Linux-Networking Digest #558, Volume #10 Fri, 19 Mar 99 15:14:01 EST
Contents:
Re: ECommerce on LINUX {CAW} ([EMAIL PROTECTED])
Recommend Fast Ethernet Card (Jon Slater)
Re: "Network unreachable"..... help!! ("Leopold Toetsch")
Re: Samba as WinNT PDC (Iain O'Cain)
Redundant Linux firewall with autotakeover (Lars Marowsky-Bree)
Re: pppdialin/#of_rings/voice_calls (Upali Weerasinghe)
Re: Cablem Modem Questions ("Eugene")
Re: The truth about the Pentium III chip and ID --- **boycott info** (Aleksandar
Milivojevic)
Re: DNS entry for dynamic IP address ([EMAIL PROTECTED])
Re: Network shuts down on a regular basis (Upali Weerasinghe)
Re: ne.o & 3c509.o compiled ??? (Enbugger)
Re: What is the best Linux to install? (Marco Danti)
Re: Delaying at the sendmail initialization (Villy Kruse)
Re: ne.o & 3c509.o compiled ??? ("��d�W")
Re: Too Frequent Dial ( DNS/Ipfwadm/Diald) ("Phantom")
Re: The truth about the Pentium III chip and ID --- **boycott info** ("Quantum
Leaper")
Re: The truth about the Pentium III chip and ID --- **boycott info** (Tim Roberts)
Re: DNS entry for dynamic IP address ([EMAIL PROTECTED])
Urgent! NFS woes... (W. thomas)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Subject: Re: ECommerce on LINUX {CAW}
Date: 18 Mar 1999 17:54:01 GMT
Reply-To: [EMAIL PROTECTED]
In <7cpc64$[EMAIL PROTECTED]>, "The Lone Scribe" <[EMAIL PROTECTED]>
writes:
>[EMAIL PROTECTED] wrote in message
><7cogd3$1gc$[EMAIL PROTECTED]>...
>>Ok gang, what do I need to have fired up to run an ecommerce server with
>apache
>>and Linux?
>
>
>Hire someone who knows what they're doing and can program and set up (and
>maintain) the site for you. There are no 'ecommerce is on now' packages
>included in most linux distros to just fire up and start making money.
Clearly a good point. However, this isn't an instant project. and I always like
to know a bit about what i'm doing, even if I hire somebody to do it. Otherwise
you never know if you are getting take.
I probably should have phrased my question differntly.
What I was looking for were sources/ references that would outline
what it took to have an ECOMMERCE site using Linux/Apache.
sorry for the confusion.
thanks for the reply
Chris
------------------------------
From: Jon Slater <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.hardware
Subject: Recommend Fast Ethernet Card
Date: Thu, 18 Mar 1999 11:19:21 -0700
Can anyone recommend a fast PCI Ethernet card for Linux?
Thanks!
--
Jon D. Slater QualComm Inc.
[EMAIL PROTECTED] 6150 Lookout Road
Phone: (303) 247-5037 Boulder, Colorado
Fax: (303) 247-5167 80301
------------------------------
From: "Leopold Toetsch" <[EMAIL PROTECTED]>
Subject: Re: "Network unreachable"..... help!!
Date: Fri, 19 Mar 1999 18:48:52 +0100
Steph Hepburn wrote in message <7ctf1m$447$[EMAIL PROTECTED]>...
>
>But in dialing up to my ISP nothing happens
>
>in /var/log/messages I got:
>
>connect ppp0 <--> /dev/ttyS1
>Local IP address 12.34.56.78
>Remote IP address 23.45.67.89 -- or something like that anyway
>
So physically everything looks ok.
Did you configure route.conf ?
Did you setup a route for your ppp0 ?
