Linux-Networking Digest #771, Volume #11 Sat, 3 Jul 99 15:13:39 EDT
Contents:
Re: linux as router to self - colo setup (Rich Roth)
Re: RedHat or Caldera? (Remote administration via telnet.) (Dave Edick)
Re: Why not C++ (John Stevens)
Re: zmodem with cu ("Gene Heskett")
Re: Combo cards - coax or 10bT (Dave Edick)
Re: Netscape 3.01 and java (Marc Mutz)
Kernel 2.2.10 + VPN ("Dirk Leas")
Re: Why not C++ (Larry Blanchard)
RPC: sendmsg returned error 105 (Garry Wright)
Re: 2.2.10 upgrade => broken networking (mist)
Re: Network Card Collisions (MadHatter)
Samba +Win95 probs (BorgholdR)
@ Home and Redhat 6 ^ Networking problem ("Anthony M. Abate")
Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft Retest
News (Anthony Ord)
----------------------------------------------------------------------------
From: Rich Roth <[EMAIL PROTECTED]>
Subject: Re: linux as router to self - colo setup
Date: Sat, 03 Jul 1999 13:19:52 -0400
John Bell wrote:
> Rich Roth wrote:
> >
> > (BTW, if you want a challege there is an NT box next to our linux box
> > that is doing this fine (different but similer set of IP addresses
> > but since NT is so visual we can't make sense out of the routing table.)
> >
>
> ???
>
> Open up a CMD.EXE shell on NT. Run the 'route' command.
> Read what it says about the routing table. Mimic :-).
Sorry, not helpful - and a poor Linux response <g>
Actually that was the first thing we did - there is something else in how the
interfaces are configured.
--
Rich Roth On-the-Net
[EMAIL PROTECTED] http://www.east.on-the-net.com
~~~ Add Instant Depth to your Website from www.i-depth.com ~~~
~~~ Adding depths to Web presences and Internet providers ~~~~~
------------------------------
From: [EMAIL PROTECTED] (Dave Edick)
Subject: Re: RedHat or Caldera? (Remote administration via telnet.)
Reply-To: [EMAIL PROTECTED]
Date: Sat, 03 Jul 1999 17:27:27 GMT
If you consider RedHat's configuration files to be so arcane, I'd recommend
that you get acquainted with Linuxconf. RedHat ships with it and you can
run it via telnet or over a web browser. Either will work fine over a 56k
connection.
Personally, I consider RedHat to be pretty well organized in the conf file
department. They keep data outside of the startup scripts and in a standard
location (/etc/sysconfig). I can't speak for Caldera since I've never used
it.
For the record, I use RedHat and Solaris at work, but run SuSE at home.
>Krazy Kat wrote:
>
>> I was wondering which of Linux distributions (aside from Slackware;
>> sorry!) would be best for remote administation via telnet/ssh?
>> I want easy upgrade paths (RPM), easy setup (no dselect), and
>> files that I can edit by hand - or a command-line tool that
>> will modify their contents for me. If RedHat is still the way
>> it was when I used 5.0 (networking, etc., can only really be
>> configured from the control panel, unless you somehow figure
>> out the more-arcane-than-usual text configuration files) then
>> it's out of the question. This box will be across the street,
>> and needs to be easily administered from a command-line.
>>
>> (We have a 56k leased line between the two locations, and
>> I'm *not* going to run X Window System over that.)
>>
>> Thanks much!
>>
>> - David
------------------------------
From: [EMAIL PROTECTED] (John Stevens)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: Why not C++
Date: Sat, 03 Jul 1999 17:25:46 GMT
On 1 Jul 1999 20:47:54 -0700, david parsons <[EMAIL PROTECTED]> wrote:
>In article <[EMAIL PROTECTED]>,
>Bruce Hoult <[EMAIL PROTECTED]> wrote:
>
>>Dylan is not several hundred percent slower than C as Java is.
>
> Have you benchmarked Java vs C on the same machine? (No, I don't
> mean benchmarking C on a machine vs Java running on a p-machine
> on that machine); there's certainly nothing in the design of
> Java that would make it much slower than C on the same machine.
