Linux-Networking Digest #447, Volume #11 Tue, 8 Jun 99 01:13:30 EDT
Contents:
ugh, network and apache not talking ("Mike Kelly")
Re: Anyone get Redhat 6.0 + Cable Modem working????? ("NORML")
Re: BNC "grounded type" terminators. ("Larry Brasfield")
HELP: Can't connect to my ISP (SLK230)
Re: ipfwadm only works for sometimes ("RJHM van den Bergh")
Re: Linksys Combo PCMCIA Ethernet Card (Brent Willcox)
Firewall and Router separate? ("Paul Vienneau")
Re: SAMBA routing ? (M. Buchenrieder)
Re: What could it be? Help needed (Dann Church)
Re: IPFWADM and 3 NIC Firewall ([EMAIL PROTECTED])
Re: [Fwd: virtual ethernet devices] (Dann Church)
IBM token-ring cards PCI (Frank Pikelner)
Re: Networking question - Looking for a Howto ("Bobby D. Bryant")
Re: Need help convincing my company Linux TCP/IP stack is safe. (Nicholas E Couchman)
Re: SAMBA routing ? ("Bono")
Re: telnet as root (Henrik OEsterberg)
Re: RedHat 6.0-PortForwarding Problem ("Greg")
Re: NIC Cards needed badly (Marc Holubow)
Re: How do you strip hostnames from mail? ("Cowles, Steve")
Re: IP masq works, but can't read news (Mike)
Xisp works, kppp doesn't (Brian Witowski)
Need help convincing my company Linux TCP/IP stack is safe. (Christian Hudon)
----------------------------------------------------------------------------
From: "Mike Kelly" <[EMAIL PROTECTED]>
Subject: ugh, network and apache not talking
Date: Tue, 08 Jun 1999 01:45:55 GMT
I have a small network that is not connected to the internet. I use it
mostly for testing and fiddling.
It is primarily a Windows network, but I wanted to get to know Linux. I have
a NT server, with a 95 box and a 98 box. They all use TCP/IP and Ethernet to
talk to each other.
The Linux box I added is a RH5.1 in server mode. I can ping it. I can FTP to
it. I can even play QuakeWorld on it from the 95 and 98 machines.
What I can't do is talk to it from a web browser. It won't even talk to
itself as a web browser.
Ah Ha, you say the Apache server is down. Correct. When I try to start it, I
get a message saying:
Cannot determine local host name.
Use ServerName to set it manually.
So being the semi-literate person I am, I try that, and then try and restart
Apache.
I am trying to see what is going on so all I type is httpd. I have tried the
full path/method for stopping and starting Apache, but I don't get any
messages.
I tracked down the error_log and it's only entries were a few lines with:
Created shared memory segment (with a number following a #)
Got a few lines of this and with different numbers each time.
I am guessing that something in my naming scheme is messed up, but I can't
figure out where.
I put a dummy domain in the netcfg field and rebooted, still no change. As
far as I can tell the names I have used are legit.
Any directions would be muchly appreciated.
Thanks for looking at this rambling, whining, plea for help.
--
~mek~
(remove cc from atkincc to email me)
------------------------------
From: "NORML" <[EMAIL PROTECTED]>
Crossposted-To:
linux.redhat.announce,linux.redhat.digest,linux.redhat.install,linux.redhat.list,linux.redhat.misc,linux.redhat.rpm,linux.samba
Subject: Re: Anyone get Redhat 6.0 + Cable Modem working?????
Date: Tue, 08 Jun 1999 01:31:41 GMT
Well I have an NT Server running Microsoft Proxy 2, 2 network interfaces one
configured for internal use and the other configured for external access to
the internet via Cable Modem. I have everything configured correctly on
both ends. Now on my linux box which is running Redhat 6.0 w/Samba, access
the internet via the Proxy with Netscape. My problem is I can not get an
IRC client rather its with socks or not through to the Net, same goes for
Newsgroup clients. I have been banging my head against a wall for 2 weeks
trying to get these working. I need to move on to my ThinkPad and finish
configuring it. Any help would be appreciated.
Thanks
Derrick
John Pisini wrote in message <[EMAIL PROTECTED]>...
>> Mediaone as far as I can tell uses the network card to controll access.
