Linux-Networking Digest #318, Volume #12 Sat, 21 Aug 99 22:13:38 EDT
Contents:
Mail Server Problem ("sago")
Re: Is this possible? ("Tick Notes")
What's this: IP_MASQ:reverse ICMP: failed checksum from (Drake Christensen)
3 COM Etherlink III port setup from BNC to TP ("darkfader")
Re: IP Masqerading - Win95 client wont work!! (Ken Booth)
NCD X-Terminal: Asking Linux for XDMCP session (Neil Zanella)
Re: Linux Cookbook Project Officially begins! (Allen)
Ethernet card vs 2.0.32 What am I misssing here? (Sam Walton)
IP MASQ and RoadRunner ([EMAIL PROTECTED])
IP Masqerading - Win95 client wont work!! ("Sunil P. Khatri")
socket problem ("Rajat Monga")
Re: Problem with ppp - syslog supplied (Clifford Kite)
Re: Linux & Lan (Raymonds Doetjes)
loading NIC module automatically ("Christopher Lu")
Re: mgetty question (Raymonds Doetjes)
Connection woa's ("J. Flechtner")
Truncated-ip (Wisquatuk)
Re: Home network problems revisited ("Eric L. Damron")
Apache won't server pages to Windows Clients! ("Eric L. Damron")
nfs mounting problem (James Ranson)
Removing Telnet Access for a User? ("Matt")
Symlink not working from FTP ("Matt")
----------------------------------------------------------------------------
From: "sago" <[EMAIL PROTECTED]>
Subject: Mail Server Problem
Date: Sat, 21 Aug 1999 19:18:46 -0400
I have just setup a linuxbox as a mailserver. I can send mail to the
internet but cant receive. If I send a mail from some other host , say
Hotmail to this mail server of mine it comes back saying :
__________________________________________________________
Hi. This is the qmail-send program at hotmail.com.
I'm afraid I wasn't able to deliver your message to the following addresses.
This is a permanent error; I've given up. Sorry it didn't work out.
<[EMAIL PROTECTED]>:
205.142.28.17 does not like recipient.
Remote host said: 553 <[EMAIL PROTECTED]>... this site does not relay
Giving up on 205.142.28.17.
__________________________________________________________
205.142.28.17 above is my service provider and hotbuilder.com is my
domain with a permenent IP.
If I do a MX query in nslookup :
MX with preference 10 is pointed to my server mail.hotbuilder.com
MX with preference 30 is pointed to my provider.
If I do /usr/lib/sendmail -t -v from some other host and send a mail to my
server ([EMAIL PROTECTED] ) then it says
[EMAIL PROTECTED] Deferred : Connection refused by
mail.hotbuilder.com
Can somebody tell me whats happening and whats the solution.
Thanx
Sagolsem
------------------------------
From: "Tick Notes" <[EMAIL PROTECTED]>
Subject: Re: Is this possible?
Date: Sat, 21 Aug 1999 15:54:03 -0800
Bill Thank you
Bill Pitz wrote in message <[EMAIL PROTECTED]>...
>tick <[EMAIL PROTECTED]> gave us the interesting posting of:
>> I've got dsl and one ip address with mydomain.com registered to it.
Would
>> it be possible to register a my2nddomain.com with the same ip address but
>> then redirect it to an internal web server.
>
>Yes, it's possible. Register the domain to the same IP address. Setup
>a name-based virtual host in apache, then have it redirect to another
>port number that forwards to your internal web server. It's not going
>to be really easy, but it is definitely possible to do.
>
>-Bill
>--
>Bill Pitz bill at svn.net
>Silicon Valley North, Inc. www.svn.net
>Internet and World Wide Web Services (707) 781-9999
------------------------------
From: [EMAIL PROTECTED] (Drake Christensen)
Subject: What's this: IP_MASQ:reverse ICMP: failed checksum from
Date: Sat, 21 Aug 1999 21:53:17 GMT
I'm running RH 6.0 with IPChains.
IP_MASQ:reverse ICMP: failed checksum from 216.xxx.xxx.xxx
I've been seeing this error message fairly regularly, say every couple of
days. They tend to show up in pairs, from two different IP addresses.
About the time these show up on my Linux box my Windows machine starts
having trouble seeing anything on the net.
I've tracerouted a couple of the IP addresses and they end up in
Cinncinati. One said something about an adsl router. So far they've all
been in 216.xxx.xxx.xxx.
Is this someone probing my network just for grins? (I don't have
anything worth stealing.) Or is this just a corrupt packet?