Excuse, but:
Java, even compiled to machine code, will be slower than C or C++.
Java requires a run time system (ala Objective-C), making it both more
dynamic, and slower.
In fact, Java is more like Objective-C, except for the truly
horrible syntax, than C++.
John S.
------------------------------
Date: 03 Jul 99 12:42:03 -0500
From: "Gene Heskett" <[EMAIL PROTECTED]>
Subject: Re: zmodem with cu
Unrot13 this;
Reply to: <[EMAIL PROTECTED]>
Gene Heskett sends Greetings to Jonathan Guthrie;
[...]
JG> According to the file zmodem.txt that comes from Omen, Inc.'s
JG> Website
JG> (actually, I got them first from the BBS "lo these many years
JG> ago") "ZMODEM was developed for the public domain under a Telenet
JG> contract. The ZMODEM protocol descriptions and the Unix rz/sz
JG> program source code are public domain. No licensing, trademark,
JG> or copyright restrictions apply to the use of the protocol, the
JG> Unix rz/sz source code and the ZMODEM name."
JG> This conflicts with what Chuck Forsberg says about rz/sz on his
JG> Web site which says that that is copyrighted code whose use is
JG> restricted. Anyway, the description of ZMODEM is that document,
JG> the "xymodem.txt" file, and zmodem.h from the rz/sz source code.
JG> For your convenience, I have placed xymodem.txt, zmodem.txt, and
JG> rzsz.zip at ftp://ftp.brokersys.com/pub/zmodem
:-)
JG> With that said, rz/sz really doesn't help that fellow much.
JG> Instead, he needs the lrzsz package and the man pages therein.
JG> Further, he has to figure out how to run programs locally with
JG> cu. (I've not used cu in years.)
Well, I must confess that my last contact with it, other than useing it
against the seriously broken versions we have on the amiga over a null
modem cable, (busted perms handling anybody?) was compiling it for os9
level 2 on a trs-80 color computer 3, where it ran at about 380 cps due
to the character oriented nature, and its call to the crc routine for
every character handled. That was version 3.36, ages ago now. I got it
up to 700 cps on a good day with some assembly code plugged in to
replace his crc code in C. Had that been 'packetized' properly so that
the crc was called at the end of the packet to do the whole packet
before returning the ack/nak, the speed could probably have been
quadrupled, this on a machine with a 1.79mhz cpu clock! But, I didn't
feel I was comfortable enough walking around in somebody elses code to
do that. It might have been a piece of cake for some, but that code is
just big enough to start hiding its trees in the forest to me.
That version still had the 'pay me xx (20?) bucks, and send me your code
for approval before you can take out the closing splash and 20 second
timekiller' copyright notices in the sources. A short time killer on
that machine was just another day at the office, so thats the version
the os9 user world is useing yet today. You can get it, with those
sources, by annonymous ftp from <os9archive.rtsi.com>, up the os9-6x09
tree someplace. My name, or the handles I used on delphi over the
years, is in that archive many times.
Yeah, well, time does march on, and I can be safely ignored in most
cases. But, I do like to comment on those parts where "I've been there,
and done that". Can I plead Old Fart Syndrome? :-)
Cheers Jonathon, Gene
--
Gene Heskett, CET, UHK |Amiga A2k Zeus040 50 megs fast/2 megs chip
Ch. Eng. @ WDTV-5 |A2091,GuruRom,1g Seagate,CDROM,Multiface III
|Buddha + 4 gig WDC drive, 525 meg tape
|Stylus Pro, EnPrint, Picasso-II, 17" vga
RC5-Moo! 690kkeys/sec isn't much, but it all helps
email gene underscore heskett at iolinc dot net
--
------------------------------
From: [EMAIL PROTECTED] (Dave Edick)
Subject: Re: Combo cards - coax or 10bT
Reply-To: [EMAIL PROTECTED]
Date: Sat, 03 Jul 1999 18:09:34 GMT
I can't speak for the clone cards. But there are two ways to change the
media on the 509.