>
>I have moved the network card I got from them into three seperate computers
no
>problem can conect with any of them.
>If I use any of these computers with another nic it fails to connect I
know the
>cards work because the machines show up on my
>network. It is the only thing I can think of. If you are using mediaone
just get
>the numbers for them from a windows box they say you have a dynamic IP but
i've
>been using the same ip address for almost a year with no problems.
>
>
------------------------------
From: "Larry Brasfield" <[EMAIL PROTECTED]>
Subject: Re: BNC "grounded type" terminators.
Date: Mon, 07 Jun 1999 23:50:27 GMT
Posted and emailed.
A serious danger is evident here.
Christopher R. Barry <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Are any of you familiar with the "grounded type" BNC terminator? When
> I bought all my networking hardware last week the standard 50 ohm
> terminators without the little grounding chain were $5.00 a pop, but
> the grounded type were $0.88. They are also 50 Ohm and the back of the
> package seems to indicate they will work fine without connecting the
> grounding chain to anything. Is this correct?
Yes. The chain is solely for tying the coax
outer conductor to an actually grounded
chassis. The outer conductor will not be
connected to anything whatsoever unless
one of those terminators is used and the
chain connected. The correct way to use
it is to connect the chain to only a single
chassis, typically at one end of the cable.
This is not strictly necessary for signalling,
but can preclude strange problems that
might arise due to static pickup or RFI.
> A few days ago when I
> was fiddling around with the back of my PC where all of my cards are
> plugged in, I touched the grounding chain and got a surging, throbbing
> jolt. That chain is just dangling there and if it touched the metal
> chassis of my case I get a feeling the consequences might not be so
> great....
You are lucky you were not electrocuted
and that nobody else has been, yet.
I urge you to have a qualified electrician or
someone knowledgable about electrical
systems check all systems that were
connected to that network. The fact that
you were able to get a continuing shock
means that some source of voltage made
a connection to that conductor. Such a
connection should not have occurred and
should not even be possible given a fault
in one of the NIC's. I suspect that one of
those systems is connected to a miswired
outlet, which is an extremely dangerous
situation. Do not suppose that things are
OK simply because nobody has yet been
shocked. You could easily have a fatal
accident waiting to happen.
In case it will help you give credence to
my caution, I have a BSEE and extensive
experience with electrical systems. My
warning springs from my conviction that
there is a dangerous fault somewhere
among that collection of systems. If you
wish, you may phone me at 2062324396
to discuss what should be done about it.
--
--
Larry Brasfield
Above opinions may be mine alone.
(Humans may reply at unundered [EMAIL PROTECTED] )
------------------------------
From: SLK230 <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.help,comp.os.linux.setup
Subject: HELP: Can't connect to my ISP
Date: Mon, 07 Jun 1999 18:41:08 -0700
Hello,
I have dual boot NT/Linux on my Dell Latitude CPi, everything is working
OK on both OS, except PPP setup on my Linux.
I tried to dial in to my ISP, modem made a dial tone then disconnect
after that and came up an error said that "timeout to send
Config-Request....
I used two IPs address for my nameservers, which I used them on NT side,
but still did not work...
Please advice what did I miss during the setup...
Thanks in advance
------------------------------
From: "RJHM van den Bergh" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: ipfwadm only works for sometimes
Date: Tue, 8 Jun 1999 02:54:41 +0200
One of the sites is
www.worldonline.be
if I do a nslookup its IP is 212.233.1.66
But when I try a a reverse lookup the hostname from the IP it isn't found.
Conclusion their DNS server isn't configured the right way.
Also 212.233.1.66 seems also to be a name server to.
Another site is
www.ifcn.nl its IP is 195.86.34.10
195.86.34.10 corresponds with ns.1-4-all.nl
So this is also a name server
So I'm beginning to suspect a misconfiguration of dns servers.
At least the DNS server of worldonline.be isn't correctly configured.
No referse lookups
But it should be strange if this is the cause.
Rob,
[EMAIL PROTECTED]
RJHM van den Bergh wrote in message
<[EMAIL PROTECTED]>...
<SNIP>
>I do have a problem with Linux RedHat 5.1 and ipfwadm.
