PPPD Connect Using Broadband Connection (problem) - PPP

This is a discussion on PPPD Connect Using Broadband Connection (problem) - PPP ; Hello, For the past 10+ days I have been pulling my hair out trying to configure pppd (on my SUSE Linux 7.3 system) using my cable modem broadband connection. I have to use SUSE linux 7.3 because im using linux ...

+ Reply to Thread
Results 1 to 13 of 13

Thread: PPPD Connect Using Broadband Connection (problem)

  1. PPPD Connect Using Broadband Connection (problem)

    Hello,

    For the past 10+ days I have been pulling my hair out trying to configure
    pppd (on my SUSE Linux 7.3 system) using my cable modem broadband
    connection. I have to use SUSE linux 7.3 because im using linux for Kylix
    development and thats the only OS that is currently supported.

    My situation is that I am trying to connect my home linux system to a remote
    office. The remote office is using Microsoft RAS server, hence I need pppd
    to connect to it. Before I began I configured my home Windows XP system
    such that it connected to the remote office using the VPN, this worked
    therefore
    acknowledging the connection parameters I was using were correct. I then
    proceeded to install the latest version of pppd on my linux box, and
    configured
    the pppd config files as follows:

    /etc/ppp/options.pptp
    ----------------------------------------------------------------------------
    ---------------
    debug
    noauth
    crtscts
    lock
    nodetach
    lcp-echo-interval 30
    lcp-echo-failure 4
    idle 600
    noipx
    nobsdcomp
    nodeflate
    debug
    mppe-40
    mppe-128
    mppe-stateless
    -proxyarp
    ----------------------------------------------------------------------------
    ---------------

    NOTE: If I set 'proxyarp' instead of '-proxyarp' then I get the following
    error when
    I start pppd:

    Cannot determine ethernet address for proxy

    Im not sure if this is a problem....


    /etc/ppp/tunnel
    ----------------------------------------------------------------------------
    ---------------
    pty "pptp xxx.xxx.150.2 --nolaunchpppd"
    name remotename PPTP
    #require-mschap-v2
    # require-mppe-128
    file /etc/ppp/options.pptp
    ipparam tunnel
    ----------------------------------------------------------------------------
    ---------------

    NOTE: I have to comment out 'require-mschap-v2' and 'require-mppe-128'
    otherwise
    when I try and start pppd I get the following error:

    pppd: In file /etc/ppp/peers/tunnel: unrecognized option 'require-mschap-v2'

    Is this a major problem?

    When I use the above configuration scripts and execute 'pppd call tunnel' to
    connect
    I notice that the system does seem to connect with the remote RAS server - I
    dont
    see any error messages, however I dont see a message stating connection
    established
    either... Provided below is the output that I receive from the pppd command
    when
    I attempt to connect:

