And after shutting down the eth0 or whatever.
Do a new dialout, if you have a result like
$ cat /etc/resolv.conf
nameserver 207.172.3.8
nameserver 207.172.3.9
you should then have functioning DNS navigaton
MarvS
Marvin Stodolsky wrote:
> US Robotics 56k model 3CP5610A must be a controller chip modem, using
> the standard serial.o driver under 2.4.n kernels
> and under 2.6.n modules in:
> # ls /lib/modules/2.6.3-test1/kernel/drivers/serial/
> 8250.ko 8250_pci.ko serial_core.ko serial_cs.ko
> ------------
> serial_core and either 8250.ko or 8250_pci
> See what is loaded with # lsmod
> The failure of ping tests means that naviagation is NOT working.
> This can be due to eth0 or some other COMM mode being active. Check with
>
> $ /sbin/ifconfig
> lo Link encap:Local Loopback inet addr:127.0.0.1
> Mask:255.0.0.0
> UP LOOPBACK RUNNING MTU:16436 Metric:1
> RX packets:210 errors:0 dropped:0 overruns:0 frame:0
> TX packets:210 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:0 RX bytes:11216 (10.9 KiB)
> TX bytes:11216 (10.9 KiB)
>
> ppp0 Link encap:Point-to-Point Protocol inet
> addr:209.122.217.155 P-t-P:208.59.249.11 Mask:255.255.255.255
> UP POINTOPOINT RUNNING NOARP MULTICAST MTU:1500 Metric:1
> RX packets:571 errors:0 dropped:0 overruns:0 frame:0
> TX packets:620 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:3 RX bytes:327714 (320.0
> KiB) TX bytes:106428 (103.9 KiB)
>
> The lo block is beneficial loopback testing.
> If you have any other block, say eth0 , do
> # ifdown eth0
> and/or # ifconfig eth0 down
>
> The 1st Attachment explains.
> Also please download
> http://linmodems.technion.ac.il/packages/scanModem.gz
> under Linux
> gunzip scanModem.gz
> chmod +x scanmodem
> ./scanModem
> will run diagnostics and output advice.
> Send me the output ModemData.txt but NOT the others
>
> MarvS
> scanModem maintainer
>
>
>
> Thomas Beresford wrote:
>
>> I don't know which driver my modem uses but it's an US Robotics 56k
>> model 3CP5610A and it used to work perfectly with the old kernel.
>> I´m pretty sure that I'm connecting to my ISP. I'm using wvdial to
>> connect and I can see all the messages (sending login, password,
>> authentication, starting pppd, etc). It doesn´t seem to be a DNS
>> problem since I´ve installed the testing ppp package and the ping
>> still doesn´t work, nor the numerical address nor the named address.
>> What could it be?
>>
>> PS: I noticed that the new kernel is loading the ipv6 module, could
>> it be the problem?
>>
>>
>> ----- Original Message -----
>> From: Marvin Stodolsky <marvstod@rcn.com>
>> Date: Sat, 03 Apr 2004 18:42:48 -0500
>> To: Thomas Beresford <number18@linuxmail.org>
>> Subject: Re: Dial up problems when upgrading 2.4 to 2.6 kernel
>>
>>
>>
>>> Which driver does your modem use?
>>> Are you sure you actually CONNECT to the ISP
>>> Compare:
>>> # ping 30.57.4.70
>>> 64 bytes from 130.57.4.70: icmp_seq=0 ttl=48 time=239.2 ms
>>> 64 bytes from 130.57.4.70: icmp_seq=1 ttl=48 time=240.0 ms
>>> 64 bytes from 130.57.4.70: icmp_seq=2 ttl=48 time=230.0 ms
>>> 64 bytes from 130.57.4.70: icmp_seq=3 ttl=48 time=230.0 ms
>>> with the named address
>>> # ping novell.com
>>> If the numerical address works but not the named address,
>>> Then it is only a DNS problem.
>>>
>>> You may need some PPP support files from the testing distribution
>>> for DNS to work
>>> Try rebooting under your old kernel.
>>> Within /etc/apt/sources.list, make a duplicate of all lines.
>>> Within the 2nd set, replace "stable" with "testing", for example
>>> deb http://ftp.lug.udel.edu/debian/ stable main non-free contrib
>>> goes to ====
>>> deb http://ftp.lug.udel.edu/debian/ testing main non-free contrib
>>> Then
>>> apt-get update
>>> apt-get -s install ppp
>>> (where -s is stimulate)
>>> If nothing dangerous is shown
>>> apt-get install ppp
>>>
>>> MarvS
>>>
>>>
>>>
>>>
>>>
>>> Thomas Beresford wrote:
>>>
>>>
>>>
>>>> Hello fellas,
>>>>
>>>> I upgraded my debian system kernel from 2.4 stable to 2.6 testing.
>>>> But now my dialup connection doesn't work anymore! I get to connect
>>>> to my ISP, but i can't navigate. Could someone help me, please?
>>>>
>>>>
>>>
>>
>>
>>
>>
>>
>>
>
>------------------------------------------------------------------------
>
>After Bruno Magnani, magnani.bruno AT alitalia.it
>tersely reported on resolving problems of coincident use of ethernet and ppp connections,
>I asked him to please expand the report for inclusion in the Lucent DOCs/
>He returned a very nice report, which with very little editing is that below.
>The issue is much more generic than usage with the Lucent modem,
>so it is appropriate to also provide it broadly to Discuss@Linmodems.org
>
>Thanks again Bruno.