1) Hack the driver source. The driver doesn't have an option to change
media, unfortunately. It reads the EEPROM on the card and sets the media
accordingly. It would be relatively easy to just hack the source to ignore the
EEPROM setting and set it to coax. Or even better, change the driver to
allow media type to be specified and submit it to the kernel developers.
If you want to try this, but don't feel up to modifying the code yourself,
here's a quick, nasty, and untested hack on the 1.16 driver that ought to force
the media type to coax. Find the 3c509.c source file and either use patch
or delete the lines manually with an editor. Then recompile the kernel.
Be sure to keep a copy of the original source in case it doesn't work.
461d460
< if (dev->if_port == 3)
464,468d462
< else if (dev->if_port == 0) {
< /* 10baseT interface, enabled link beat and jabber check. */
< EL3WINDOW(4);
< outw(inw(ioaddr + WN4_MEDIA) | MEDIA_TP, ioaddr + WN4_MEDIA);
< }
876d869
< if (dev->if_port == 3)
879,883d871
< else if (dev->if_port == 0) {
< /* Disable link beat and jabber, if_port may change ere next
open(). */ < EL3WINDOW(4);
< outw(inw(ioaddr + WN4_MEDIA) & ~MEDIA_TP, ioaddr + WN4_MEDIA);
< }
2) Use the DOS configuration utility. It sets the EEPROM, so the existing
driver will respond to it.
On Sat, 3 Jul 1999 11:22:27 -0400, Drew M. Mooney <[EMAIL PROTECTED]> wrote:
>How do you tell linux which media to use on a combo card if it can't sort it
>out for itself?
>
>Cards in question are both 3Com 509b ISA and some off-brand PCI combos.
>[that work fine configured as NE2000 PCI]
>
>They both appear to default to 10bT, and I've got a coax backbone to support
>in a couple of days.
>
>TIA
>
>-Drew-
>
>
------------------------------
Date: Sat, 03 Jul 1999 20:20:11 +0200
From: Marc Mutz <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.x
Subject: Re: Netscape 3.01 and java
Gene Heskett wrote:
>
> MK> Has anyone succeeded in getting Netscape 3.01 to stay
>
> NS3.01? I'll be kind, and just say thats an awfully old version.
> Its up to 4.61, and freely downloadable from their site.
>
Wow, Gene was a _real_ help, wasn't he? (BTW: he said he was working on
a 486, did _you_ work with communicator 4.xx on a 486?)
Anyway, if you want to stick to the 3.xx series of netscape, check out
3.04. It's pretty stable. If that does not work, check the hint someone
else gave you before me... :-)
Marc
--
Marc Mutz <[EMAIL PROTECTED]> http://marc.mutz.com/
University of Bielefeld, Dep. of Mathematics / Dep. of Physics
PGP-keyID's: 0xd46ce9ab (RSA), 0x7ae55b9e (DSS), 0x31748570 (DH)
------------------------------
From: "Dirk Leas" <[EMAIL PROTECTED]>
Subject: Kernel 2.2.10 + VPN
Date: Sat, 03 Jul 1999 16:13:08 GMT
I've read through the "Linux VPN Masquerade" document and am having trouble
identifying the steps necessary to build a VPN between two networks given
to linux 2.2.10-based firewalls (happens to be RH 6.0+kernel upgrade). Most
of the content caters to 2.0 kernels.
There's lots of options while building the kernel, but I can't find any
details on them (e.g. IP: tunneling, IP: GRE tunnels over IP, etc.). I've
read the help in the kernel build front-ends (make xconfig), but still can't
find the details to make it all come together.
Here's the idea...
eastCoast:
[inside] <-> eth0 (172.31.1.0/255.255.255.0)
[outside] <-> eth1 (24.1.244.100/255.255.242.0)
westCoast:
[inside] <-> eth0 (172.31.2.0/255.255.255.0)
[outside] <-> eth1 (24.1.211.100/255.255.248.0)
I need to connect using a VPN between eastCoast's 24.1.244.100 and
westCoast's 24.1.211.100 across the internet. They can already ping each
other, and both internal networks can successfully ipmasq out to the
internet.