    ----------------------------------------------------------------------------
    ---------------
    using channel 60
    Using interface ppp0
    Connect: ppp0 <--> /dev/pts/4
    sent [LCP ConfReq id=0x1 ]
    rcvd [LCP ConfReq id=0x0
    < 17 04 00
    45>]
    sent [LCP ConfRej id=0x0 < 17 04 00 45>]
    rcvd [LCP ConfAck id=0x1 ]
    rcvd [LCP ConfReq id=0x1
    ]
    sent [LCP ConfAck id=0x1
    ]
    sent [LCP EchoReq id=0x0 magic=0x8180c1ef]
    cbcp_lowerup
    want: 2
    rcvd [CHAP Challenge id=0x0 <17f0fce699daa9d2461fd66b35950441>, name =
    "BACKUP-SERVER"]
    sent [CHAP Response id=0x0
    660407210c3b228fee45700>, name = "xxxxxxxx"]
    rcvd [LCP EchoRep id=0x0 magic=0xb351445]
    rcvd [CHAP Success id=0x0 "S=4FE5BC46FA73E96FC1E8947AD669201302223914"]
    Remote message: S=4FE5BC46FA73E96FC1E8947AD669201302223914
    cbcp_open
    cbcp_req CONF_NO
    sent [CBCP Request id=0x1 < NoCallback>]
    rcvd [CBCP Request id=0x1 < NoCallback>]
    length: 2
    Callback: none
    cbcp_resp cb_type=2
    cbcp_resp CONF_NO
    sent [CBCP Response id=0x1 < NoCallback>]
    sent [IPCP ConfReq id=0x1 ]
    sent [CCP ConfReq id=0x1 ]
    rcvd [CBCP Ack id=0x1 < NoCallback>]
    rcvd [proto=0x803f] 01 03 00 1f 03 05 00 05 01 02 16 00 01 00 01 00 00 42 41
    43 4b 55 50 2d 53 45 52 56 45 52 00
    Unsupported protocol 'NETBIOS Framing Control Protocol' (0x803f) received
    sent [LCP ProtRej id=0x2 80 3f 01 03 00 1f 03 05 00 05 01 02 16 00 01 00 01
    00 00 42 41 43 4b 55 50 2d 53 45 52 56 45 52 ...]
    rcvd [IPXCP ConfReq id=0x4 ]
    Unsupported protocol 'Novell IPX Control Protocol' (0x802b) received
    sent [LCP ProtRej id=0x3 80 2b 01 04 00 12 01 06 38 da b8 a5 02 08 00 00 00
    00 00 01]
    rcvd [CCP ConfReq id=0x5 ]
    sent [CCP ConfNak id=0x5 ]
    rcvd [IPCP ConfReq id=0x6 ]
    sent [IPCP ConfAck id=0x6 ]
    rcvd [IPCP ConfRej id=0x1 ]
    sent [IPCP ConfReq id=0x2 ]
    rcvd [CCP ConfNak id=0x1 ]
    sent [CCP ConfReq id=0x2 ]
    rcvd [CCP ConfReq id=0x7 ]
    sent [CCP ConfAck id=0x7 ]
    rcvd [IPCP ConfNak id=0x2 ]
    sent [IPCP ConfReq id=0x3 ]
    rcvd [CCP ConfAck id=0x2 ]
    MPPE 128 bit, stateless compression enabled
    rcvd [IPCP ConfAck id=0x3 ]
    local IP address 192.168.2.57
    remote IP address 192.168.2.50
    Script /etc/ppp/ip-up started (pid 8731)
    Script /etc/ppp/ip-up finished (pid 8731), status = 0x0
    sent [LCP EchoReq id=0x1 magic=0x8180c1ef]
    rcvd [LCP EchoRep id=0x1 magic=0xb351445]
    sent [LCP EchoReq id=0x2 magic=0x8180c1ef]
    rcvd [LCP EchoRep id=0x2 magic=0xb351445]
    ....
    ....
    ....
    ----------------------------------------------------------------------------
    ---------------

    The interesting thing is that when I reach this point (thinking I am
    connected) I contact
    the remote office and I asked the administrator whether he could acknowledge
    that I
    was actually connected, he said that I had successfully connected. The
    problem is that
    once connected I can not ping any of the servers. The only IP addresses
    that I can
    ping are:

    192.168.2.57 <---- my assigned IP address
    192.168.2.50 <---- the remote MS RAS server

    If I try pinging any other server IP address I get 'From 172.18.6.61
    Destination Host
    Unreachable'. I am completely confused because I have absolutely no idea
    where the
    IP address 172.18.6.61 is coming from, neither do the administrators at the
    remote
    office.

    I would have assummed that there may have been a problem with the
    config/routing
    at the remote office, but I ruled this out because when I connect using my
    Windows XP
    machine I am able to ping all the servers successfully.

    Im not sure if this will help, but I saw a post in another newsgroup that
    explained
    I could monitor the ping requests going out using the command 'tcpdump -i
    ppp0'. I
    invoked this command and then issued the ping command, however the tcpdump
    returned absolutely nothing - however when I monitored eth0 ''tcpdump -i
    eth0' I
    noticed all the ping requests were appearing here. Im not sure what this
    means....

    Thats all the information I have, I would be *exceptionally* grateful if
    someone could
    possibly help me on this issue.

    Best regards

    Spencer
    (spencer@dvsmedia.co.uk)




  2. Re: PPPD Connect Using Broadband Connection (problem)

    Hi Spencer,

    you wrote:
    > The interesting thing is that when I reach this point (thinking I am
    > connected) I contact
    > the remote office and I asked the administrator whether he could acknowledge
    > that I
    > was actually connected, he said that I had successfully connected. The
    > problem is that
    > once connected I can not ping any of the servers. The only IP addresses
    > that I can
    > ping are:
    >
    > 192.168.2.57 <---- my assigned IP address
    > 192.168.2.50 <---- the remote MS RAS server


    You need to specify which networks or hosts are reachable via the PPtP
    connection. You could do this by
    1) Using the netmask option of pppd.
    2) Setting up static routes within the ip-up script. E.g. with
    /sbin/route add -net 192.168.2.0 netmask 255.255.255.0 "$1"

    Notice, if you need or want to set a route to the dialin servers public
    IP via the PPP device, you have to use the advanced routing features
    (iproute2 front-end) to force the packets of the PPtP-connection to use
    the normal internet connection.