>
>Situation:
>ppp0 with external ISDN TA
>Private LAN 10.0.0.0
>The Linux box acts as a gateway to the internet.
>Attached on the Linux Box is an NT workstation.
>
>With the Linux box I can surf to every site.
>With the NT box I can surf to allmost every site.
>I use ipfwadm on the Linux box to do masquarading.
>
>At the end fo this email I attached the firewall script.
>I'm totaly confused why I can't reach some sites.
>DNS seems to work loging the ipfwadm shows
>Jun 7 20:33:46 jads kernel: IP fw-in acc ppp0 TCP 212.233.1.66:80 194
>.159.226.43:61250 L=44 S=0x00 I=4374 F=0x0040 T=118
.......
------------------------------
From: [EMAIL PROTECTED] (Brent Willcox)
Crossposted-To: comp.os.linux.hardware
Subject: Re: Linksys Combo PCMCIA Ethernet Card
Date: Mon, 7 Jun 1999 20:50:33 -04-59
Reply-To: [EMAIL PROTECTED]
I have the same card as you, and I think I can help you.
Difference is my laptop is running Redhat 5.1.
>I looked in /etc/pcmcia/config and it lists the Linksys Combo PCMCIA
>Ethernet Card so I guess I have the software for card services loaded and
>this PCMCIA Ethernet card is supported. When the system boots it always
>says that ethernet initialization is being delayed.
My Toshiba says that upon bootup. It is because the PCMCIA drivers
have not been loaded at that point in time yet. They get loaded later on.
During the daemon start up sequence.
>But when I try to configure the ethernet eth0, I am never given the chance
>to use "pcnet_cs" which is listed in the /etc/PCMCIA/config file. The only
>choices I'm given appear to be non-PCMCIA ethernet cards. What am I doing
>wrong?
I did not need to specify a module in linuxconf or netcfg for the card.
When I plug in the card I get the two high beeps, and the log will say
that eth0is available. My linuxconf fields just list Net Device as eth0
and the lines below that (kernel module, etc) are all blank.
Cardmgr takes care of all the PCMCIA cards when you plug them in.
You'd only need to specify a module for a ethernet adapter on a desktop system.
All I had to do was plug in the card, then I configured eth0 in
linuxconf. Took about 5 minutes if even that.
>Also, although I have a man page for cardinfo, there is no cardinfo
>executable anywhere to be found. The 3Com Noteworthy 56k PCMCIA modem card
>works fine.
Redhat appears to have left cardinfo out of their PCMCIA distribution,
as I don't have it either. It is the X based card information manager.
The command line cardctl is quite helpful though. I'd suspect a goof in their
build environment, as they have the manpage but not the binary.
Report it to them as a bug. They will fix it.
>Any ideas?
Hope this is helpful.
--
Brent Willcox
University of North Texas
mail: bwillcox at jove acs unt edu
Spamblock in use: remove "+usenet" to reply.
------------------------------
From: "Paul Vienneau" <[EMAIL PROTECTED]>
Subject: Firewall and Router separate?
Date: Mon, 7 Jun 1999 23:26:30 -0400
Just finished reading Paul Sery's "LINUX Network Toolkit" and while very
informative, left me with a question. Let me preface this by saying that I
am new to the realm of Linux/Networking and if an applicable HOWTO is out
there, please point me to it.
The Question:
For what seems to be an extremely simple network topology wise, he suggests
putting the router and firewall on separate boxes; each dedicated to their
applicable task. Now, I've been following the discussions here, especially
those related to my topology (DSL line with 10 or so machines), and the
reoccurring theme is:
Dedicate machine to host firewall using ipchains and masquerading.
So, what does one get by establishing a separate firewall network (other
than more configuration time).
He seems to suggest its because of security, however, from what I've read so
far (HOWTO's, discussion groups), this isn't the popular choice
Thanks
-paul
------------------------------
From: [EMAIL PROTECTED] (M. Buchenrieder)
Subject: Re: SAMBA routing ?
Date: Mon, 7 Jun 1999 21:39:56 GMT
Zoran Cutura <[EMAIL PROTECTED]> writes:
[...]
>I don't know about a sollution to the
>name problem,
[...]
WINS service or LMHOSTS files. See other posting.
