Linux-Networking Digest #24, Volume #10 Wed, 27 Jan 99 20:13:54 EST
Contents:
Re: ppp-server problem (Josh Gentry)
Re: PPP internet connection problem (Luca Filipozzi)
automount / umount SMB partition (Kelvin Leung)
Re: Simple? Sendmail newbie question (Wile E. Coyote)
Re: Win95-Linux VPN (Dai to)
Re: IPX/PPP Client Problem (Andy Neverowsky)
Re: RJ-45 network needs hub ? (David 'Septimus' De Ridder)
Compex RL100/ATX and kernel 2.1.x (Nenad Ocelic)
Re: Kernel 2.2 in RPM (Dan Nguyen)
Re: machine crash (David Kirkpatrick)
Re: Telnet stops working when Internet connection goes down (Matthew Willcock)
Re: PPP connects but I don't get out anywhere (Darren Greer)
DNS caching server on PPP router, elegant solution needed ("Victor Chou")
Re: is my system overloaded? (Stefan Sundkvist)
Re: This is Linux, not Windows, so why not superior flexibility AND idiot-friendly?
(NF Stevens)
Re: Adding Routes (David Efflandt)
PC linux accessible through Mac (Matthieu Buchs)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Josh Gentry)
Crossposted-To: alt.os.linux,aus.computers.linux,comp.os.linux.misc,comp.os.linux.setup
Subject: Re: ppp-server problem
Date: Sun, 24 Jan 1999 16:23:35 -0800
You may want to consult my document on dialin server setup, although I do
not cover the use of chap.
http://www.swcp.com/~jgentry/dialin2.html
*** Posted from RemarQ - http://www.remarq.com - Discussions Start Here (tm) ***
------------------------------
From: [EMAIL PROTECTED] (Luca Filipozzi)
Subject: Re: PPP internet connection problem
Date: Sun, 24 Jan 1999 16:18:04 -0800
In article <78g8m0$3uk$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] says...
> Help! I got my modem to dial my ISP but that's about it. I can't seem to
> connect to their DNS, let alone accessing via a web browser.
>
> my /etc/resolv.conf is as follows:
> search .
> nameserver xxx.xxx.xxx.xxx (I obtain the nameserver from my ISP)
Looks OK. Don't think you need the search directive.
> Connect:ppp0<--->/dev/modem
> local IP address ###.###.###.###
> Remote IP address ###.###.###.###
>
>
> Everything seemed good. Once connected, I tried to ping the DNS (specified
> in /etc/resolv.conf) both by IP and my host. Neither one worked. It just
> hung. Do I need to set my default route in my routing
> table? Did I enter the nameserver incorrectly or did I enter it in the
> wrong file? Please email and thanks in advance.
That's exactly right. You need to set the default route to the remote IP
address.
Try this command, "traceroute -n xxx.xxx.xxx.xxx"and see where your
machine wants to send the packets.
Hope this helps,
Luca
--
Luca Filipozzi <[EMAIL PROTECTED]>
------------------------------
From: [EMAIL PROTECTED] (Kelvin Leung)
Crossposted-To: comp.os.linux.setup,comp.os.linux.misc
Subject: automount / umount SMB partition
Date: Wed, 27 Jan 1999 14:41:10 -0700
Hello,
I got a RH 5.1 server running as Samba server for windoze machines. And my
own Linux workstation is on the same network. I would like to know how to
setup the automount of Samba partition from the Samba server (do something
in /etc/fstab?). So that I don't need to smbmount and smbunmount everytime
I reboot the system... Thanks
Kelvin
------------------------------
Subject: Re: Simple? Sendmail newbie question
From: [EMAIL PROTECTED] (Wile E. Coyote)
Crossposted-To: comp.os.linux.setup
Date: Sun, 24 Jan 1999 18:17:52 -0600
Thanks to everybody who responded, I'm going off to "try" some things now...
In article <uDg#lpuR#[EMAIL PROTECTED]>, [EMAIL PROTECTED] says...
>
>OK, my Linux server is up & running RedHat 5.1, with kernel version 2.0.35.