    Michael

    --
    Linux@TekXpress
    http://www-users.rwth-aachen.de/Mich...kxp/tekxp.html

  3. Re: PPPD Connect Using Broadband Connection (problem)

    Spencer wrote:

    > For the past 10+ days I have been pulling my hair out trying to
    > configure pppd (on my SUSE Linux 7.3 system) using my cable modem
    > broadband connection. I have to use SUSE linux 7.3 because im
    > using linux for Kylix development and thats the only OS that is
    > currently supported.


    > My situation is that I am trying to connect my home linux system to
    > a remote office. The remote office is using Microsoft RAS server,
    > hence I need pppd to connect to it. Before I began I configured my
    > home Windows XP system such that it connected to the remote office
    > using the VPN, this worked therefore acknowledging the connection
    > parameters I was using were correct. I then proceeded to install
    > the latest version of pppd on my linux box, and configured the pppd
    > config files as follows:


    ....

    The log and pppd parameters that weren't accepted showed you aren't
    using a standard pppd 2.4.1 or even the pppd 2.4.2b3 beta available
    at ftp.samba.org. But the PPP link appeared to me to be negotiated okay.

    > The interesting thing is that when I reach this point (thinking I am
    > connected) I contact the remote office and I asked the administrator
    > whether he could acknowledge that I was actually connected, he
    > said that I had successfully connected. The problem is that once
    > connected I can not ping any of the servers. The only IP addresses
    > that I can ping are:


    > 192.168.2.57 <---- my assigned IP address
    > 192.168.2.50 <---- the remote MS RAS server


    You may just need to add the pppd option defaultroute.

    > If I try pinging any other server IP address I get 'From 172.18.6.61
    > Destination Host Unreachable'. I am completely confused because I
    > have absolutely no idea where the IP address 172.18.6.61 is coming
    > from, neither do the administrators at the remote office.


    It's a reserved (aka private) IP address that isn't forwarded by Internet
    routers. Try "traceroute 172.18.6.61" and see what shows up.

    > I would have assummed that there may have been a problem with the
    > config/routing at the remote office, but I ruled this out because
    > when I connect using my Windows XP machine I am able to ping all
    > the servers successfully.


    > Im not sure if this will help, but I saw a post in another newsgroup
    > that explained I could monitor the ping requests going out using the
    > command 'tcpdump -i ppp0'. I invoked this command and then issued
    > the ping command, however the tcpdump returned absolutely nothing -
    > however when I monitored eth0 ''tcpdump -i eth0' I noticed all the
    > ping requests were appearing here. Im not sure what this means....


    What does eth0 connect to? If it has a default route then the unknown
    IP address might well somehow be associated with the routing for it,
    and that would also be the reason the ping echo-requests appear on that
    interface.

    If you really need an eth0 default route and just want to talk to sites
    on a private network on the other side of the tunnel then add

    /sbin/route add -net aaa.bbb.ccc.ddd/xx $1

    to /etc/ppp/ip-up, where aaa.bbb.ccc.ddd/xx is the network on the other
    side, and forget about the pppd option defaultroute. The MS RAS server
    should know what to do with traffic to that network.

    --
    Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    PPP-Q&A links, downloads: http://ckite.no-ip.net/
    /* The wealth of a nation is created by the productive labor of its
    * citizens. */

  4. Re: PPPD Connect Using Broadband Connection (problem)

    Hi Bill,

    Thanks *alot* for your information - it provided me with some really
    good pointers
    and basically enabled me to solve the problem. The main problem was
    with the
    routing as you stated, I basically had to invoke the following line
    after setting up my
    ppp connection:

    # route add -net 192.168.0.0 gw 192.168.2.50 netmask 255.255.0.0

    Once the above line had been issued everything worked perfectly.

    I do still have one issue, and this is how to get the above command to
    be invoked
    when the pppd daemon is started. Basically I added the above line to
    the
    /etc/ppp/ip-up script - but this did not create the route. The ip-up
    script that
    my system uses is used for ip-up and ip-down, passed parameters
    determine
    what functionality to invoke. I added the above route command under the
    ip-up section - but it doesnt get invoked. Any ideas?

    Regardless of the above issue, many thanks for your exceptionally quick
    reply and help.

