16:00:36 wuntang WvDial: enabled Nov 25 16:00:36 wuntang WvDial: Entering PPP Mode. Nov 25 16:00:36 wuntang WvDial: IP address is xxx.xx.x.xx Nov 25 16:00:36 wuntang WvDial: MTU is 1524. Nov 25 16:00:36 wuntang WvDial: Looks like a welcome message. Nov 25 16:00:36 wuntang pppd[587]: Serial connection established. Nov 25 16:00:36 wuntang pppd[587]: Using interface ppp0 Nov 25 16:00:36 wuntang pppd[587]: Connect: ppp0 <--> /dev/ttyS2 Nov 25 16:01:07 wuntang pppd[587]: LCP: timeout sending Config-Requests Nov 25 16:01:07 wuntang pppd[587]: Connection terminated. Nov 25 16:01:07 wuntang pppd[587]: Hangup (SIGHUP) Nov 25 16:01:07 wuntang pppd[587]: Exit. My ifcfg-ppp0 looks like this: DEVICE=ppp0 NAME=xxxxxxx WVDIALSECT=xxxxxxx MODEMPORT=/dev/ttyS2 LINESPEED=57600 PAPNAME=xxxxxx USERCTL=true ONBOOT=no DNS1=xxx.xxx.xxx.xx DNS2=xxx.xxx.xxx.xx PPPOPTIONS=noauth My wvdial.conf looks like this: [Modem0] Modem = /dev/ttyS2 Baud = 57600 Init1 = ATZ Init2 = ATQ0 V1 E1 S0=0 SetVolume = 1 Dial Command = ATDT Init4 = ATM1L3 [Dialer Defaults] Modem = /dev/ttyS2 Baud = 57600 Init1 = ATZ SetVolume = 1 Dial Command = ATDT Init4 = ATM1L3 [Dialer Teleport] Username = xxxxxx Password = xxxxxx Modem = /dev/ttyS2 Phone = 972-2800 Baud = 57600 Stupid mode = 0 Inherits = Dialer Defaults My ISP doesn't use PAP; when I use Windows NT I just send a clear text password. Please help, I love Linux, but I NEED the internet connection. Please note that I can actually dial into my ISP and see PPP starting on the ISP side, but I can't get it started on my side. The dialer keeps redialing and dieing at the Config-Requests point!!. I have the latest updates including those put out on 11/23/1999.
This does not appear to have anything to do with PAP. This timeout sending config requests means that the PPP daemon is unable to negotiate a working PPP connection with the remote side. There are all sorts of reasons this could be going wrong. Try adding the line DEBUG=yes to your ifcfg-ppp0 file and see what additional information it produces.
Added DEBUG=yes to ifcfg-ppp0 file, and I am getting the following: Unsupported protocol (0x80fd) received
Looks like protocol 0x80fd is unsupported and the other side insists that it work in order to have a connection. I don't know what protocol 0x80fd is, though, so I don't have much idea what to tell you at this point.
The unsupported protocol problem only showed once and did not appear ever again. Now, do I need anything in the /etc/resolv.conf, /etc/hosts.deny, hosts.allow, /etc/networks files??
80fd, it turns out, is compression control protocol, which should be supported. Not sure why you got the error message, but glad to know it isn't a continuing problem. If you have all the latest errata updates, /etc/resolv.conf will be updated automatically. You should not need to set up /etc/sysconfig/network or the /etc/host.* files to get your PPP connection working. You are still getting "LCP: timeout sending Config-Requests", right?
Yep, I am still getting the time-outs. I am going to delete evrything from my host.* files and see what happens. I also have the following entry in my /etc/networks file loopback 127.0.0.0 localnet 0.0.0.0 Should I have this? I took this from an old config. manual
I really don't think those are related. I have nothing in /etc/networks and I haven't touched the hosts.* files, but they really shouldn't have anything to do with this. Have you tried lowering your LINESPEED=57600 to LINESPEED=38400 to see if that helps?
I haven't tried lowering my line speed. Does that have anythong to do with 'pppd' at all or is it just related to the connection? That's about the only recommended thing I haven't done so far!
That changes the speed that ppp sets the modem to before connecting. Please try it.
I lowered my modem speed, LINESPEED=38400, but the same thing happened. At this point, I am really baffled. I am just gonna go back to RedHat 6.0. What's the significance of LCP?
LCP is the link control protocol; it's one of the base protocols that pppd uses to set up other PPP protocols. It isn't totally broken -- if it were, then no one would be able to connect via PPP at all. One other thing that has caused this problem in the past is a modem with misconfigured interrupts. If Linux is misinformed about what interrupt the modem is on, it will default to polled mode, and that goes slowly -- too slowly for LCP to work. Are your kernel messages at boot time about what IRQ it is using for ttyS2 correct? Is it a different IRQ from ttyS0? (It needs to be different.)
That could be it. When I use RP3 to detect the modem, it looks like it finds something on /dev/ttyS0 (says determining modem type, etc,etc) then comes back and says no modem detected. Also, I have an /etc/rc.d/rc.serial with the following command: setserial /dev/ttyS2 auto_irq autoconfig skip_test uart 16450 Do I need this command? Another thing, how do I set up the irq for /dev/ttyS2 to be different from /dev/ttyS0 's irq? I know pppd is not broken at all; I have a Compaq Netelligent plug and play modem.
Oooh, a 16450 uart can't handle PPP very well at speeds that are at all high, particularly if you have IDE hard drives. plug and play modems are often unsupported "winmodems"; you are sure that this is not the case? Try using the isapnp program to set the IRQ for that modem to something that is not used anywhere on the system, and particularly neither 3 nor 4 (the common ttyS{0,1} IRQs. In any case, this does not sound like a bug, it sounds like a hardware configuration issue, so I've marked it as not a bug.
How do I set up ppp to work with MSCHAP? Do I need to recompile the PPP packages or is MSCHAP compiled in? I found out my ISP uses an NT server and I might require MSCHAP.
Our ppp package does not currently include MSCHAP support because MSCHAP requires DES encryption, which historically we have not been able to legally export. The README.MSCHAP80 file in the ppp source package includes information on rebuilding pppd with MSCHAP support. However, before you go to the trouble, I'd suggest asking your ISP to enable normal CHAP authentication. If they refuse and you have other options, you may wish to take your business elsewhere.