# route add default gw $REMOTEIP dev ppp0
s. PPP-HOWTO-14.html or PPP-FAQ-7.html
hope this helps
leo
------------------------------
From: [EMAIL PROTECTED] (Iain O'Cain)
Subject: Re: Samba as WinNT PDC
Date: 19 Mar 1999 12:07:01 -0700
Reply-To: [EMAIL PROTECTED]
[posted and mailed]
Andrea Borgia <[EMAIL PROTECTED]> wrote:
>[EMAIL PROTECTED] (Iain O'Cain) wrote:
>
>>Actually, it's just "unsupported" at this point, but it works fine. I'm
>>using Samba 2.0.3 as a PDC for a small office network. Not all the PDC
>>functionality is supported, but for single-logon and roaming profiles,
>>it seems to be the easiest way to go.
>
>BTW, do you mind telling me how you did it? I tried following the
>(very incomplete) instructions included in the docs, but got nowhere.
Grab the NT-Domain FAQ from the Samba docs. Yes, it's very incomplete,
since it's not yet officially supported. The basic idea tho' is just to
set up Samba with MS "encrypted" password support, and set the options
to have it advertise itself as the domain controller. Unless you're
willing to wade through the confusing docs and get yourself really
familiar with the current state of Samba, I'd recommend you wait for
the official support, though.
>TIA,
YW. ZQRVX. EIEIO?
- Iain
------------------------------
From: [EMAIL PROTECTED] (Lars Marowsky-Bree)
Subject: Redundant Linux firewall with autotakeover
Date: 19 Mar 1999 19:19:02 GMT
Good morning,
I need to implement a setup which involves a Linux firewall solution.
All involved systems need to be redundant, which is not the problem (a
Cisco LocalDirector certainly helps ;)
But the secondary Linux packetfilters need to take over when the primary
fails. Automatically, on both sides, reliably. If the other box comes back
up, it of course needs some check if the other side is working and then
not go active.
Both boxes have 2 ethernet cards. On each segment, there is a virtual IP
as the gateway used from the inside.
I have thought about running a conversation across both cards between them,
and pinging the default gateway on each side. They would exchange data about
their wellbeing and if they can ping the default gw, maybe CPU, disk failures
etc too.
This would allow both boxes to detect which interfaces have failed, or whether
itself or the other box went down completely. (At least I think the
information gathered should suffice)
The box which failed would then halt itself and warn, while the other uses
"fake" to take over. If one of the boxes reboots, it runs this check after
boot and tries to determine if itself is defunc (interfaces etc) or if the
other box is active, in which case it does not take up the virtual gateway
IP.
Would this a) work, b) has someone already implemented something like this?
Sincerely,
Lars Marowsky-Br�e
--
Lars Marowsky-Br�e
Network Management
teuto.net Netzdienste GmbH - DPN Verbund-Partner
------------------------------
From: Upali Weerasinghe <[EMAIL PROTECTED]>
Subject: Re: pppdialin/#of_rings/voice_calls
Date: Fri, 19 Mar 1999 14:36:38 GMT
In this case I would use mgetty, it's more advance than uu_getty. You
can setup modem to answer 1 to 9 rings or second call.
means ring_back, please read info. at your Linux prompt type info then
press s , mgetty and enter
all the configuration files should be in /etc/mgetty+sendfax
also you can tell modem not to answer when ever you want, over here you
have to create /etc/nologin.ttyS2 file and use crontab to remove it or
put it back
Upali
Harry Park wrote:
> I am using uu_getty(2.07j-3) for dialin connectivity to my Linux 5.1
> machine.
> I have 2 questions:
>
> Can I make uugetty ignore voice phone calls on the line? As it is now
>
> when
> someone calls the house the modem trys to talk. Useful for
> telepromoters
> but
> not for family members.
>
> How do I set the number of rings to 5 rings?