    Best regards

    Spence


    "Bill Unruh" wrote in message
    news:bjr6nt$tlp$1@string.physics.ubc.ca...
    > "Spencer" writes:
    >
    > The short answer is that this has nothing to do with pppd. It is a
    > routing problem.
    > You need to tell your system, using the route or the ip command which
    > addresses to use ppp for.
    >
    >
    >
    > ]Hello,
    >
    > ]For the past 10+ days I have been pulling my hair out trying to configure
    > ]pppd (on my SUSE Linux 7.3 system) using my cable modem broadband
    > ]connection. I have to use SUSE linux 7.3 because im using linux for

    Kylix
    > ]development and thats the only OS that is currently supported.
    >
    > ]My situation is that I am trying to connect my home linux system to a

    remote
    > ]office. The remote office is using Microsoft RAS server, hence I need

    pppd
    > ]to connect to it. Before I began I configured my home Windows XP system
    > ]such that it connected to the remote office using the VPN, this worked
    > ]therefore
    > ]acknowledging the connection parameters I was using were correct. I then
    > ]proceeded to install the latest version of pppd on my linux box, and
    > ]configured
    > ]the pppd config files as follows:
    >
    > ]/etc/ppp/options.pptp
    > ]-------------------------------------------------------------------------

    ---
    > ]---------------
    > ]debug
    > ]noauth
    > ]crtscts
    > ]lock
    > ]nodetach
    > ]lcp-echo-interval 30
    > ]lcp-echo-failure 4
    > ]idle 600
    > ]noipx
    > ]nobsdcomp
    > ]nodeflate
    > ]debug
    > ]mppe-40
    > ]mppe-128
    > ]mppe-stateless
    > ]-proxyarp
    >
    > get rid of all mention of proxyarp.
    >
    >
    > No defaultroute althought I suspect that you already have one on your
    > ethernet. Eitehr get rid of that and put defaultroute into this list, or
    > put a specific route into /etc/ppp/ip-up (and take it down in
    > /etc/ppp/ip-down)
    >
    >
    >
    >
    > ]-------------------------------------------------------------------------

    ---
    > ]---------------
    >
    > ]NOTE: If I set 'proxyarp' instead of '-proxyarp' then I get the

    following
    > ]error when
    > ]I start pppd:
    >
    > ]Cannot determine ethernet address for proxy
    >
    > ]Im not sure if this is a problem....
    >
    >
    > ]/etc/ppp/tunnel
    > ]-------------------------------------------------------------------------

    ---
    > ]---------------
    > ]pty "pptp xxx.xxx.150.2 --nolaunchpppd"
    > ]name remotename PPTP
    > ]#require-mschap-v2
    > ]# require-mppe-128
    > ]file /etc/ppp/options.pptp
    > ]ipparam tunnel
    > ]-------------------------------------------------------------------------

    ---
    > ]---------------
    >
    > ]NOTE: I have to comment out 'require-mschap-v2' and 'require-mppe-128'
    > ]otherwise
    > ]when I try and start pppd I get the following error:
    >
    > ]pppd: In file /etc/ppp/peers/tunnel: unrecognized option

    'require-mschap-v2'
    >
    > ]Is this a major problem?
    >
    > ]When I use the above configuration scripts and execute 'pppd call tunnel'

    to
    > ]connect
    > ]I notice that the system does seem to connect with the remote RAS

    server - I
    > ]dont
    > ]see any error messages, however I dont see a message stating connection
    > ]established
    > ]either... Provided below is the output that I receive from the pppd

    command
    > ]when
    > ]I attempt to connect:
    >
    > ]-------------------------------------------------------------------------

    ---
    > ]---------------
    > ]using channel 60
    > ]Using interface ppp0
    > ]Connect: ppp0 <--> /dev/pts/4
    > ]sent [LCP ConfReq id=0x1

    ]
    > ]rcvd [LCP ConfReq id=0x0


    > ] < 17 04 00
    > ]45>]
    > ]sent [LCP ConfRej id=0x0 < 17 04 00 45>]
    > ]rcvd [LCP ConfAck id=0x1

    ]
    > ]rcvd [LCP ConfReq id=0x1


    > ] ]
    > ]sent [LCP ConfAck id=0x1


    > ] ]
    > ]sent [LCP EchoReq id=0x0 magic=0x8180c1ef]
    > ]cbcp_lowerup
    > ]want: 2
    > ]rcvd [CHAP Challenge id=0x0 <17f0fce699daa9d2461fd66b35950441>, name =
    > ]"BACKUP-SERVER"]
    > ]sent [CHAP Response id=0x0
    > ]
    986
    > ]660407210c3b228fee45700>, name = "xxxxxxxx"]
    > ]rcvd [LCP EchoRep id=0x0 magic=0xb351445]
    > ]rcvd [CHAP Success id=0x0 "S=4FE5BC46FA73E96FC1E8947AD669201302223914"]
    > ]Remote message: S=4FE5BC46FA73E96FC1E8947AD669201302223914
    > ]cbcp_open
    > ]cbcp_req CONF_NO
    > ]sent [CBCP Request id=0x1 < NoCallback>]
    > ]rcvd [CBCP Request id=0x1 < NoCallback>]
    > ]length: 2
    > ]Callback: none
    > ]cbcp_resp cb_type=2
    > ]cbcp_resp CONF_NO
    > ]sent [CBCP Response id=0x1 < NoCallback>]
    > ]sent [IPCP ConfReq id=0x1 ]
    > ]sent [CCP ConfReq id=0x1 ]
    > ]rcvd [CBCP Ack id=0x1 < NoCallback>]
    > ]rcvd [proto=0x803f] 01 03 00 1f 03 05 00 05 01 02 16 00 01 00 01 00 00 42