>Now I want to get sendmail setup. I would like to fix an immediate problem
>that I have first, and learn about the inner workings of sendmail later, if
>possible, so here's my question:
>
>I want to set up my server so that an outside user can use it as an SMTP
>server. It works fine as a POP server, but when I try to send mail through it
>from another computer using it as the SMTP server, it says that it doesn't
like
>my destination address, and it only seems to take destinations that are on
that
>server itself, ie. [EMAIL PROTECTED]
>
>Ideas? I am trying to get this piece working first, then I'll end up buying a
>good book on sendmail later and end up tweaking it more extensively....
>
>TIA,
>Scott
>
------------------------------
From: Dai to <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: Win95-Linux VPN
Date: Sun, 24 Jan 1999 14:21:00 -0800
Yes it can be done, but how depends on what type of a connection you are
looking for. Is just a telnet session? If so look at
>http://www.ssh.fi/sshprotocols2/index.html
If you want more you might look at http://www.redcreek.com. Red Creek is a
great out of the box vpn solution, but just bring money.
Wayne
Eric Rossing wrote:
> We have an office in another city with a few non-networked computers. Here
> in the main office, our network is connected to the Internet through a
> Linux-based firewall/server.
>
> Is it possible to have our out-of-town users connect to the Internet through
> a local (to them) ISP, and then establish a VPN connection through our Linux
> server to access our network?
>
> If so, where can I go to find more information?
>
> Thanks!
>
> Eric Rossing
> Intec Company, Inc.
> [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Andy Neverowsky)
Subject: Re: IPX/PPP Client Problem
Date: Wed, 27 Jan 1999 12:42:52 GMT
On Tue, 26 Jan 1999 19:20:40 +0100, Malware
<[EMAIL PROTECTED]> wrote:
>Hi Andy,
>
>you wrote:
>> 2)Near line 514, from:
>> #define ACKCIPROTO(opt, neg, val) \
>> if (neg) { \
>> To:
>> #define ACKCIPROTO(opt, neg, val) \
>> if (neg && (len>0)) { \
>
>Did you try this out? I do not see how it does work. In the first run of
I try - this works.
>the loop [ do {...} while(0);] len does get less 0 in ACKCINAME (I
>assume go->neg_name to be 1 as I see no way to get it 0 than a reject
>received for this option). With your patch it does skip the ACKCIPROTO
>lines and then does run into the loop again because len!=0. On the top
>of the loop it then will be trown out of the loop by the check "if ((len
>-= CILEN_NETN) < 0)" in ACKCINETWORK. After the loop it does return 0
>and the frame is still declared as being invalid. I'am wrong about? If
>yes where is my point of failure?
>
>The sense of the other part of your patch I see. Just wondering there
>was no warning by gcc about the "if(internal & 0)" never beeing true.
First of all I must say that I talk about ppp-2.3.5. (I hav't 2.3.4
and don't know about it).
Pppd sends IPXCP ConfigReq packet with information about our network
number, node, may be IPX router name (I don't use it), and IPX
router protocol. But if we use RIP/SAP (default) it don't send
any info about router proto.
If all are OK, remote host reply to us with IPXCP ConfigAck with
SAME fields. This is handled by function ipxcp_ackci:
...
/*
* Process the ACK frame in the order in which the frame was assembled
*/
do {
ACKCINETWORK (IPX_NETWORK_NUMBER, go->neg_nn,
go->our_network);
ACKCINODE (IPX_NODE_NUMBER, go->neg_node,
go->our_node);
ACKCINAME (IPX_ROUTER_NAME, go->neg_name, go->name);
ACKCIPROTO (IPX_ROUTER_PROTOCOL, go->neg_router,
go->router);
ACKCIPROTO (IPX_ROUTER_PROTOCOL, go->neg_router,
go->router);
ACKCIPROTO (IPX_ROUTER_PROTOCOL, go->neg_router,
go->router);
/*
* This is the end of the record.
*/
if (len == 0)
return (1);
} while (0);
...