>
>============================================================================================
>
>Premise
>-----------
>These notes report about running a Lucent Microelectronics LT winmodem
>driver on a Compaq Armada E500 that is mostly used to connect to a
>Company's LAN through a network card.
>
>Goals:
> - set up a handy configuration for dialup ISP accounts
> - avoid touching existing network configuration files
>Technical context :
> - distribution Red Hat 7.2 (Enigma)
> - Kernel 2.4.7-10
>Dialers:
> - wvdial 1.41
> - KDE kppp 2.0.8 (KDE 2.2-11, Qt 2.3.1)
>Evidence of applicability to a different context: NONE
>Please check the output of the following commands to verify your settings:
> # uname -a
> # wvdial --version
> # kppp --version
>
>Startup Assumption
>--------------------------
>I installed the source package ltmodem-6.00b9, and assume the process
>of compiling, installing and autoloading the new drivers on your machine went ok.
>
>After that I now have: the new drivers installed at:
> /lib/modules/2.4.7-10/kernel/drivers/char/lt_modem.o
> /lib/modules/2.4.7-10/kernel/drivers/char/lt_serial.o ;
>the device node /dev/ttyLT0 (as the port to be referred in dialup scripts);
>the symbolic link /dev/modem --> /dev/ttyLT0 ;
>an updated /etc/modules.conf as well as
>an updated module configuration database.
>
>Setting up name resolution with wvdial
>----------------------------------------
>The configuration file for this dialer is /etc/wvdial.conf, you can read through the wvdial man pages
>and the documentation file ../DOCs/wvdial.txt for setting it up (remember to create the symbolic link
>"ln -sf /dev/ttyLT0 /dev/ttyS14" as suggested).
>
>Here is what I get when attempting a connection:
>-----------------------------------------------
> /home/bruno> wvdial libero
> --> WvDial: Internet dialer version 1.41
> --> Initializing modem.
> --> Sending: ATZ
> ATZ
> OK
> --> Sending: ATM1L3
> ATM1L3
> OK
> --> Modem initialized.
> --> Sending: ATDT 00688531010
> --> Waiting for carrier.
> ATDT 00688531010
> CONNECT 48000 V42bis
> --> Carrier detected. Waiting for prompt.
> Username:
> --> Looks like a login prompt.
> --> Sending: xxxxxxx@libero.it
> xxxxxxx@libero.it
> Password:
> --> Looks like a password prompt.
> --> Sending: (password)
> ~Entering PPP mode.}-}*Async interface address is unnumbered (Loopback0)}-}*Your IP
> address is 151.24.220.124. MTU is 1500 bytes}-}*}-}*jO~
> --> Looks like a welcome message.
> --> Starting pppd at Fri Feb 22 17:08:31 2002
>
>And the corresponding output of "tail -f /var/log/messages"
>------------------------------------------------------------
> kernel: CSLIP: code copyright 1989 Regents of the University of California
> kernel: PPP generic driver version 2.4.1
> pppd[1955]: pppd 2.4.1 started by root, uid 0
> pppd[1955]: Using interface ppp0
> pppd[1955]: Connect: ppp0 <--> /dev/ttyLT0
> kernel: PPP BSD Compression module registered
> kernel: PPP Deflate Compression module registered
> N4206624 pppd[1955]: local IP address 151.24.220.124
> N4206624 pppd[1955]: remote IP address 151.5.160.151
> pppd[1955]: primary DNS address 193.70.192.25
> pppd[1955]: secondary DNS address 193.70.152.25 (*)
>
>(*) The last two lines are shown if you add the line "usepeerdns" in /etc/ppp/options
>as suggested by Emin Liman when I raised a DNS issue in discuss@linmodem.org.
>
>However when connected:
>----------------------
>I did not get name server resolution after this, that is
>I could ping using dotted quad notation IP addresses,
>but domain name server (DNS) addressing failed.
>
>Even adding the two DNS addresses to /etc/ppp/resolv.conf was unsuccessfull.
>The only solution I found was to prepare a specific DNS configuration file for
>the ISP account (e.g. resolv.conf.libero) and write a simple script to substitute
>/etc/resolv.conf before the connection, then restoring the usual network configuration
>upon exiting.
>
>//Comment from MarvS
>If ethernet is not "ifdown eth0" before attempting ppp without the further configuration
>work described below, there are frequently Netscape crashes and/or freezing of Xwindows
>//
>
>Setting up name server resolution with kppp
>--------------------------------------------
>Setting up kppp is easily done by starting the tool and clicking on setup,
>the various steps are well documented in the manual that is accessible through the ? button.
>This process creates a configuration file in ~/.kde/share/config/kppprc
>listing all the accounts information.
>If you provide your ISP's domain name, DNS addresses and set the option:
> "Disable Existing DNS Seervers during Connection"
>in the DNS section of kppp setup ~/.kde/share/config/kppprc
>will include lines such as
>
> DNS=212.17.192.216,212.17.192.56
> .....
> Domain=infinito.it
> ExDNSDisabled=1
>
>Given this, DNS resolution (magically) works with out any need to touch /etc/resolv.conf.
>
>However, when I use kppp I get a weird message:
>
> ....modprobe: modprobe: Can't locate module ppp0
> ... kernel: CSLIP: code copyright 1989 Regents of the University of California
> ...
>
>in /var/log/messages. This can be (inelegantly) removed by inserting "alias ppp0 off" in
>/etc/modules.conf.
>
>Endnote:
>----------------------------------
>Please consider that starting a serial connection on a public network
>while logged onto a LAN raises severe security issues and should be avoided.
>
>
>
>
>
>