    41
    > ]43 4b 55 50 2d 53 45 52 56 45 52 00
    > ]Unsupported protocol 'NETBIOS Framing Control Protocol' (0x803f) received
    > ]sent [LCP ProtRej id=0x2 80 3f 01 03 00 1f 03 05 00 05 01 02 16 00 01 00

    01
    > ]00 00 42 41 43 4b 55 50 2d 53 45 52 56 45 52 ...]
    > ]rcvd [IPXCP ConfReq id=0x4 ]
    > ]Unsupported protocol 'Novell IPX Control Protocol' (0x802b) received
    > ]sent [LCP ProtRej id=0x3 80 2b 01 04 00 12 01 06 38 da b8 a5 02 08 00 00

    00
    > ]00 00 01]
    > ]rcvd [CCP ConfReq id=0x5 ]
    > ]sent [CCP ConfNak id=0x5 ]
    > ]rcvd [IPCP ConfReq id=0x6 ]
    > ]sent [IPCP ConfAck id=0x6 ]
    > ]rcvd [IPCP ConfRej id=0x1 ]
    > ]sent [IPCP ConfReq id=0x2 ]
    > ]rcvd [CCP ConfNak id=0x1 ]
    > ]sent [CCP ConfReq id=0x2 ]
    > ]rcvd [CCP ConfReq id=0x7 ]
    > ]sent [CCP ConfAck id=0x7 ]
    > ]rcvd [IPCP ConfNak id=0x2 ]
    > ]sent [IPCP ConfReq id=0x3 ]
    > ]rcvd [CCP ConfAck id=0x2 ]
    > ]MPPE 128 bit, stateless compression enabled
    > ]rcvd [IPCP ConfAck id=0x3 ]
    > ]local IP address 192.168.2.57
    > ]remote IP address 192.168.2.50
    > ]Script /etc/ppp/ip-up started (pid 8731)
    > ]Script /etc/ppp/ip-up finished (pid 8731), status = 0x0
    > ]sent [LCP EchoReq id=0x1 magic=0x8180c1ef]
    > ]rcvd [LCP EchoRep id=0x1 magic=0xb351445]
    > ]sent [LCP EchoReq id=0x2 magic=0x8180c1ef]
    > ]rcvd [LCP EchoRep id=0x2 magic=0xb351445]
    > ]...
    > ]...
    > ]...
    > ]-------------------------------------------------------------------------

    ---
    > ]---------------
    >
    > ]The interesting thing is that when I reach this point (thinking I am
    > ]connected) I contact
    > ]the remote office and I asked the administrator whether he could

    acknowledge
    > ]that I
    > ]was actually connected, he said that I had successfully connected. The
    > ]problem is that
    > ]once connected I can not ping any of the servers. The only IP addresses
    > ]that I can
    > ]ping are:
    >
    > ]192.168.2.57 <---- my assigned IP address
    > ]192.168.2.50 <---- the remote MS RAS server
    >
    > ]If I try pinging any other server IP address I get 'From 172.18.6.61
    > ]Destination Host
    > ]Unreachable'. I am completely confused because I have absolutely no idea
    > ]where the
    > ]IP address 172.18.6.61 is coming from, neither do the administrators at

    the
    > ]remote
    > ]office.
    >
    > ]I would have assummed that there may have been a problem with the
    > ]config/routing
    > ]at the remote office, but I ruled this out because when I connect using

    my
    > ]Windows XP
    > ]machine I am able to ping all the servers successfully.
    >
    > ]Im not sure if this will help, but I saw a post in another newsgroup that
    > ]explained
    > ]I could monitor the ping requests going out using the command 'tcpdump -i
    > ]ppp0'. I
    > ]invoked this command and then issued the ping command, however the

    tcpdump
    > ]returned absolutely nothing - however when I monitored eth0 ''tcpdump -i
    > ]eth0' I
    > ]noticed all the ping requests were appearing here. Im not sure what this
    > ]means....
    >
    > ]Thats all the information I have, I would be *exceptionally* grateful if
    > ]someone could
    > ]possibly help me on this issue.
    >
    > ]Best regards
    >
    > ]Spencer
    > ](spencer@dvsmedia.co.uk)
    >
    >
    >




  5. Re: PPPD Connect Using Broadband Connection (problem)

    Michael Mueller wrote:
    > Hi Bill,


    > you wrote:
    >> ]You need to specify which networks or hosts are reachable via the PPtP
    >> ]connection. You could do this by
    >> ]1) Using the netmask option of pppd.
    >>
    >> The netmask option of pppd does nothing and is useless. ppp is point to
    >> point.