I'm trying to securely route samba, and some of the other standard protocols
(http, ftp, etc.). Any references or examples would be GREATLY appreciated!
I'd be happy to start a "Linux Kernel 2.2.x VPN for dummies" if I can bring
this together (there definitely seems to be a quite a bit of traffic asking
similar questions).
TIA,
Dirk
------------------------------
From: Larry Blanchard <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: Why not C++
Date: 03 Jul 1999 09:13:23 PDT
Reply-To: [EMAIL PROTECTED]
Enough already!
Can we let this thread die now? It's had more arguments than Northern
Ireland!
--
Larry Blanchard - Old roses, old motorcycles, and old trains
Homo Sapiens is a goal, not a description.
------------------------------
From: [EMAIL PROTECTED] (Garry Wright)
Subject: RPC: sendmsg returned error 105
Date: 3 Jul 1999 17:19:04 GMT
After a samba failure caused by putting a mounted machine to sleep
the attempt to unmount samba lead to the errors reported from the
kernel
RPC: sendmsg returned error 105
neighbor table overflow
This continues ad nauseum, even after a reboot.
Also I observer there is no lo device present. The failure to
smbumount kept reporting io failure related to the mount point
prior to the reboot.
Can someone point me in the right direction?
[EMAIL PROTECTED]
------------------------------
From: mist <[EMAIL PROTECTED]>
Subject: Re: 2.2.10 upgrade => broken networking
Date: Sat, 3 Jul 1999 17:22:53 +0100
Reply-To: mist <new$[EMAIL PROTECTED]>
Dave Warner <[EMAIL PROTECTED]> scribed to us that -
>Did both. The problem has to be more fundamental since even ping
>doesn't work; I think ping is wire -> kernel -> inetd -> kernel -> wire,
>and maybe even inetd isn't involved.
>
I'm not sure quite what you mean by that description of ping. Ping
works by sending specific ICMP messages out and waiting for replies.
For it to work over a VMware system you'll need to make sure that you
have enabled ICMP masquerading in the kernel.
--
Mist.
------------------------------
From: MadHatter <[EMAIL PROTECTED]>
Subject: Re: Network Card Collisions
Date: Sat, 03 Jul 1999 17:25:19 GMT
I understand that collisions are a natural occurence. however, i have a
small network setup at home with 2 hosts, both running linux. when i
try and FTP files, 100KB or more, i get constant collisions. since both
hosts wait for a certain amount of time, i get heavy delays and end up
transferring data at about 1kb/s, which is slower than a modem. can
anyone tell me why that happens? how do i fix that? both hosts are
conected to a hub and there are no settings on the hub. also, how do i
figure out how to set the duplex settings for my nics?
please help
In article <[EMAIL PROTECTED]>,
Frank Sweetser <[EMAIL PROTECTED]> wrote:
> Martin Ng <[EMAIL PROTECTED]> writes:
>
> > Hi,
> >
> > Does any one know about the cause of collision of network card? I
can
> > see a number of collision when issuing 'ifconfig'.
> > I suspect that it is the cause of why I telnet to it very slowly.
>
> when two hosts try to send a packet out at the same time, rather than
have
> two bad packets transmitted, a collision is generated. both hosts
detect
> this, wait a random amount of time, and then try again. however, the
only
> reason that collisions would be causing a problem is if a) the
segement is
> *totally* saturated, b) there's an underlying problem (ie, dead card,
> frayed cable), or c) the hub port you're plugged into is set to full
duplex
> while your network card is set to half.
>
> --
> Frank Sweetser rasmusin at wpi.edu fsweetser at blee.net | PGP key
available
> paramount.ind.wpi.edu RedHat 5.2 kernel 2.2.5 i586 | at public
servers
> Sometimes we choose the generalization. Sometimes we don't.