ACKCINETWORK: we send netnumber - we received it. OK
ACKCINODE: same with node. OK
ACKCINAME: Hm... I don't use it and have go->neg_name == 0.
Therefore this macro is skiped.
ACKCIPROTO: Oops !!! If we have RIP/SAP we did't send any proto
info. And remote host DO NOT reply us with it!!!
len == 0 in this point(!!), but go->neg_router NEVER be
zero !
Macros works and decrease len. len become <0.
Function returns error.
>From Joe Loyall:
> Jan 25 21:04:00 Rastaman pppd[2015]: Received bad configure-ack: 01 06
> 23 90 32 31 02 08 00 00 00 00 5d 6f
"Decript" of this message :
01 - IPX_NETWORK_NUMBER (from ipxcp.h)
06 - CILEN_NETN. Length of this field include id and length bytes.
23 90 32 31 - Joe net number (0x31329023 - big endian)
02 - IPX_NODE_NUMBER
08 - CILEN_NODEN
00 00 00 00 5d 6f - node number
Really no errors but not info about router proto. My patch resolve
this.
But if I should be writing this function, I wrote this fragment
of code like (shematicaly):
#define IPXCI_ERR_NO 0
#define IPXCI_ERR_BADLEN 1
#define IPXCI_ERR_BADID 2
int error = IPXCI_ERR_NO;
while (len > 2) {
GETCHAR(citype, p);
GETCHAR(cilen, p);
len -= 2;
switch (citype) {
case IPX_NETWORK_NUMBER:
...
break;
case IPX_NODE_NUMBER:
...
break;
case IPX_COMPRESSION_PROTOCOL:
...
break;
case IPX_ROUTER_PROTOCOL:
...
break;
case IPX_ROUTER_NAME:
...
break;
case IPX_COMPLETE:
...
break;
default: // we don't know this id
... //error handler
error = IPXCI_ERR_BADID;
}
}
if (error) {
...
}
------------------------------
From: [EMAIL PROTECTED] (David 'Septimus' De Ridder)
Subject: Re: RJ-45 network needs hub ?
Date: Wed, 27 Jan 1999 12:46:48 GMT
Reply-To: [EMAIL PROTECTED]
Whereto with speedy words John Smith repli'd :
>You can use a hub and full duplex.. if you have a really good one :o)
Such as ?
I hope you're not referring to 'switching hubs' etc.
I was talking about hubs as 'bit-repeaters', or
electrical-signal-repeaters, if you like.
In this context, hubs and FDX don't mix, kids.
(It may/will work, but with terrible performance).
Regards,
+-----------------------------------------------------------------+
David 'Septimus' De Ridder <[EMAIL PROTECTED]>
"You are alone. Maybe I can help you."
- UNKLE
+-----------------------------------------------------------------+
------------------------------
From: Nenad Ocelic <[EMAIL PROTECTED]>
Subject: Compex RL100/ATX and kernel 2.1.x
Date: Wed, 27 Jan 1999 13:44:55 +0100
Reply-To: [EMAIL PROTECTED]
My Compex RL100/aTX cards worked fine with kernel 2.0.36 using
winbond-840 module. After I swithced to development kernel (2.1.131)
winbond module could not find MII PHY any more. Same thing with 2.2.0.
Might it be wrong kernel compilation options (PCI, PNP?) or just
winbond-840.o not designed for new kernels?
I would appreciate any suggestions.
Thnx,
NO
[EMAIL PROTECTED]
------------------------------
From: Dan Nguyen <[EMAIL PROTECTED]>
Crossposted-To: alt.linux,alt.os.linux,linux.redhat,linux.redhat.misc
Subject: Re: Kernel 2.2 in RPM
Date: 28 Jan 1999 00:12:41 GMT
In alt.os.linux John K <[EMAIL PROTECTED]> wrote:
: Can somebody compile the new kernel in a RPM package for "newbies", please.
: I have RedHat 5.2 and I look forward for something easy to upgrade without
: making it a pain, or making a mistake as a result to have to re-install
: again Linux for the 400th time.