    > False atleast for the pppd version upto 2.4.1 on Linux 2.4+ since the
    > Linux kernel thinks different about this. If one does set a netmask for
    > a PPP interface - and pppd does so - this is reflected within the
    > automatic network route created.


    Huh? I don't think pppd does so. Here is a list of extra options
    (which are printed by the connection script and added to a common
    option set) for a recent pppd connection with pppd 2.4.2b3 to test
    that:

    netmask 255.255.255.0 novjccomp idle 300

    And here is my routing table for that connection:

    corncob:~# route -n
    Kernel IP routing table
    Destination Gateway Genmask Flags Metric Ref Use Iface
    207.44.129.90 0.0.0.0 255.255.255.255 UH 0 0 0 ppp0
    192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0 dummy0
    127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo
    0.0.0.0 207.44.129.90 0.0.0.0 UG 0 0 0 ppp0

    OTOH, maybe I'm missing something subtle?

    --
    Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    PPP-Q&A links, downloads: http://ckite.no-ip.net/

  6. Re: PPPD Connect Using Broadband Connection (problem)

    Michael Mueller writes:

    ]Hi Clifford,

    ]you wrote:
    ]> >> The netmask option of pppd does nothing and is useless. ppp is point to
    ]> >> point.
    ]>
    ]> > False atleast for the pppd version upto 2.4.1 on Linux 2.4+ since the
    ]>
    ]> Huh? I don't think pppd does so. Here is a list of extra options

    ]My stupid fault. Thought I can skip the final pppd test because I did
    ]not like to stop my connection. Only tried with ifconfig and looked
    ]through the code if netmask option still does anything.

    The ifconfig netmask option certainly does something. The pppd netmask
    option does not.

    From sys-linux.c

    /*
    * Set the netmask.
    * For recent kernels, force the netmask to 255.255.255.255.
    */
    if (kernel_version >= KVERSION(2,1,16))
    net_mask = ~0L;


  7. Re: PPPD Connect Using Broadband Connection (problem)

    "Spencer" writes:

    ]Hi Bill,

    ] Thanks *alot* for your information - it provided me with some really
    ]good pointers
    ] and basically enabled me to solve the problem. The main problem was
    ]with the
    ] routing as you stated, I basically had to invoke the following line
    ]after setting up my
    ] ppp connection:

    ] # route add -net 192.168.0.0 gw 192.168.2.50 netmask 255.255.0.0

    ] Once the above line had been issued everything worked perfectly.

    ] I do still have one issue, and this is how to get the above command to
    ]be invoked
    ] when the pppd daemon is started. Basically I added the above line to
    ]the
    ] /etc/ppp/ip-up script - but this did not create the route. The ip-up
    ]script that
    ] my system uses is used for ip-up and ip-down, passed parameters
    ]determine
    ] what functionality to invoke. I added the above route command under the
    ] ip-up section - but it doesnt get invoked. Any ideas?

    I have no idea what "my system" is. the standard pppd calls
    /etc/pp/ip-up as root once the ppp link is up. If your pppd does not,
    get rid of it and get a standard one, instead of using a broken one.

    If the script has been so munged up that it does not behave properly,
    fix it.

    Some distributions, you do not state what yours is, think that they are
    very clever, and instead break a things they understand poorly.
    Unfortunately it is up to the users to try to glue the pieces back
    together.


    ] Regardless of the above issue, many thanks for your exceptionally quick
    ] reply and help.

    ] Best regards

    ]Spence

  8. Re: PPPD Connect Using Broadband Connection (problem)

    Michael Mueller wrote:

    > Specifiy the absolute path to the route command. pppd does start ip-up
    > in an enviroment without the PATH variable set.