Michael
--
Michael Buchenrieder * [EMAIL PROTECTED] * http://www.muc.de/~mibu
Lumber Cartel Unit #456 (TINLC) & Official Netscum
Note: If you want me to send you email, don't munge your address.
------------------------------
From: Dann Church <[EMAIL PROTECTED]>
Subject: Re: What could it be? Help needed
Date: Mon, 07 Jun 1999 20:58:40 -0600
Denis,
Only two possibilities that I can think of:
1) DNS entry for godzilla has changed and that hasn't propagated to all the
servers that serve that DNS domain. Check to make sure that the servers listed
in Win98 match the servers listed in Linux (win98 -> winipcfg, linux ->
/etc/resolv.conf) and in the same order.
2) Godzilla is trying to authenticate who you and is unable. Not highly
likely, but possible. At least have seen stranger things in this regard.
Check (in linux) if you can do:
nslookup <ip address assigned to ppp interface>
good luck!
-Dann Church
Denis wrote:
> hi,
>
> I encountered a problem that I don't understand at all.
> It's a problem with my ISP connection, which before worked
> just fine for many months.
>
> Here it is:
> I CAN
> connect to my school (ISP) via a modem, can surve the web (with
> Netscape, for example), can read/post to newsgroups like this one,
> BUT I CAN'T
> withdraw emails from the school, nor can I even telnet it from a terminal
> window - in Win95 on the other hand everything remains fine, there both
> things work. I also tried telneting to another account - it worked and then
> from that account to the one where my email is - it worked, too.
> (when trying to get new mail, I get "Server was not found",
> when "[denis@localhost denis]$ telnet godzilla.acpub.duke.edu", I get
> "godzilla.acpub.duke.edu: Host name lookup failure ")
>
> my linux experiences are driving me crazy, if somebody can please help me
> understand what is going on?
>
> thanks a lot.
> Denis
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: IPFWADM and 3 NIC Firewall
Date: 8 Jun 99 02:11:45 GMT
With the configuration described below you are allowing the ICMP ECHO
packets from network A to the External Network fine, but you aren't
allowing the reply packets back.
The reason that bringing the subnet back one bit fixed it is- if you
ignore the last bit of the third octet then network A and the External
network are the same. (Therefore you were passing all traffic both
directions.)
If you want traffic to flow both directions, you will have to explicitly
allow it, and then block TCP SYN packets from the External network from
reaching the inside network. (I believe that's the -y option with ipfwadm.
I know it is with the 2.2.x IPCHAINS package, anyway.)
Try doing testing with TCP connections (ie: Telnet) in addition to PING.
PING is not stateful and often behaves differently through simple packet
filters (like Linux 2.0.x) than TCP circuits do. Excuse my ambiguity,
I'm a little rusty on the ipfwadm filtering rules.
-K.C.
David Akins <[EMAIL PROTECTED]> wrote:
: I have a machine with 3 NIC's in it acting as a firewall. IP Forwarding
: is enabled and everything is working fine...except IPFWADM is not
: working correctly for me (or I don't quite understand the syntax)
: External Network : 10.100.110.0 eth0
: Internal Network A: 10.100.111.0 eth1
: Internal Network B: 10.100.112.0 eth2
: OK, I do the following:
: ipfwadm -O -p accept
: ipfwadm -I -p accept
: ipfwadm -F -p deny
: Then I flush any rules:
: ipfwadm -O -f
: ipfwadm -I -f
: ipfwadm -F -f
: ipfwadm -A -f
: Bascially, I want to allow Internal Network A to have unlimited access
: to the External Network and the External Network to have No access to
: Internal Network A. I tried the following:
: ipfwadm -F -a accept -P all -S 10.100.110.0/24 -D 0.0.0.0/0
: After running this command I still cannot ping the external network from
: Internal Network A. However if I change the subnet mask to 23 bits as
: in:
: ipfwadm -F -a accept -P all -S 10.100.110.0/23 -D 0.0.0.0/0 it will
: ping. What gives??
------------------------------
From: Dann Church <[EMAIL PROTECTED]>
Subject: Re: [Fwd: virtual ethernet devices]
Date: Mon, 07 Jun 1999 21:37:42 -0600
I agree with with Malware. Either in Solaris or Linux, interfaces get
config'd by nothing more than simple scripts. Thus, if you wanted Solaris to
have a file which dictated all the interfaces instead of a file for each, you
could rewrite the appropriate startup script to look for interface names in
your file.
