Linux-Networking Digest #235, Volume #10 Wed, 17 Feb 99 20:13:37 EST
Contents:
Re: diald: first connection doesn't respond (Philboyd Studge)
Help with ACKs ("Ayad Al-Adnani")
Re: FTP not working w/ipchains, help desperately needed (Kevin Dick)
IP Masquerade & MAC Numbers ("JSH")
Problem level of PPP+linux 2.2.1? (Daren Scot Wilson)
Re: IPCHAINS/Firewalling: END OF OUR ROPE (Marc Haber)
IPCHAINS/Firewalling: END OF OUR ROPE (Kevin Dick)
Can't get NFS rw-access (Stefan Lingdell)
Re: Cable lengths (Stuart R. Fuller)
Samba Setup (jimterm)
Re: can't get computers to recognise each other (Darren Greer)
Re: Newbie wants to install networking. Help! ("x")
trouble building IMAP-4.5.BETA, Re: Pop3 server for linux ("Cameron Spitzer")
Re: New Redhat installation - cannot telnet to localhost. (Villy Kruse)
Newbie question: how do I set up networking? ("Brian Inderwies")
Redhat 5.2 - DEC Tulip problem (Ben Lovett)
Re: mail server (Luca Filipozzi)
PPP only gets to ISP ("Jay Beatty")
Re: Mail and Diald (Matteo Valsasna)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Philboyd Studge)
Crossposted-To:
linux.debian.qa,linux.debian.user,comp.os.linux.help,comp.os.linux.questions
Subject: Re: diald: first connection doesn't respond
Date: Wed, 17 Feb 1999 23:05:57 GMT
add this line to /etc/rc.d/rc.local:
echo 1 > /proc/sys/net/ipv4/ip_dynaddr
that will make the kernel 'hold its water' until the defaultroute is
established
On Tue, 09 Feb 1999 06:15:34 GMT, MikeF <[EMAIL PROTECTED]> wrote:
:::I'm running diald .16.5-2 in debian 2.0. It will dial automatically when it
:::senses ip traffic, but any connections (http,tcp,ping,udp,icmp) that are started
:::before/during the dialup scripts are running don't go through. They just
:::time-out. How can I fix this? I'm including any relevant config files and my
:::ppp log (which includes the diald logs).
------------------------------
From: "Ayad Al-Adnani" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.ms-windows.networking.windows,comp.os.ms-windows.programmer.networks
Subject: Help with ACKs
Date: Wed, 17 Feb 1999 22:43:07 -0000
When using piggybacking for ACKs, a receiver will be in the position of
wanting to send an ACK without any data to piggback on. What happens if the
receiver waits too long or too short for a data packet to piggyback the ack
before sending a reply???
------------------------------
From: Kevin Dick <[EMAIL PROTECTED]>
Subject: Re: FTP not working w/ipchains, help desperately needed
Date: Wed, 17 Feb 1999 10:32:48 -0600
We're having similar problems with ftp-data traffic. You're not alone.
We're using ipchains with Kernel 2.2.1. I haven't tried "insmod
ip_masq_ftp" yet.
Clinton Pierce wrote:
> Had IP Masquerading working PERFECTLY with Kernel 2.0.36, but for
> reasons beyond my control I had to upgrade to 2.2.1.
>
> It almost works! All I want (for now) is my internal LAN to be able
> to access the 'Net using IP Masquerading. So under 2.0.36 I was
> doing:
>
> # Kernel 2.0 Forwarding...
> /sbin/ipfwadm -F -f
> /sbin/ipfwadm -F -p deny
> /sbin/ipfwadm -F -a m -S 192.168.1.0/24 -D 0.0.0.0/0
>
> /sbin/insmod ip_masq_irc
> /sbin/insmod ip_masq_ftp
> /sbin/insmod ip_masq_raudio
>
> And everything worked PERFECTLY. Under the 2.2 kernel, through a
> confusing array of HOWTOS on ipmasq, ipfwadm and ipchains..I figured I
> needed only this:
>
> # Kernel 2.2 Forwarding
> /sbin/ipchains -A forward -j MASQ -s 192.168.1.0/24\
> -d 0.0.0.0/0
> /sbin/insmod ip_masq_irc
> /sbin/insmod ip_masq_ftp
> /sbin/insmod ip_masq_raudio
>
> And that ALMOST works, except that nobody on the LAN can reliably FTP
> out onto the 'Net going through the Masquerading host. (FTP hangs,
> gives messages like "unable to create data socket", etc...) HTTP
> works, telnet works, everything else works (even ICQ!). I realize
> that FTP requires a two-way connection but was hoping that the module
> (which was recompiled for 2.2, thank you) would take care of that like
> it had before.