Drake Christensen
------------------------------
From: "darkfader" <[EMAIL PROTECTED]>
Subject: 3 COM Etherlink III port setup from BNC to TP
Date: Fri, 20 Aug 1999 02:03:02 +0200
Hi!
I'm using Suse Linux 6.0 in a system I set up today, and it works great
apart from one VERY big problem:
The drvier automatically selected the NIC's BNC port, though I connected it
on TP!
Is there anyone, who can tell me how and where to tell the driver what port
to use???
Thanks,
Florian
------------------------------
From: Ken Booth <[EMAIL PROTECTED]>
Subject: Re: IP Masqerading - Win95 client wont work!!
Date: Sat, 21 Aug 1999 17:20:30 -0500
Sunil,
If you post the results from the following commands, we may be able to
help you better:
netstat -rn
and
ipfwadm -F -ln
--
Regards, Ken
I AM.
"Sunil P. Khatri" wrote:
> I installed IP masquerading, and my gateway and (linux) clients
> work just fine with the configuration and strong ipfwadm rulesets
> that are described in the latest IP-Masquerading HOWTO (v. 1.77).
> My gateway runs kernel 2.0.30.
>
> However, when I put a Win95 client on the local network, with the
> settings as described in the IP-Masquerading HOWTO (I set the IP
> address, netmask, gateway address and DNS server addresses in the
> TCP_IP->NIC properties) the win95 client can only ping the other
> machines on the local network, but cant ping any outside machine.
>
> Any idea why this is the case?
>
> My conjecture is that the IP-Masquerading HOWTO was written for win95
> and win98 clients that support the ICS (internet connection sharing)
> option in networking. My win95 client does not have this module. Does
> this make sense?
>
> Muchas thanks
> Sunil
------------------------------
From: Neil Zanella <[EMAIL PROTECTED]>
Subject: NCD X-Terminal: Asking Linux for XDMCP session
Date: Sat, 21 Aug 1999 19:20:40 -0230
Hello,
I would like to configure a Red Hat Linux 5.2 box so that
it gives an NCD X Terminal an XDMCP session. The NCD X Terminal
is already configured to accept XDMCP sessions from some other
hosts on the TCP/IP based network.
Thanks for all your help,
Neil Zanella
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Allen)
Crossposted-To: comp.os.linux.advocacy,comp.os.linux.help,comp.os.linux.setup
Subject: Re: Linux Cookbook Project Officially begins!
Date: 22 Aug 1999 00:11:17 GMT
On Fri, 20 Aug 1999 23:41:44 -0700, "W.A. Scheer" <[EMAIL PROTECTED]>
wrote:
...snip...
>
>I need the following:
>
> 1) Volunteers to edit, test and/or review copy.
I'll be happy to volunteer for testing, as I can't get much more
bald from testing the existing docs for RH5.2 and RH6.0, and I can
certainly give an accurate point of view from an ignorant newbie
standpoint, 'cause I have no UNIX or Linux experience save the bruising
I've gotten this past 8 month's trying to get my Linux machine to go
online. I'm not shy of or afraid of the command line interface, just
not familiar with the unix or bash (non-dos) versions of it. I would
hope that this is also covered too, since my extensive dos experience
suggests that there *are* times that one still needs to go to the
command line to get something done, or done well/efficiently.
Allen
(email addy; furnished upon request?)
------------------------------
From: Sam Walton <[EMAIL PROTECTED]>
Subject: Ethernet card vs 2.0.32 What am I misssing here?
Date: Sat, 21 Aug 1999 18:41:10 -0400
Reply-To: [EMAIL PROTECTED]
If someone would bear with me, I'll try to give complete info.
B/G: Power mac user exploring Linux as an AppletalkIP server and SMB.
Machine is a 586 clone with 24m ram. I bought "Linux in 24 hours" and a
D-Link 220e 10bt card. The Linux kernel is 2.0.32. I tried making sure
NE2000 was installed when installing for the first time. From my
research, I'm not surprised the installer failed to see it.
Visited:
http://metalab.unc.edu/mdw/HOWTO/Ethernet-HOWTO-1.html
and determined that I must turn off PlugnPlay on the card in DOS and
have put in the base address of 300 in I/O (tried putting in 0x300,
too) and the interrupt or something like that at 3 and failing that,
10.
I go to kerneld, kernel Configurator and add an eth0 interface with
"ne" module. I hit the "restart kerneld"
(what's odd is that after attempting ping to test it, I go back to
Kernel Configurator and the eth0 interface is gone. So I've had plenty
of practice of putting in these numbers).