As for making some interfaces not be active upon reboot, again you would write
your script so that it ignores, for example, all lines beginning with a '#'
character. Then just modify your interface file before reboot.
Good luck!
--Dann Church
Malware wrote:
> Hi George,
>
> you wrote:
> > We are migrating to Linux from solaris. Linux seems to bring up all
> > interfaces based on the IP addresses as configured in
> > /etc/sysconfig/network-scripts/ :
> > ifcfg-eth0
> > ifcfg-eth0:0
> > ifcfg-eth0:1
> > ifcfg-eth0:2
> > etc....
>
> This is not Linux in general, that's how it is done in the RedHat
> distribution by default.
>
> > Under solaris, the interfaces are brought up based the presence of a
> > /etc/hostname.eth0:### file, ### being the virtual interface ID for that
> > IP address.
> >
> > This gets messy when the machine is set to respond to 400 ips. Since
> > Linux does not require these files, the /etc directory is much less
> > cluttered.
>
> Why didn't you change that?
>
> > Question is : Is there a way to prevent an interface from being
> > activated
> > on a reboot outside of doing an ifconfig eth0:### 0 down for each device
> > we wish to deactive?
>
> Yes, you are free to costumize the init-scripts in the fashion you want.
> This could end up with a database based approach e.g.
>
> > Some addresses are not in use but the machine still brings up all
> > interfaces. These addressed do not resolve on a reverse DNS lookup.
> > This causes many messages like :
> >
> > Jun 7 11:41:18 anvil sendmail[19719]: gethostbyaddr(216.42.98.182)
> > failed: 1
>
> You could switch of the interface probing of sendmail.
>
> Malware
------------------------------
From: Frank Pikelner <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: IBM token-ring cards PCI
Date: Tue, 08 Jun 1999 04:21:48 GMT
Hello,
I've read the Token Ring HOWTO, and it states that the IBM token ring II
adapter is NOT supported. Has there been any updates to allow this
adapter to work? I'm using Suse 6.1 and red Hat 6.0
I have tried to install the adapters, but the AUTOPROBE does not pick
them up. For the manual install, I'm not sure what parameters to enter -
any help would be much appreciated.
IBM has no information on their token ring adapters and Linux (none I
could find). My adapters are using IO=7800-78FFF, IRQ=15
Thank you.
Frank Pikelner
mailto:[EMAIL PROTECTED]
------------------------------
From: "Bobby D. Bryant" <[EMAIL PROTECTED]>
Subject: Re: Networking question - Looking for a Howto
Date: Mon, 07 Jun 1999 21:20:38 -0500
Bobby D. Bryant wrote:
> I've been wanting to set up a similar environment just for learning about
> networking, but I'm a complete newbie. ...
Thanks to those who responded. I got the hardware today, and will give it a try
later tonight.
Bobby Bryant
Austin, Texas
------------------------------
From: Nicholas E Couchman <[EMAIL PROTECTED]>
Subject: Re: Need help convincing my company Linux TCP/IP stack is safe.
Date: Tue, 08 Jun 1999 04:19:43 GMT
Tell them it's safer than WinNT, because it truly is. The best place to go
for Linux docs is www.linux.org. They have info on the latest kernels, as
well as documentation on networking and TCP/IP. www.linux.org is running
off an actual Linux server (2 of them, but who's counting?), plus there are
countless web servers for people's personal home pages that have switched
from NT to Linux for there primary storage systems.
Don't let them take away Linux!! If you need to, take your Linux machine
down for a while and (hope) to prove to them that your box isn't causing
the problem.
--Nick
Christian Hudon wrote:
> Hi,
>
> the company I work for has been experiencing networking problems
> recently, and they've started to take a look at everything that's
> connected to their internal network. That includes my Linux box. So I'd
> need help convincing them that Linux's TCP/IP stack doesn't cause
> network floods, is well implemented, etc. I know this is a bit silly,
> but...