    I was recently "reminded" that there may be a default PATH for ip-up
    anyway, at least if you use the bash shell. It depends on the whim of
    whomever compiled your bash. Here it's

    PATH = /usr/gnu/bin:/usr/local/bin:/usr/ucb:/bin:/usr/bin:.

    which, of course, causes a program in /sbin, /usr/sbin, or /usr/local/sbin
    to fail when run from ip-up without the absolute path. :/

    --
    Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    PPP-Q&A links, downloads: http://ckite.no-ip.net/
    /* The generation of random numbers is too important to be left
    to chance. */

  9. Re: PPPD Connect Using Broadband Connection (problem)

    Clifford Kite writes:

    ]Michael Mueller wrote:

    ]> Specifiy the absolute path to the route command. pppd does start ip-up
    ]> in an enviroment without the PATH variable set.

    ]I was recently "reminded" that there may be a default PATH for ip-up
    ]anyway, at least if you use the bash shell. It depends on the whim of
    ]whomever compiled your bash. Here it's

    ]PATH = /usr/gnu/bin:/usr/local/bin:/usr/ucb:/bin:/usr/bin:.

    ??? I certainly have nothing like this.
    I dump the whole environment into a file when ip-up runs

    echo "`/usr/bin/env`">>/tmp/pppenv


    Here it is
    PPPD_PID=17022
    PPPLOGNAME=unruh
    SPEED=115200
    IPREMOTE=142.103.234.11
    IFNAME=ppp0
    DNS2=137.82.1.1
    DNS1=142.103.236.1
    PWD=/
    SHLVL=1
    DEVICE=/dev/ttyS0
    USEPEERDNS=1
    ORIG_UID=500
    IPLOCAL=142.103.234.12
    _=/usr/bin/env

    Note the absence of any PATH environment.
    (Mandrake 9.0 bash 2.0.5b-6mdk)


    Of course you may have a public bashrc file which is run whentever bash
    is started up.


    ]which, of course, causes a program in /sbin, /usr/sbin, or /usr/local/sbin
    ]to fail when run from ip-up without the absolute path. :/

    This is certainly true.

    So either set an explicit path in /etc/ppp/ip-up, or use full path names
    to executables.


    ]--
    ]Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    ]PPP-Q&A links, downloads: http://ckite.no-ip.net/
    ]/* The generation of random numbers is too important to be left
    ] to chance. */

  10. Re: PPPD Connect Using Broadband Connection (problem)

    Bill Unruh wrote:
    > Clifford Kite writes:


    > ]Michael Mueller wrote:


    > ]> Specifiy the absolute path to the route command. pppd does start ip-up
    > ]> in an enviroment without the PATH variable set.


    > ]I was recently "reminded" that there may be a default PATH for ip-up
    > ]anyway, at least if you use the bash shell. It depends on the whim of
    > ]whomever compiled your bash. Here it's


    > ]PATH = /usr/gnu/bin:/usr/local/bin:/usr/ucb:/bin:/usr/bin:.


    > ??? I certainly have nothing like this.
    > I dump the whole environment into a file when ip-up runs


    Notice that I carefully said *may* be a default PATH, and otherwise
    qualified the answer as well.

    BTW, I just checked the bash man pages and this is what I found:

    PATH The search path for commands. It is a colon-sepa-
    rated list of directories in which the shell looks
    for commands (see COMMAND EXECUTION below). The
    default path is system-dependent, and is set by the
    administrator who installs bash. A common value is
    ``/usr/gnu/bin:/usr/local/bin:/usr/ucb:/bin:/usr/bin:.''.

    That's a little misleading since that PATH is very likely compiled
    into bash by default. The bash executable that I use comes with my
    distribution, compiled by either GNU or my distributor.

    > Of course you may have a public bashrc file which is run whentever bash
    > is started up.


    It's /etc/profile here and I'd checked it, but ip-up apparently doesn't
    inherit the PATH from a user since pppd executes it.

    --
    Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    PPP-Q&A links, downloads: http://ckite.no-ip.net/

  11. Re: PPPD Connect Using Broadband Connection (problem)

    Clifford Kite writes:

    ]Bill Unruh wrote:
    ]> Clifford Kite writes:

    ]> ]Michael Mueller wrote:


    ]> ??? I certainly have nothing like this.
    ]> I dump the whole environment into a file when ip-up runs

    ]Notice that I carefully said *may* be a default PATH, and otherwise
    ]qualified the answer as well.

    ]BTW, I just checked the bash man pages and this is what I found:

    ] PATH The search path for commands. It is a colon-sepa-
    ] rated list of directories in which the shell looks
    ] for commands (see COMMAND EXECUTION below). The
    ] default path is system-dependent, and is set by the
    ] administrator who installs bash. A common value is
    ] ``/usr/gnu/bin:/usr/local/bin:/usr/ucb:/bin:/usr/bin:.''.