>
> Any help is appreciated.
>
> PS: To the author of the IPCHAINS-HOWTO: if you read this I'd be MORE
> THAN HAPPY to help re-write the section entitled "I'm confused!
> Routing Masquerading, ipautofw..." You never explained the differences
> between those technologies and left the whole thing a confusing mess.
> --
> "If you rush a Miracle Man, you get rotten miracles"
> --Miracle Max, The Princess Bride
> DNRC: "Grand Inquisitor of Out At 5 Doctrine" 06/96
------------------------------
From: "JSH" <[EMAIL PROTECTED]>
Subject: IP Masquerade & MAC Numbers
Date: Wed, 17 Feb 1999 11:43:56 -0500
I want to set up an aDSL connection. My Linux box would be connected to the
DSL modem via eth0 and to a Windows98 box via eth1, like this:
W --- (eth1) L (eth0) --- ISP
(W is the Windows box on the internal network and L is the masquarade
gateway).
The ISP, however, tracks my MAC number to make sure that I only have one
computer using the DSL connection. Will IP Masquerade replace the MAC number
on packets coming from the internal network with its own before shipping
them out onto the Internet, or will the MAC numbers remain unchanged?
Any help you could give would be appreciated. Thanks!
Dominik Slusarczyk
[EMAIL PROTECTED]
------------------------------
From: Daren Scot Wilson <[EMAIL PROTECTED]>
Crossposted-To: comp.protocols.ppp,comp.os.linux.setup
Subject: Problem level of PPP+linux 2.2.1?
Date: Wed, 17 Feb 1999 16:50:55 +0000
I'm new to these newsgroups,would like the judgement of old timers: I'm
wondering if the number of PPP problems related to the new linux 2.2.1 is
the "normal level" of trouble routinely found here, or has there been a
sharp rise from past levels? If the latter, would you attribute this to
changes in the software, or a sharp rise in beginners?
PPP has always worked fine for me, or at least nothing requiring spelunking
the newsgroups, HOWTOs and everything, all through 1996, 1997, 1998, but
suddenly with linux 2.2.1, kablewy!!!
daren wilson
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Marc Haber)
Subject: Re: IPCHAINS/Firewalling: END OF OUR ROPE
Date: Wed, 17 Feb 1999 15:35:53 GMT
Kevin Dick <[EMAIL PROTECTED]> wrote:
>I am coming close to giving up on Linux for our firewall. We've been
>unable to determine how to use IPCHAINS for port forwarding of inbound
>mail (and web and etc) to alternative machines through our linux
>firewall under construction.
You will have to give more details about your network topology.
Greetings
Marc
--
====================================== !! No courtesy copies, please !! =====
Marc Haber | " Questions are the | Mailadresse im Header
Karlsruhe, Germany | Beginning of Wisdom " | Fon: *49 721 966 32 15
Nordisch by Nature | Lt. Worf, TNG "Rightful Heir" | Fax: *49 721 966 31 29
------------------------------
From: Kevin Dick <[EMAIL PROTECTED]>
Subject: IPCHAINS/Firewalling: END OF OUR ROPE
Date: Wed, 17 Feb 1999 09:08:27 -0600
I am coming close to giving up on Linux for our firewall. We've been
unable to determine how to use IPCHAINS for port forwarding of inbound
mail (and web and etc) to alternative machines through our linux
firewall under construction.
Please help. We are trying NOT to buy an upgrade for Raptor on Windows
NT; this is a $5000 expense.
Kevin
------------------------------
From: [EMAIL PROTECTED] (Stefan Lingdell)
Subject: Can't get NFS rw-access
Date: 17 Feb 1999 23:23:00 GMT
--
Hi,
I'm trying to boot a diskless SparcStation ELC by using tftp to load the
kernel and then NFS to mount the root file system. Booting the kernel
works fine, but I am having some problems with the NFS mount; I can't
get write permissions! The NFS mount seems to be readonly, even though
I have specified rw-access in /etc/exports. This is what my
/etc/exports looks like:
/tftpboot (rw,insecure,no_root_squash)
Anyone have any suggestions?