> -- Larry Wall in <[EMAIL PROTECTED]>
>
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (BorgholdR)
Subject: Samba +Win95 probs
Date: 03 Jul 1999 17:43:51 GMT
Hallo,
ich habe hier ein gro�es Problem mit dem Datentransfer: PC --> Linux : in
Sekunden, Linux --> PC in Minuten ! Ich habe Linux 6.1 (Suse) und auf dem
anderen Rechner Win95 C. Per FTP klappt es Prima in beiden Richtungen.
I have a great Problem with Datatransfer between Linux and Win 95. In the
direction Win 95 -> Linux = fast. The other direction is very slow (2MB = two
Minutes) FTP is running fast in both directions, any help ? Linux is installed
with new kernel and lastest version of Samba
Many thanks for your help
�ber eine kurze Mail w�rde ich mich sehr freuen
Tschau Reiner
------------------------------
From: "Anthony M. Abate" <[EMAIL PROTECTED]>
Subject: @ Home and Redhat 6 ^ Networking problem
Date: Sat, 03 Jul 1999 18:39:12 GMT
Ok, so before I commit ritual suicide I need some help. I have a D-Link 220
NIC (ISA) in my Linux box. I have a small LAN setup in my house with NT and
95 on the other machines. My Linux machine can see the NIC, ping itself but
not get to the Gateway. So here's more info I'm not using DHCPCD, I have a
static address. I've double checked all my settings they are matching my
other machine setups. I've tried DHCPCD with the -h switch. I've edited
all the files to ensure that the information is correct. Yet all of this to
no avail, I try to ping my DNS server I get Network unavailable:.
Additionally, I'm seeing a lot of TX errors and dropped packets when I try
to ping any IP but my own. I set one of my Windows machines up to use the
IP and then started the Linux bow ..... the 95 machine registered an IP
conflict .... So ... I'm lost .... Could someone help me figure this out
TIA
Tony
------------------------------
From: [EMAIL PROTECTED] (Anthony Ord)
Crossposted-To:
omp.os.ms-windows.nt.advocacy,comp.os.linux.advocacy,comp.infosystems.www.servers.unix,comp.os.linux.misc
Subject: Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft
Retest News
Date: Sat, 03 Jul 1999 16:40:17 GMT
On Thu, 1 Jul 1999 11:25:38 -0700,
[EMAIL PROTECTED] ("Bob Taylor") wrote:
>In article <[EMAIL PROTECTED]>,
> De Messemaeker Johan <[EMAIL PROTECTED]> writes:
>> Chad Mulligan wrote:
>>
>>> >And how do you know this actually happened? Were you present? You read
>>> >it in a book written by an American hater? On behalf of the Americans
>>> >who died saving your sorry ass in WWII, I *demand* a retraction and
>>> >appology for such a vicious attack!
>>>
>>> I would have to agree with Mr taylor on this one. The stories I've heard, some
>first
>>> hand from German POW's in the US, friends of my Granddad who was a Latvian pressed
>>> into German service, all said they were treated well and some even remained friends
>>> with their jailers.
>>
>> True, that happened. But it wasn't always like that. Here in Belgium, there was a
>> concentrationcamp named 'Breendonk'. When the americans arrived, they torchered the
>> germans. That's a historical fact. And there's a lot more. But what do we expect ?
>It was
>> war, with all pleasant and unpleasant stuff ... sort of ...
>
>If by "torchered"
Oh look! A spelling flame! Tell me Mr Taylor, how many
languages do you speak? I would rather he wrote imperfect
English than perfect French.
>you mean forced to tour the concentration camps, then
>I will agree.
So the guards who ran the concentration camp called
"Breendonk" would be tortured by being forced to tour
concentration camps? I would have thought it would become a
bit samey during the first tour, never mind the rest.
>Otherwise, this is another vicious attack with no
>foundation in truth. Historical fact my tired ass!
I constantly find it amazing that there is never any doubt
in some people's minds about anything. But don't worry, I'm
sure if you look hard in the Yellow Pages there will be
someone who reloads expended clue-sticks.
Regards
Anthony
--
=========================================
| And when our worlds |
| They fall apart |
| When the walls come tumbling in |
| Though we may deserve it |
| It will be worth it - Depeche Mode |
=========================================
------------------------------
** 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
******************************