It's not that hard to compile a kernel. There is no advantage with
having someone else compile your kernel, because you won't gain any
advantages.
--
Dan Nguyen | There is only one happiness in
[EMAIL PROTECTED] | life, to love and be loved.
http://www.cse.msu.edu/~nguyend7 | -George Sand
------------------------------
From: David Kirkpatrick <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup,comp.os.linux.misc,comp.os.linux.hardware
Subject: Re: machine crash
Date: Wed, 27 Jan 1999 17:45:19 +0000
Reply-To: [EMAIL PROTECTED]
Brad,
Look at /usr/src/linux/Documentation/opps-tracing.txt. I
think
the addresses from the dump will apply to your build only unless
you have some generic thing. This also seems like only part of
the information that could be dumped from an oops. It should
show
the Process:, Stack, Call Trace, registers and etc. May be you
can increase the debug level and do it again. See man page for
klogd man page options -c -d etc for getting full debug
information.
With more info you can probably at lest isolate the process and
function. If the oops gives a full dump to the logs then you
will have the stack and register and can probably nail the line.
d
[EMAIL PROTECTED] wrote:
--
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Matthew Willcock)
Subject: Re: Telnet stops working when Internet connection goes down
Date: Wed, 27 Jan 1999 13:00:35 GMT
On Wed, 27 Jan 1999 05:04:16 GMT, Tim Gibson <[EMAIL PROTECTED]>
wrote:
>I use a Redhat 5.1 server for gateway Internet access using IP Masq.
>Everything works great until the Internet connection drops. Then it
>takes either a veeery long tim to get a login porpt in a telnet session
>or it jsut times out. What's the deal? I suspect it might have
>something to do with DNS lookups but I am somewhat of a newbie and would
>know how to fix it.
Make sure your /etc/hosts files has entries for all the computers on
your local network - this solved the problem for me.
Matthew
Matthew Willcock
[EMAIL PROTECTED]
http://www.mwillc.free-online.co.uk
------------------------------
From: [EMAIL PROTECTED] (Darren Greer)
Subject: Re: PPP connects but I don't get out anywhere
Date: Thu, 28 Jan 1999 00:36:10 GMT
Could you post your /etc/ppp/options file please....then we might be
able to help you a little better.
DrGreer
On Wed, 27 Jan 1999 17:57:46 -0500, "Jim Orfanakos" <[EMAIL PROTECTED]>
wrote:
-->I am trying to setup PPP on my RedHat 5.1 system. The system dials,
-->connects, and stays connected. IFCONFIG PPP0 shows my ip address from my
-->ISP.
-->
-->I have configured my /etc/reslove.conf and ROUTE shows my routes that I hard
-->coded with my subnet and gateway....but still no luck. I cannot ping, surf,
-->etc.
-->
-->Any ideas?
-->
-->
--> ------------------------------------------------------
--> Jim Orfanakos
--> mailto:[EMAIL PROTECTED]
--> http://home.sprynet.com/sprynet/djo3
--> ------------------------------------------------------
-->
-->
------------------------------
From: "Victor Chou" <[EMAIL PROTECTED]>
Subject: DNS caching server on PPP router, elegant solution needed
Date: Wed, 27 Jan 1999 21:33:17 +0800
I have set up a Linux route to act as a gateway to the net for my other
computer. Managed to set up everything required (ipfwadm, bind and squid).
Then I realised the dns server will not be able to find the forwarders when
the there isn't a ppp connection to the ISP. So trying to be clever I added
a few lines to the ppp scripts to swap the bind config files around (one for
online, one offline) and run the command to restart the dns server. However
every so often things get completely messed up and the dns server ended up
not being killed properly.
The router is running:
RH5.0 (kernel 2.0.35)
BIND 8
and I'm calling the script that controls BIND inside /etc/rc.d ( I think )
usually bad things happen when I repeatedly do connects and
disconnects...perhaps something to do with the scripts being run in the
background?
------------------------------
From: Stefan Sundkvist <[EMAIL PROTECTED]>
Subject: Re: is my system overloaded?