>
> So, I'd appreciate pointers to resources showing that Linux's TCP/IP
> stack is implemented according to the RFCs. Pointers to resources
> showing that people actually use Linux on the Internet without causing
> problems (so-and-so % of
> the Internet's web servers are running Linux, company x relies on
> Linux's TCP/IP stack for their business, company y uses Linux for its
> servers, etc.) would also be great. Anything.
>
> If there are other appropriate forums for this kind of questions, I'd
> like hearing about them too.
>
> Thanks in advance.
>
> Christian
------------------------------
From: "Bono" <[EMAIL PROTECTED]>
Subject: Re: SAMBA routing ?
Date: Tue, 8 Jun 1999 12:50:40 +0900
Thanks for your advices.
At last, I succeeded in logging on the winxx boxes in the other network.
However, I didn't use WINS or even SAMBA at all. I turned off SAMBA, still I
can find and access the windows boxes in the other networks.
Just TCP/IP protocol and proper routing. Plus, the entry of IP addresses in
"Hosts" file. That's all.
I even eliminated NetBEUI protocol in windows machines. Still it works. I
can share files, printers and use Network Drives.
I'm very happy now.
But, I got another questions Instead.
1) TCP/IP also does all these things between windows boxes, then why do I
need NetBEUI(which is not routable)?
2) If it(TCP/IP) is routable, can I log on all windows boxes anywhere it
reaches through internet?
I didn't try this, but I think it possible.
Regards,
Bono.
------------------------------
From: Henrik OEsterberg <[EMAIL PROTECTED]>
Subject: Re: telnet as root
Date: Tue, 08 Jun 1999 01:51:08 +0200
Be aware that sending any kind of password over a telnet session is extremely
insecure.
maybe you should check out ssh instead.
www.ssh.fi
Malware wrote:
> Hallo Marten,
>
> you wrote:
> > How can I make a telnet session to a linux computer in my little network
> > as "root", so that I don't need a graphic card and a monitor in this
> > computer ?
>
> 1. Telnet in as normal user and warp to root with "su".
> or
> 2. Edit /etc/login.defs to allow root logins on ttyp0, ttyp1 and so on.
>
> Malware
------------------------------
From: "Greg" <[EMAIL PROTECTED]>
Subject: Re: RedHat 6.0-PortForwarding Problem
Date: Mon, 7 Jun 1999 23:02:36 -0400
Not sure why ipchains and masquerading with have to do with anything
if the box is on the same lan and routed for the interface..???
Excuse me if I miss understood your question,
Would it not be easy to use a alias in the /etc/hosts file... ???
#hosts file
201.101.1.5 your.domain.com linux1:1603
Greg.
Richard Gintz <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> I have ipchains performing masquerading, but am having a problem with port
> forwarding. What I want to do is setup my Linux box such that when
someone
> does "telnet linux1:1603", it actually forwards the packet to another box,
> say 201.101.1.5 port 23; such that the telnet is actually going through
the
> Linux box and hitting the other Unix box. Is there a way to do that with
> "ipchains" or do I need something in addition to that?
> thx,
> [EMAIL PROTECTED]
>
>
>
------------------------------
From: [EMAIL PROTECTED] (Marc Holubow)
Subject: Re: NIC Cards needed badly
Date: Tue, 08 Jun 1999 04:33:13 GMT
[EMAIL PROTECTED] wrote:
>Hello could someone tell me where I can get NIC Cards for RedHat 5.2??
>I have read the http://www.redhat.com/hardware page.
>No luck in finding one of the supported NICs though.
>If someone could point me to a page that has supported NICs I would be
>very very thankful.
>Thanks In Advance.
>Vince
>
The Hardware-HOWTO lists all supported, and some non-supported,
hardware for Linux (this pertains to the kernel, and not a particular
distribution like Red Hat or Slackware).
http://metalab.unc.edu/pub/Linux/distributions/slackware/docs/Hardware-HOWTO
------------------------------
From: "Cowles, Steve" <[EMAIL PROTECTED]>
Subject: Re: How do you strip hostnames from mail?
Date: Tue, 08 Jun 1999 04:01:37 GMT
Matthew,
I your using sendmail, you will need to edit the following line in the
"/etc/sendmail.cf" file from:
# who I masquerade as (null for no masquerading) (see also $=M)
DM
to the following:
# who I masquerade as (null for no masquerading) (see also $=M)
DMdomain.net
This will re-address (masquerade) all outbound emails to "[EMAIL PROTECTED]"
BTW: You will need to "restart" sendmail for the change to take effect.
