There are several possibilities.
1. "Bridging" was compiled into the kernel (see the appropriate question below).
2. Broadcasts from the interfaces ware being forwarded by ISDN (see the appropriate question below).
3. If IP connections are still open with the line is disconnected and IP addresses are dynamically assigned, then the disaster is inevitable. Then a new connection is started to bring down the open IP connections, which fails because the IP address is now different. The line is hung up, but the IP connections are still open, the line is dialed again, and so on... (see the appropriate question below).
Maybe the kernel option "bridging" was mistakenly activated?
Michael Pieper
michael@nexus1.tng.oche.de
wrote on 10 Nov 1996:
ARP requests or broadcasts? You should run ifconfig with the options -arp and -broadcast to keep from opening connections in this way.
Gernot Zander
hifi@scorpio.in-berlin.de
wrote on 18 Jan 1997:
You can bring the interface "down" then back "up". When you do this, it will try to dial out. But if you have removed the outgoing telephone number, then "no outgoing number..." appears in the syslog, and as soon as the interface is "up", all connections will be closed.
Gernot Zander
hifi@scorpio.in-berlin.de
wrote on 10 Jan 1997:
I delete the telephone number of the interface (or set an invalid one - Ed.) Then I can see right away from the complaints in the syslog whether a process wants to send packets out to the world.
Sascha Ottolski
sascha@alzhimer.isdn.cs.tu-berlin.de
added on
28 Oct 1996:
For me, what is happening is what you want: the ISDN system is "down", Netscape immediately shows an error message a la: "the Server doesn't have a DNS entry" or something similar. You probably have to delete the route so that this will happen. I do it like this:
/sbin/route del default /sbin/isdnctrl system off /sbin/ifconfig ippp0 down
/sbin/isdnctrl system on /sbin/ifconfig ippp0 up /sbin/route add $GATE-IP dev ippp0 /sbin/route add default ippp0
In the next version of isdnlog (2.52) it will be able to work as a watchdog and prevent such dialouts (with a reboot, if necessary).
Karsten Keil
keil@temic-ech.spacenet.de
wrote on 11 Feb 1997:
I'm guessing, that with the status enquiry (in Switzerland - Ed.) you
simply want to make sure that when the user side has crashed, the connection
is broken. This is in addition to the Layer 2 monitoring and is not
totally senseless, since with many cards/end devices the ISAC is run in
auto mode and therefore a crash would keep the connection open.
However, i4l runs the ISAC in nonauto mode, meaning that when interrupts are no longer being process, the connection is broken after a maximum of about 1/2 a minute. This is not the reason for using nonauto mode, but this is a safety feature ;-), but doesn't mean that the charge unit disaster is impossible.
Dirk Lutzebaeck
lutzeb@wadk-berlin.de
wrote on 5 Nov 1996:
Unfortunately, I don't know of any syncPPP encapsulation patch for tcpdump. If you use ipppd, then your only chance is to turn off one daemon after the other and see if things have finally quieted down. Likely candidates are named, sendmail, and also smbd (Samba) that are likely to open connections.(On tcpdump see the appropriate question under "General" in the "Configuration" section. On named and sendmail see the following questions - Ed.) Christoph Trautwein
trautw@fzi.de
added on 5 Nov 1996:
I too was only able to find this out by killing suspicious processes.
More information on the search for these processes and how to make
them quit can be found at:
http://www.fzi.de/sim/people/trautw/i4l/index.html
Herbert Rosmanith
herp@wildsau.idv.uni-linz.ac.at
added on 24 Nov 1996:
Try to find out which lookup triggers the connection with "isdnctrl verbose 3". Then a message should appear in the kernel message queue (visible with "dmesg") a'la: OPEN: 141.76.60.54 - 193.171.67.253 TCP, port: 1686 - 540 In this example, our computer is trying to pick up mail on port 540 (UUCP over TCP/IP over ISDN).(only the triggering packet will be logged - Ed.)
Stefan Luethje
luethje@sl-gw.lake.de
wrote further on 27 Nov 1996:
Another tip. There are a lot of daemons on the Linux side that broadcasts on all interfaces. This leads to frequent autodials. In this case you can redirected the broadcast address to the dummy0 interface. It's not clean, but it works.See all the last question in this section.
Stefan Luethje
luethje@sl-gw.lake.de
wrote on 27 Nov 1996:
If in Wintel the name server of your provider is given, and Windows 3.11/95 is started, then it has to talk to the name server (only Bill Gates knows why).
Jens Ey
jens@jeyhh.shnet.org
wrote on 29 Nov 1996:
Turn on debug level 1 in named and look at the logfile in /var/tmp.