Then I go to Network Configurator and add the eth0 interface with an IP
address of, oh say, 127.0.0.2 and a netmask of 255.0.0.0. I have tried
both "none" and "dhcp" in the Interface configuration protocal setting.
(I really wish I had a web page to help me figure out what I should be
doing in Network Configurator. I've looked in the redhat pages for that
term and got air biscuits.)
(also, I want to try IPMasq if I ever get this figured out and it seems
I've got to figure this Configurator out if that's going have a prayer
to work)
I then activate that setting and save. I go to the shell window and
type "ping -r 168.127.68.12" which happens to be the TCP address my Mac
is using for its DHCP setting. I have also tried setting a static
address on this Mac when I'm offline to eliminate that as a variable.
Air biscuits. I use Contol-Z to make it stop, is that cool?
I go to my Mac using OT tools to ping the 127.0.0.2 address. Air
biscuits again.
I have checked all the relevant parts of
http://metalab.unc.edu/mdw/HOWTO/Ethernet-HOWTO-3.html
to get some ideas. Nothing has worked.
I typed "dmesg | more" and so no reference to eth0 or IRQ 10. Netatalk
and Samba seem to come up, though.
Then I typed "cat /proc/net/dev" and got
65 0 0 0 0 65 0 0 0 0
whatever that means.
I'm probably done something really simply wrong but I'm big enough to
admit if I've done a bonehead thing.
BTW, can an ISA card fit in a PCI slot?
Looking for smarter people than I, sam
------------------------------
From: [EMAIL PROTECTED]
Subject: IP MASQ and RoadRunner
Date: Sun, 22 Aug 1999 01:39:37 GMT
I am moving to RoadRunner and their cable modem. Had the nice man install all that
last weekend
on a Windows 98 machine and it works fine. Got the linux (redhat 5.2) set up and it
sees both
the local LAN and the internet. The windows machines see the internet for browsing but
I seem
to NOT be able to get my mail or use newsgroups from the machines that are connecting
via ip
masq. This all works when the Windows 98 machine is directly connected.
Anyone have any ideas? I would be happy to supply any other information that would be
deemed
useful. Thanks in advance for any help.
tom
------------------------------
From: "Sunil P. Khatri" <[EMAIL PROTECTED]>
Subject: IP Masqerading - Win95 client wont work!!
Date: Sat, 21 Aug 1999 21:23:38 GMT
I installed IP masquerading, and my gateway and (linux) clients
work just fine with the configuration and strong ipfwadm rulesets
that are described in the latest IP-Masquerading HOWTO (v. 1.77).
My gateway runs kernel 2.0.30.
However, when I put a Win95 client on the local network, with the
settings as described in the IP-Masquerading HOWTO (I set the IP
address, netmask, gateway address and DNS server addresses in the
TCP_IP->NIC properties) the win95 client can only ping the other
machines on the local network, but cant ping any outside machine.
Any idea why this is the case?
My conjecture is that the IP-Masquerading HOWTO was written for win95
and win98 clients that support the ICS (internet connection sharing)
option in networking. My win95 client does not have this module. Does
this make sense?
Muchas thanks
Sunil
------------------------------
From: "Rajat Monga" <[EMAIL PROTECTED]>
Subject: socket problem
Date: Sat, 21 Aug 1999 15:41:19 -0700
i am writing a chat server on linux with a client on windows. when i close
the connection using closesocket on the windows side, while a thread on
linux server is blocked on a receive, it does not come out with an error.
Sometimes even the first write to the socket gets blocked even though the
length of the data is just 4 bytes.
any suggestions are welcome.
Thanks
Rajat
------------------------------
From: kite@NoSpam.%inetport.com (Clifford Kite)
Subject: Re: Problem with ppp - syslog supplied
Date: 21 Aug 1999 19:38:11 -0500
Jo Knight ([EMAIL PROTECTED]) wrote:
: Ive just recompiled my kernel so i could get my soundcard to work, i made
: sure that i included ppp support. My problem is that when i connect to the
: net i cannot get to any sites, so i checked the syslog, here is the output:
: **** This was when i connect with wvdial*****
: Aug 19 18:07:14 sensei pppd[9164]: PAP authentication failed
: Aug 19 18:10:57 sensei pppd[9181]: unrecognized option '-v'
: Aug 19 18:15:20 sensei pppd[9200]: PAP authentication failed
Pppd has a -v option set somewhere and that's not a valid pppd option.
Find it and remove it. I can't tell you where wvdial (whatever that is)
puts it. Pppd itself can keep options in /etc/ppp/options.