Below are som extracted logs (if it is of any help):
mountd
======
[root@eggman nfs-server-2.2beta37]# rpc.mountd -F -d auth
mountd[598] 02/17/99 23:04 can't get my own address
mountd[598] 02/17/99 23:04 clnt <anon clnt> exports:
mountd[598] 02/17/99 23:04 /tftpboot rw
mountd[598] 02/17/99 23:04 flushed host access cache
mountd[598] 02/17/99 23:04 NFS mount of /tftpboot/193.10.242.167 attempted
from 193.10.242.167
mountd[598] 02/17/99 23:04 check unknown clnt addr 193.10.242.167
mountd[598] 02/17/99 23:04 Anonymous request from 193.10.242.167.
mountd[598] 02/17/99 23:04 auth_path(/tftpboot/193.10.242.167): mount
point /tftpboot, (all_squash insecure rw)
mountd[598] 02/17/99 23:05 /tftpboot/193.10.242.167 has been mounted by
193.10.242.167
nfsd (same options):
===================
nfsd[599] 02/17/99 23:07 auth_path(/tftpboot/193.10.242.167/root/.bashrc):
mount point /tftpboot, (all_squash insecure rw)
nfsd[599] 02/17/99 23:07 auth_path(/tftpboot/193.10.242.167/usr/bin/mesg):
mount point /tftpboot, (all_squash insecure rw)
nfsd[599] 02/17/99 23:07
auth_path(/tftpboot/193.10.242.167/usr/local/bin): mount point /tftpboot,
(all_squash insecure rw)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
E-mail: [EMAIL PROTECTED] Tel: 031-182697
Snailmail: Stefan Lingdell ICQ: 5602825
Gibraltargatan 94/303 IP: 193.10.242.145
412 79 G�teborg
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
------------------------------
From: [EMAIL PROTECTED] (Stuart R. Fuller)
Subject: Re: Cable lengths
Reply-To: [EMAIL PROTECTED]
Date: Wed, 17 Feb 1999 17:00:03 GMT
[EMAIL PROTECTED] wrote:
: I've got two buildings, about 100' apart.
: I'm wondering what the best way to cable the
: two together will be. (We're getting xDSL
: installed in one building, but residents in
: the other want a piece of the access as well.)
: We'll be stringing the cable outside, basically
: from tree to tree. =)
:
: I guess my questions are these:
:
: 1. 10BaseT is listed as having a 100 meter cable
: run limit, which would be just about long enough
: to run from one hub (directly connected to our
: server/router via 10BaseT) to the windowsill of
: the other building, where we could stick an
: additional hub. Is the 100 meter limit a total
: limit per network segment, or could we go hub-to-
: hub with a cable of roughly 250'?
: 2. Would we be better to invest in a pair of inexpensive
: hubs with 10BaseT and 10Base2 ports, and run 10Base2
: from one building to a hub on the window-sill of the
: other?
:
: We're college students doing this out-of-pocket, so
: expensive esoteric gear is kinda out of the question.
: Practical, cheap advice is always much appreciated! :-)
For cheap advice, go to the electrical engineering side of your campus, and
ask them the wisdom of running a copper wire between the two buildings.
I'd advise running fiber optic between the buildings. It's less likely to
want to take a lightning strike.
Stu
------------------------------
From: jimterm <[EMAIL PROTECTED]>
Subject: Samba Setup
Date: Wed, 17 Feb 1999 09:34:47 -0500
Admittedly, I am a Linux "newbie/wannabe". I have installed Linux on a
test
machine. I want to eventually use it as a server to Win95/98
computers.
I have successfully installed Red Hat Linux 5.2. My Win98/95 machines
can see that "Linuxserver" is there in the Network Neighborhood. But
when I try to access it - I get a dialogue box asking me to login to
"\\Linuxserver\IPC$" - which tells me I haven't got some parameter set
right, probably a "share".
"Testparm" says everything is ok.
Any suggestions, help or other sources will be greatly appreciated.
------------------------------
From: [EMAIL PROTECTED] (Darren Greer)
Subject: Re: can't get computers to recognise each other
Date: Wed, 17 Feb 1999 23:38:36 GMT
Perhaps you should post this in a Win98 networking group. As this
group pertains to Linux networking.