In my case, I found regular DNS requests from Windows machines. The problem was that names like WORKGROUP.domain.de were requested, i.e. names that the DNS could not know. I'm assuming that the Windows machine was looking for its master browser or a domain controller.
This was causing me so many problems for my Internet connection with Linux in a LAN that I installed an external terminal adapter, set up a proxy server, and set up diald that only DNS requests from the Linux machine were allowed to be carried out. Then the connections are established only when they are actually needed. The (caching) local DNS is only used after the connection has been established.
Eike Stepper
isdn@esc-net.de
wrote on 30 Nov 1996:
Why not simply set the "Use DNS for Windows Names Resolution" (or similar) to No? Then it should be quiet, at least it is for me.
First you have to get sendmail to no long open any DNS connections. You need to activate the following features: "nodns", "nocanonify".
If you have a smarthost, you need to make sure that this name does not call the name server. You can either set it directly as an IP address, or add the name to /etc/hosts (/etc/host.conf should then contain "order hosts bind")
You should set all non-local mailers as "expensive" ("define(SMTP_MAILER_FLAGS, e)"), and then forbid sendmail with "define(`confCON_EXPENSIVE', `True')" from automatically connection to expensive mailers. The call to sendmail should no longer include a time for the "-q" option (e.g. only "-bd -os -q"). "-os" means that all mail will be queued (which won't prevent local mail from being delivered immediately). The only catch is that when booting, mail that might still be in the queue will be sent by sendmail, even though the network is not yet up. Therefore, when booting you should remove all mail from /var/mqueue before starting sendmail, and then return it once sendmail has been started.
Mail to expensive mailers will now only be send with the explicit call "sendmail -q".
Andreas Glahn
andreas@tao.westfalen.de
wrote on 31 Jan 1997:
I had this problems too. Then when starting the samba daemon I gave
it the internal IP address I use here at home. Since then a samba
request is no longer sent to default, but stays here.
Take a look at the configuration with netstat and tcpdump. With tcpdump you can quickly find out to which IP address samba is trying to connect.
My internal Linux computer has, e.g.: 192.168.99.99
My Win95 computer: n 192.168.99.88
On the Linux computer I started samba with:
nmdb -S -B 192.168.99.255 -I 192.168.99.99
Most likely in the preferences a non-local home page has been listed. Only a home page that Netscape is able to load immediately (e.g. "file://localhost/xxx") won't cause an immediate dialout. Alternatively you can also set up a cache daemon that saves pages that are often needed.
A proxy should not cause a dial out, even when the complete name is
entered. Only when a new proxy is given does Netscape do a DNS
lookup (and in this special case cause a dialout.
However, on 17 Mar 97 Steffan Henke
henker@Informatik.Uni-Bremen.DE
wrote:
Unfortunately reality has caught up with us. I've heard that Netscape now in version.4.02 really does establish a connection...
Michael Stiller
michael@toyland.ping.de
wrote on 23 Oct 1996:
Tip for ftp:
ftp://ftp.gwdg.de/pub/misc/isdn/linux/isdn4linux-gwdg
There is the patch: "tcpdump-3.0.4-1-isdn.dif.gz"
and the rest is at:
/pub/linux/mirrors/funet/PEOPLE/Linus/net-source/tools/tcpdump-3.0.4-1.tar.gz
You might need to hack some, depending on the name of your ISDN interface (mine is bri0). By default, it recognizes only isdn* and isdnY* as interface names.
Henning Schmiedehausen
henning@pong.iconsult.com
further wrote on
30 Oct 1996:
After finding the patch from Eberhard Moenkeberg at ftp.gwdg.de cannot dump cisco HDLC, I made my own patch for tcpdump-3.0.4 that asks the interface which encapsulation it used and sets itself accordingly. The patch is against a tcpdump-3.0.4-1.tar.gz distribution, for example at
ftp://ftp.funet.fi/pub/Linux/PEOPLE/Linus/tools
This patch recognizes rawIP, ISDN-IP and CISCO-HDLC and can dump these packets.(The patch was attached to the message - it should be easy to find in the mailing list archive - Ed.)
Sascha Ottolski
sascha@alzhimer.isdn.cs.tu-berlin.de
gave the following
tip on 5 Nov 1996:
This is a isdn4k-utils-2.0/tcpdump-3.0.3-isdn.diff ! It work, if one makes some changes: In the file tcpdump-3.0.3-isdn/libpcap-0.0/pcap-linux.c after patching you find the following: else if (strncmp("ppp", device, 3) == 0) Either you name your ppp devices pppX instead of ipppX, or change this line, e.g. else if (strncmp("ippp", device, 4) == 0) ^^^^ ^^ Then tcpdump will also recognize syncPPP. At least it does for me.