    ]That's a little misleading since that PATH is very likely compiled
    ]into bash by default. The bash executable that I use comes with my
    ]distribution, compiled by either GNU or my distributor.

    Try doing
    string /bin/bash|grep gnu
    to see if the path is actually in bash. I suspect that it is obtained
    from somewhere. Note that since pppd runs as root, and in particular
    starts up ip-up as root, it is the root user who should give bash its
    default path, probably set in /root/.bash_profile as a non-login user.


    ]> Of course you may have a public bashrc file which is run whentever bash
    ]> is started up.

    ]It's /etc/profile here and I'd checked it, but ip-up apparently doesn't
    ]inherit the PATH from a user since pppd executes it.

    pppd is running as root at the time it starts ip-up.


  12. Re: PPPD Connect Using Broadband Connection (problem)

    Bill Unruh wrote:
    > Clifford Kite writes:
    > ]Bill Unruh wrote:
    > ]> Clifford Kite writes:


    > ]Notice that I carefully said *may* be a default PATH, and otherwise
    > ]qualified the answer as well.


    > ]BTW, I just checked the bash man pages and this is what I found:


    > ] PATH The search path for commands. It is a colon-sepa-
    > ] rated list of directories in which the shell looks
    > ] for commands (see COMMAND EXECUTION below). The
    > ] default path is system-dependent, and is set by the
    > ] administrator who installs bash. A common value is
    > ] ``/usr/gnu/bin:/usr/local/bin:/usr/ucb:/bin:/usr/bin:.''.


    > ]That's a little misleading since that PATH is very likely compiled
    > ]into bash by default. The bash executable that I use comes with my
    > ]distribution, compiled by either GNU or my distributor.


    > Try doing
    > string /bin/bash|grep gnu
    > to see if the path is actually in bash. I suspect that it is obtained
    > from somewhere. Note that since pppd runs as root, and in particular
    > starts up ip-up as root, it is the root user who should give bash its
    > default path, probably set in /root/.bash_profile as a non-login user.


    ~/shop/d4x-2.03$ strings /bin/bash|grep gnu
    rl_gnu_readline_p
    i386-slackware-linux-gnu
    /usr/gnu/bin:/usr/local/bin:/usr/ucb:/bin:/usr/bin:.
    ..
    ..
    ..

    > ]> Of course you may have a public bashrc file which is run whentever bash
    > ]> is started up.


    > ]It's /etc/profile here and I'd checked it, but ip-up apparently doesn't
    > ]inherit the PATH from a user since pppd executes it.


    > pppd is running as root at the time it starts ip-up.


    Okay, as a test I put
    echo $PATH > /root/ip-up.path
    just after the "# !/bin/sh" at the beginning of my ip-up. This is
    what was in ip-up.path after a PPP link came up:
    /usr/gnu/bin:/usr/local/bin:/usr/ucb:/bin:/usr/bin:.

    So ip-up doesn't inherit root's PATH, it inherits the bash shell's
    built-in PATH. And pppd doesn't have a built-in path, doing
    strings /usr/sbin/pppd|egrep 'bin:|bin$'
    should convince you of that.

    --
    Clifford Kite Email: "echo xvgr_yvahk-ccc@ri1.arg|rot13"
    PPP-Q&A links, downloads: http://ckite.no-ip.net/
    /* The wealth of a nation is created by the productive labor of its
    * citizens. */

  13. Re: PPPD Connect Using Broadband Connection (problem)

    In comp.protocols.ppp you write:

    ]Bill Unruh wrote:
    ]> Try doing
    ]> string /bin/bash|grep gnu
    ]> to see if the path is actually in bash. I suspect that it is obtained
    ]> from somewhere. Note that since pppd runs as root, and in particular
    ]> starts up ip-up as root, it is the root user who should give bash its
    ]> default path, probably set in /root/.bash_profile as a non-login user.

    ]~/shop/d4x-2.03$ strings /bin/bash|grep gnu
    ]rl_gnu_readline_p
    ]i386-slackware-linux-gnu
    ]/usr/gnu/bin:/usr/local/bin:/usr/ucb:/bin:/usr/bin:.

    OK, I agree, your bash has that path compiled in. Mine apparently has
    /usr/local/bin:/bin:/usr/bin

    Certainly not any sbin.




    ]So ip-up doesn't inherit root's PATH, it inherits the bash shell's
    ]built-in PATH. And pppd doesn't have a built-in path, doing
    ] strings /usr/sbin/pppd|egrep 'bin:|bin$'
    ]should convince you of that.

    I do not doubt that. I am just surprised that when pppd runs ip-up, it
    does not read /etc/profile or /root/.bash_login or /root/.bashrc.