You may get better luck by doing so.
Darren
On Wed, 17 Feb 1999 19:42:29 GMT, [EMAIL PROTECTED] wrote:
-->I am trying to use the Linksys Ether 16 cards to link my P75 and P200 together
-->for my kids. Both systems are running win98. Both computers show only
-->themselves in network neighborhood. Both pass their tests quickly. I have not
-->been able to set up a master and am getting very frustrated. Thanks Elisa
-->
-->-----------== Posted via Deja News, The Discussion Network ==----------
-->http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: "x" <[EMAIL PROTECTED]>
Crossposted-To:
athome.users-unix,comp.os.linux.help,comp.os.linux.questions,comp.os.linux.setup
Subject: Re: Newbie wants to install networking. Help!
Date: Wed, 17 Feb 1999 16:54:24 GMT
get a piece they are much easier to install
Brian Barjenbruch wrote in message ...
>Okay, here's the deal:
>
>I have a Power Macintosh G3 with a *built-in* Ethernet port. How do I
>get Linux to recognize this? What are the steps I need to go through
>to install networking with this (I have a cable modem connected to the
>Ethernet port)?
>
>Again: It's not on a card, it is a built-in Enet port. Does this make
>a difference, Linux-wise? (LinuxPPC)
>
>--
>"Its origin and purpose...still a total mystery."
> - Dr. Heywood Floyd, "2001: A Space Odyssey"
------------------------------
From: "Cameron Spitzer" <[EMAIL PROTECTED]>
Subject: trouble building IMAP-4.5.BETA, Re: Pop3 server for linux
Date: 18 Feb 1999 00:13:47 GMT
In article <[EMAIL PROTECTED]>, Jonas <[EMAIL PROTECTED]> wrote:
>A fast way is to just install IMAP rpm and restart. Now you have a IMAP4 and
>POP3 server.
I took the IMAP/POP3 server modified for Qmail from the Qmail site.
(imap-4.5.BETA + five patches)
Problem was the mailbox name was hardcoded into the binary for POP3,
and I use a different mailbox name. So I got the source RPM, same
place. The five patches installed with no problem, once I examined
their internal dates to figure out what order they go in.
(No installation instructions in the RPM, to my untrained eye.)
I then tried make slx (I've got shadow passwords) but the make failed,
couldn't resolve the "crypt" symbol. This was fixed by running
make slx EXTRALDFLAGS="-llibcrypt"
This build gets as far as trying to compile c-client/osdep.c, where it
dies as follows:
cc -g -fno-omit-frame-pointer -O6 -DNFSKLUDGE `cat OSCFLAGS` -c osdep.c
osdep.c: In function `checkpw':
osdep.c:190: `PAM_ESTABLISH_CRED' undeclared (first use this function)
osdep.c:190: (Each undeclared identifier is reported only once
osdep.c:190: for each function it appears in.)
make[3]: *** [osdep.o] Error 1
make[3]: Leaving directory `/usr/src/mail/imap-4.5.BETA/c-client'
make[2]: *** [slx] Error 2
make[2]: Leaving directory `/usr/src/mail/imap-4.5.BETA/c-client'
make[1]: *** [OSTYPE] Error 2
make[1]: Leaving directory `/usr/src/mail/imap-4.5.BETA'
make: *** [slx] Error 2
Evidently, there's another missing library besides libcrypt. Can anyone tell
me the next missing piece, where to get it, how to use it?
TIA!
Cameron
http://petra.greens.org/~cls/
------------------------------
From: [EMAIL PROTECTED] (Villy Kruse)
Subject: Re: New Redhat installation - cannot telnet to localhost.
Date: 17 Feb 1999 16:18:07 +0100
In article <7aegoq$ua6$[EMAIL PROTECTED]>, <[EMAIL PROTECTED]> wrote:
>Nope - they're installed. I've checked out the inetd config files (sorry can't
>remember the name - linux hasn't burnt itself into my nervous system yet) and
>they all contain commands for running ftp and telnet. I'll check this again to
>make sure (wont hurt).