Date: Wed, 27 Jan 1999 14:42:22 +0100
probebly not, my logs are like up to 50mb and i have not any problem.,
christo skrev:
> i've a linux 2.0.35 running on a double p2 with 128Mb ram and a quantm disk
> (server style) of 6 Gbyte.
> It performs web and mail server jobs, with apache and sendmail.
>
> After 6 mounth of hard working without any problem i noticed that response
> time when you try to connect to any port are getting longer.
>
> my 'messages' is about 6 Mb and my 'access.log' is almost the same.
>
> Could this slow down be caused by the fact it needs to write logs in this
> big files?
> In this case, can i backup this file and empty them without problems?
>
> If this this is a bad diagnosis, do anyone have suggestions?
>
> Thank you in advance and sorry for my poor english.
>
> christiano presutti
> istituto per il lavoro
> +39 051 6564241
> fax +39 051 6565425
> [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (NF Stevens)
Crossposted-To:
comp.os.linux.misc,comp.os.linux.portable,comp.os.linux.powerpc,comp.os.linux.setup
Subject: Re: This is Linux, not Windows, so why not superior flexibility AND
idiot-friendly?
Date: Sun, 24 Jan 1999 23:36:42 GMT
djbaz <[EMAIL PROTECTED]> wrote:
>Similar story here............
>Mine involves supporting vendors, localisation and tech support vendors. A lot of
>them use FTP to send builds and localised apps to our servers and many have extreme
>difficuly even understanding the directory structure of the FTP server. Some don't
>want to use the most common FTP clients as they find them too 'difficult to use'.
>Most of my time at work is spent trouble shooting connection problems and trying to
>simplify issues for idiot users when I should be concentrating on other important
>stuff.
>
>If the majority of users I encounter have difficulty using win95. What would they be
>like using Linux. For Linux to succeed in a desktop environment it would need to be
>simplified for Joe Soap users.
If the users were running any form of unix you could give them a
script which would control the ftp session from start to finish. I'd
probably use expect, but there are others available. You could
program the script to make intelligent decisions based on file
names or extensions or even check strings within the files to
be uploaded. The uses would never have to know anything
about your directory structure; all they'd have to know is
the name of the script.
A flashy gui front end may be great for surfing an unknown
site but for regular tasks which can be automated command
line unix wins hands down.
Norman
------------------------------
From: David Efflandt <[EMAIL PROTECTED]>
Subject: Re: Adding Routes
Date: Thu, 28 Jan 1999 00:42:10 GMT
Reply-To: [EMAIL PROTECTED]
On 1/27/99, 2:17:53 PM, William Groombridge <[EMAIL PROTECTED]> wrote=20
regarding Adding Routes:
> I have a few Linux machines with 2 NICs, and have had to add a couple =
of
> static routes like
> route add -net 10.0.0.0 netmask 255.0.0.0 gw 10.10.10.1
> for example... anyway.. whenever I reboot the machine, the routes are
> gone, and was wondering how to make them semi-permanent???
> Should I just add the route add command to my rc.local file or is=20
there
> a proper way to do it.. thanks muchly,
> Bill
I have done that in rc.local. If you end up with a default route, you=20
should also likely delete that if you have proper -net routing=20
(especially if you want to use defaultroute option for pppd). =20
Otherwise, RedHat has a nettool in X for setting routes.
------------------------------
Date: Wed, 27 Jan 1999 13:50:07 +0000
From: Matthieu Buchs <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: PC linux accessible through Mac
Hi,
I am currently using Linux for some calculations. I will then treat
results on a Mac, and I need to see the PC linux on the Mac through
Appleshare. I have heard (and seen) that it is possible, but I don't
know how to do it.
Can anybody help me? It would be king!
Regards,
Matthieu
--
==============================
Matthieu Buchs
Department of Inorganic Chemistry
University of Fribourg (Switzerland)
Perolles
Tel: ++41 26 300 87 49
Fax: ++41 26 300 97 38
mailto:[EMAIL PROTECTED]
http://sgich1.unifr.ch/ac/phd/mbuchs/
------------------------------
** 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
******************************