Steve Cowles
Matthew Hanselman <[EMAIL PROTECTED]> wrote in message
> I'd like the mail from my site to be of the form [EMAIL PROTECTED], not with
> the hostname (right now it's sending mail as [EMAIL PROTECTED]). This
> is a problem since the hostnames don't necessarily exist to the outside
> world (masquerading et al).... any hints?
>
> I've read the HOWTOs, but couldn't find anything...
>
> Thanks. Also, if you could cc: me via email that would be great!
>
> - Matt
>
------------------------------
From: Mike <[EMAIL PROTECTED]>
Crossposted-To: news.software.nntp
Subject: Re: IP masq works, but can't read news
Date: Tue, 08 Jun 1999 03:16:28 +0000
Mike wrote:
>
<snip>
I have received several answers on this topic. The most recent answer
turns out to be the one that worked for me. It was emailed directly to
me, and I'd like to share it now with everyone, just in case someone
else out there is having similar problems.
Mike wrote:
> I set up a Linux IP-masquerading server. Both it and the Win95 box can
> send/receive email, telnet, http, ftp, ping. Only problem is reading
> news. Linux box is fine. Win95 box can get list of new groups, can find
> how many messages in each group, but when I select a group to look at
> (using Netscape v4.6) it just seems to hang. Netscape says "Receiving
> articles..." but nothing ever gets received. Looking at my external
Have you checked MTU settings? Masquerading tends to not work very well
to some hosts if the MTU of the PPP link (typically 296, 576 etc) is
different to that of the local ethernet (almost always 1500). IP masq
simply doesn't deal with MTU path discovery very well in this situation.
This would account for why it works on the linux box itself, as
connections to the news server from there aren't being masquerated.
I'm not sure what the pattern is (if it's certain OSs, or server daemons
or what), but for a long time I had problems where I could connect to
some Web sites fine but others would just hang, and the same would
happen with some IRC servers but not others. Setting the ppp MTU to 1500
fixed these problems.
This may not be your problem, and hence this suggestion may not help.
But at least it's something to try.
--
Work: [EMAIL PROTECTED] Home: [EMAIL PROTECTED]
Electrical and Electronic Engineering Dept.
University of Auckland
--
A big "thank-you" to you for providing the fix for me! I changed the mru
and mtu to 1500 and now things are working great. It has allowed
connection to the news server for the masq'ed boxen and it has cleared
up some web connection and speed problems as well! A great, great piece
of advice. Thanks again!
--
Mike Wright
------------------------------
From: Brian Witowski <[EMAIL PROTECTED]>
Subject: Xisp works, kppp doesn't
Date: Tue, 08 Jun 1999 00:00:45 -0400
Greetings,
I have a strange problem. Using Xisp I can connect to my isp and log
in, no problem. But If I use, say Kppp,
I can connect but never get a Login: or Password: prompt. After about
10 sec. it dosconnects. The same thing happens with a terminal
program. I know that xisp has it's own dialer but I need for this to
work the 'normal' way because I want to set up diald.
Please help!
Brian
------------------------------
From: Christian Hudon <[EMAIL PROTECTED]>
Subject: Need help convincing my company Linux TCP/IP stack is safe.
Date: Tue, 08 Jun 1999 04:15:19 GMT
Hi,
the company I work for has been experiencing networking problems
recently, and they've started to take a look at everything that's
connected to their internal network. That includes my Linux box. So I'd
need help convincing them that Linux's TCP/IP stack doesn't cause
network floods, is well implemented, etc. I know this is a bit silly,
but...
So, I'd appreciate pointers to resources showing that Linux's TCP/IP
stack is implemented according to the RFCs. Pointers to resources
showing that people actually use Linux on the Internet without causing
problems (so-and-so % of
the Internet's web servers are running Linux, company x relies on
Linux's TCP/IP stack for their business, company y uses Linux for its
servers, etc.) would also be great. Anything.
If there are other appropriate forums for this kind of questions, I'd
like hearing about them too.
Thanks in advance.
Christian
------------------------------
** 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
******************************