These lines will always be there and enabled even if the daemon program is
not installed. We are talking about the existence of the file in.telnetd
# grep telnet /etc/inetd.conf
telnet stream tcp nowait root /usr/sbin/tcpd in.telnetd
# ls -l /usr/sbin/in.telnetd
-rwxr-xr-x 1 root root 41040 Apr 24 1998 /usr/sbin/in.telnetd
Villy
------------------------------
From: "Brian Inderwies" <[EMAIL PROTECTED]>
Subject: Newbie question: how do I set up networking?
Date: Wed, 17 Feb 1999 19:21:01 -0500
This is my scenario: I have a Pentium PC that boots both Windows 98 and
linux. I installed linux (RedHat 5.1) the other day and am having problems
setting up access to the Internet over TCP/IP. I am on a college-dorm
Ethernet that works very well with windows TCP/IP, yet I cannot figure out
the configuration under linux. When I installed, I set up my Ethernet
adapter just fine into the kernel (a Netgear LNE100-TX as the "tulip" driver
per their instructions), and it seems to initalize okay under linux bootup.
But I am having a problem with the settings. I have an IP address, DNS
server, subnet mask, and default gateway from the winipcfg (on the Windows
98 side) but these seem to lock-up linux. Windows uses the DHCP to get an
IP address, but when I try to use dhcp in linux, I get a message that it
failed. My IP address, however, seems pretty static, so I tried the manual
approach. Nothing seems to work that way. When I try to ping another
comptuer, the program starts but never receives any packets.
This seems verbose, but if anyone could help me with this problem, I would
greatly appreciate it!!
------------------------------
From: Ben Lovett <[EMAIL PROTECTED]>
Subject: Redhat 5.2 - DEC Tulip problem
Date: Wed, 17 Feb 1999 09:05:53 -0800
I have a problem setting up a DEC Tulip type card on a system running
Redhat 5.2.... I can either have a static IP, or use DHCP to recieve an
IP. Every time I boot up Redhat, it always says:
Delaying eth0 initialization
Whats up with that? In Windows98 I can get a static IP or a dynamic
IP? But not in Linux? can anybody help?
thanks in advance
------------------------------
From: [EMAIL PROTECTED] (Luca Filipozzi)
Subject: Re: mail server
Date: Wed, 17 Feb 1999 09:20:36 -0800
In article <[EMAIL PROTECTED]>, [EMAIL PROTECTED]
says...
> Richard Sydney-Smith wrote:
> >
> > What is the best (free) mail server?
> >
> > for redhat 5.2 ?
> >
> > richard
> use sendmail (www.sendmail.org)
>
sendmail is big and hard to configure
i suggest smail
Luca
--
Luca Filipozzi <[EMAIL PROTECTED]>
------------------------------
From: "Jay Beatty" <[EMAIL PROTECTED]>
Subject: PPP only gets to ISP
Date: Wed, 17 Feb 1999 19:34:34 -0500
I've just installed RedHat 5.2, and I can't get ppp working. Or, more
precisely, I can only connect to the isp. I think the problem is that ppp is
not picking up the dynamic IP address. It keeps assigning the lan ip
address.
It dials properly. route -n shows both the host route and the default route.
ifconfig shows the client ppp address to be the same as my lan address.
I can reach my isp, but no further. Addresses are resolved by DNS. When I
ping www.linux.org, I get the ip address, but no response.
I've used netcfg. When configuring the ppp interface I've left the local and
remote ip addresses blank ( I also tried local 0.0.0.0 ). I've removed the
eth0 interface. "activitate" dials and gives no error messages. Always the
same result.
Any help appreciated.
Jay
------------------------------
From: Matteo Valsasna <[EMAIL PROTECTED]>
Subject: Re: Mail and Diald
Date: Wed, 17 Feb 1999 18:14:12 +0100
Vaughan McPherson wrote:
>
> I have set up a Linux Box running diald for a small office network which
> is made up of various windows boxes. We have a registered mail domain so
> all messages end up in one mailbox regardless of the user to which they
> were sent. I want to be able to download the messages automatically and
> then drop then into individual locally hosted mail accounts so that
> users only receive mail inteded for them. Since I'm woefully ignorant
> about mail administration can someone tell me how to go about this ?
I did not understand on which host is your single mailbox with all
messages, but I think fetchmail can do the trick
in the man pages there are examples (multidrop mailboxes) that may fit
your needs.
MAtteo
===========================================================
If You don't Sweat You don't Enjoy
Make your PC happy: give him an OS (e.g. www.linux.org)
------------------------------
** 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
******************************