: ****This was connecting thru kppp*******
: Aug 19 18:26:28 sensei pppd[9251]: Cannot determine ethernet address for
: proxy ARP
: So my question is does anyone know how i can fix these problems?
The proxy arp problem can be solved by removing the pppd proxyarp option
whereever kppp keeps them. Don't know kppp either but, again, for pppd
the file /etc/ppp/options is a common one for options.
--
Clifford Kite <kite@inet%port.com> Not a guru. (tm)
/* Governments should be changed like diapers - often and for the
* same reason. */
------------------------------
From: Raymonds Doetjes <[EMAIL PROTECTED]>
Subject: Re: Linux & Lan
Date: Sun, 22 Aug 1999 01:12:48 +0200
During installation you will have to supply these figures.
If tou did'nt back then, you can still fdo it with your setup software from
the distro
(WIth SuSE it's called yast)
and with RedHat its called control-panel (you have to run X for the
control-panel to work)
Or try linuxconf (with any other distro I guess) and otherwise use ifconfig to
setup and ip address to a nic.
Raymond
"Rudolf Traunm�ller" wrote:
> I've got 2 Computers & 2 Ethernet Ne 2000 network cards. Now I've got
> WinProxy on my Win98 Pc to dial in to the Internet with various network Pcs
> and only one connection to the net. Winproxy works.
> Now I try to set up my Linux Pc. The only thing I would need is a TcpIp
> Connection between my Linux and my Win98 Pc, so does anybody know how i
> configure my network card and other things to install such a TcpIp Lan??
>
> Thanks
------------------------------
From: "Christopher Lu" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,alt.os.linux.mandrake
Subject: loading NIC module automatically
Date: Sun, 22 Aug 1999 00:59:11 GMT
I'm using D-Link DFE530TX on my Mandrake 6. Currently I have to load the
module for the NIC (via-rhine) everytime I boot up and log on. Is there a
way to have the module load automatically when I boot? Thanks in advance.
------------------------------
From: Raymonds Doetjes <[EMAIL PROTECTED]>
Subject: Re: mgetty question
Date: Sun, 22 Aug 1999 01:29:38 +0200
Add to your mgetty command line -n #
# is the number of rings before modem will pick up the line/
Raymond
Leonard Hardy wrote:
> I have tried to use xringd to dial my machine with a specific ring
> sequence and have a connection started to my ISP, this works sometimes
> but is unreliable.
>
> I have heard that mgetty can do the same thing?? Does anyone have sample
> configuration files that work?
>
> Thanks, Len
------------------------------
From: "J. Flechtner" <[EMAIL PROTECTED]>
Subject: Connection woa's
Date: Sat, 21 Aug 1999 19:26:06 -0700
Hello all!
Version: Redhat 5.2
Modem: USR 56k data/fax v90
Recently installed a USR 56k from a USR 33.6 and have not been able to
connect since. Dip.script is fine as well as /etc/resolv.conf. Jumper
settings on the board are correct. Just noticed during LILO that it now
cannot find the USR**********(string of numbers). Where do I go to correct
this and tell it what it should be looking for? Thanks in advance!
J. Flechtner
------------------------------
From: Wisquatuk <[EMAIL PROTECTED]>
Subject: Truncated-ip
Date: Sun, 22 Aug 1999 00:54:37 GMT
Okay. So I upgraded from an old tcpdump (I forget the version) to
3.4... of course, now I have to get used to a slightly different UI,
but that's beyond the scope of this post. :) But anyway, now suddenly
I'm getting truncated-ip reports *everywhere*. Nothing else has
changed -- kernel version (still 2.2.11), etc. Is this a TCP bug? Or
is it a network bug that only the latest versions pick up? In any
case, I'd really like to fix it, cos until I do, I'm gonna hafta
downgrade to some older version (thank goodness for RPMs) when I find
an urgent need to use it -- it's kinda unreadable as-is:
20:45:44.230691 eth0 > truncated-ip - 1137 bytes missing!furball.1474
> ch2oco.bellglobal.com.www: P 2379:3608(1229) ack 51881 win
31856 <nop,nop,timestamp 16406524 263569465> (DF)
20:45:44.410691 eth0 < ch2oco.bellglobal.com.www > furball.1474:
. 51881:51881(0) ack 3608 win 65160 <nop,nop,timestamp
263569619 16406524> (DF)
20:45:45.040691 eth0 < truncated-ip - 1356 bytes
missing!ch2oco.bellglobal.com.www > furball.1474:
. 51881:53329(1448) ack 3608 win 65160 <nop,nop,timestamp
263569680 16406524> (DF)
20:45:45.050691 eth0 < truncated-ip - 1356 bytes
missing!ch2oco.bellglobal.com.www > furball.1474: P
53329:54777(1448) ack 3608 win 65160 <nop,nop,timestamp
263569680 16406524> (DF)
20:45:45.050691 eth0 > furball.1474 > ch2oco.bellglobal.com.www:
. 3608:3608(0) ack 54777 win 31856 <nop,nop,timestamp 16406606
263569680> (DF)
20:45:45.080691 eth0 < truncated-ip - 1356 bytes
missing!ch2oco.bellglobal.com.www > furball.1474:
. 54777:56225(1448) ack 3608 win 65160 <nop,nop,timestamp
263569685 16406606> (DF)
20:45:45.090691 eth0 < truncated-ip - 465 bytes
missing!ch2oco.bellglobal.com.www > furball.1474: P
56225:56782(557) ack 3608 win 65160 <nop,nop,timestamp
263569685 16406606> (DF)
(And that's just the half of it -- the linebreaks and tabs I put in
just for the sake of Usenet readability. It's pretty much impossible
to get any kinda real info from a glance at the tcpdump output, like
it used to be.)
If this has been posted before, I apologise, but I tried my best to
find answers on the web, and failed. However, at least the answer can
come quick and easy. :)
--
- Wisquatuk (name[1..4]@netrover.com to email)
=====BEGIN GEEK CODE BLOCK=====
Version: 3.12
GCS/CC/M d-(--) a--- C++(+++)>$ UL++++>$ P+++ L+++ E>++ W-(+>++) N+++
!o>++ K- w--- O- M- !V PS++(+++) PE- Y+ PGP+++@ t+@ 5 X+++@ R+ tv b+
DI+@ D+ G>+++ e- h!(++) !r z
======END GEEK CODE BLOCK======
------------------------------
From: "Eric L. Damron" <[EMAIL PROTECTED]>
Crossposted-To:
alt.linux,alt.solaris.x86,comp.os.ms-windows.networking,microsoft.public.win98.networking
Subject: Re: Home network problems revisited
Date: Sun, 22 Aug 1999 00:50:34 GMT
> Just a shot in the dark..
When you use the uplink port the hub will disable one of the ports. (Usually
the one right next to it.)
Try moving the disfunctional PC's cable to a new port.
Make sure TCP/IP is working by pinging the loopback.
ping 127.0.0.1
If that works (TCP/IP is working.) try pinging your own network card.
ping PCs_IP
If that works then try pinging the other PC.
ping WORKING_PCs_IP
If that works try pinging your router's IP.
This will help you narrow the problem.
------------------------------
From: "Eric L. Damron" <[EMAIL PROTECTED]>
Subject: Apache won't server pages to Windows Clients!
Date: Sun, 22 Aug 1999 01:05:05 GMT
My Apache web server is showing good taste by snubbing my Microsoft
clients! It servers the pages fine if my PCs are booting into Linux.
As humorous as I find this, I really need to find the configuration
error that is causing this problem.
Can anyone think of anything that would explain this?
Thanks.
------------------------------
From: James Ranson <[EMAIL PROTECTED]>
Subject: nfs mounting problem
Date: Sat, 21 Aug 1999 19:35:40 +0000
I have a computer that mounts all its filesystems from NFS. The setup
works fine with kernel 2.0.36, but when I upgraded to 2.2.11 I experienced
some problems. The kernel was able to mount the root filesystem properly,
but when it attempted to mount /usr it gave the following message:
portmap: server localhost not responding, timed out
Does anyone know what's going on?
James Ranson
http://www.cs.uregina.ca/~ranson/
------------------------------
From: "Matt" <[EMAIL PROTECTED]>
Crossposted-To: comp.security.firewalls
Subject: Removing Telnet Access for a User?
Date: Sun, 22 Aug 1999 01:26:48 GMT
How do you remove permissions for a user/group to access my machine via
telnet? (Linux)
------------------------------
From: "Matt" <[EMAIL PROTECTED]>
Subject: Symlink not working from FTP
Date: Sun, 22 Aug 1999 01:21:33 GMT
I'm trying to create a symlink to a directory on another drive for my FTP
users to access. My FTP users are in a restricted guest group which
normally wouldn't be able to access this folder. The problem I'm having is
that all of the symlinked folders in the FTP users dir don't work,
regardless of where they point to. When a regular user logs in they can
open these but not members of guest groups. What do I need to do to get
this working?
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.networking) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Networking Digest
******************************