>
> Included is my /etc/conf.uugetty.ttyS2 file
>
>
> INITLINE=cua2
> ALTLOCK=cua2
> TIMEOUT=60
> INIT="" AT\r OK\r\n
> WAITFOR=RING
> CONNECT="" ATA\r CONNECT\s\A
> DELAY=1
> DEBUG=010
>
> TIA
>
>
> --
> Harold Park---System administrator
> remove nospam from return address for
> accurate email addressing
>
>
------------------------------
From: "Eugene" <[EMAIL PROTECTED]>
Subject: Re: Cablem Modem Questions
Date: Fri, 19 Mar 1999 19:27:15 GMT
if you have a 10baseT network and a hub, you can connect the cable modem to
the hub and use IP aliasing on the Linux box (i.e. a single interfrace
(eth0) will have 2 IP addresses).
If you have a 10base2 network, you need a second network card. That way eth0
will still be on the local network, eth1 will be connected to the cable
modem.
Second network card is cheaper then a hub anyway.
For more info about this go to www.linux.org/help and read cable modem howto
(it might be mini-howto).
I set up my cable modem with 2 network cards. The gateway is also a web,
ftp, and mail server, up and running 24/7. Check it out:
www.happypenguin.dhs.org
R. Brooks wrote in message <[EMAIL PROTECTED]>...
>I have RH 5.1 running 2.2.2
>I have IP MASQ running with firewalling.
>It works great for a ppp connection.
>
>Our local provider offers a cable modem.
>Hybrid model.
>I believe they are using DHCP.
>
>Now, do I have to install a second ethernet card?
>(One for the cable modem and one for internal network.)
>
>If so, should I use a DCHP client on the second one.
>
>I basically what to remove the ppp0 and replace it with a cable modem.
>
>Any help would be appreciated.
>
>Thanks,
>
>Randall Brooks
>--
>_____________________________________________
>Randall Brooks
>[EMAIL PROTECTED]
>http://www.glendinningprods.com
>Senior Engineer
>GMP (843) 399-6146 FAX (843) 399-5005
>
>
------------------------------
From: Aleksandar Milivojevic <[EMAIL PROTECTED]>
Crossposted-To:
comp.lang.perl.misc,comp.lang.python,comp.lang.tcl,comp.mail.sendmail,comp.os.linux.setup,comp.os.ms-windows.advocacy
Subject: Re: The truth about the Pentium III chip and ID --- **boycott info**
Date: 19 Mar 1999 15:10:07 +0100
Stuart R. Fuller ([EMAIL PROTECTED]) wrote:
> Some Sun boxes return unique (at least unique between the few I
> looked at) values, and they are not related to the IP address.
> However, the Sun boxes do not constitute "every unix box in
> existance[sic]".
Every Sun box have unique hostid. It is written in NVRAM. For more
details see Sun NVRAM/hostid FAQ at:
<http://www.squirrel.com/squirrel/sun-nvram-hostid.faq.html>.
--
Aleksandar Milivojevi� | [EMAIL PROTECTED] | http://jagor.srce.hr/~alex/
Opinions expressed herein are my own.
================================ooooO=Ooooo==============================
Real Users never know what they want, but they always know when your
program doesn't deliver it.
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: DNS entry for dynamic IP address
Date: Fri, 19 Mar 1999 14:29:04 GMT
I was under the impression that a DNS name is read backwards. For example: if
I registered "enon.oh.us" with a static ip of 198.6.245.100 and made it a DNS
server, computerst that went to lookup "cheetah.enon.oh.us" at a dynamic I{
of something like 207.90.118.70 it would go to enon.oh.us at 198.6.245.100 to
see where cheetah is in that domain. Since the caching would be controlled on
that machine and I control both, why would TTL and caching be a problem?
Or,do individual DNS servers cache everything, including subnets in DNS
names?
Terry
[EMAIL PROTECTED]
In article <7ct2ct$dad$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> In article <7csfnd$gu$[EMAIL PROTECTED]>,
> "Terry Mathews" <[EMAIL PROTECTED]> wrote:
>
> > Here's the scoop. I want to put this Linux box (Red Hat 5.2) on a static IP
> > address as an Internet server, e-mail server, yadda yadda yadda. Here's the
> > trick. I'd like to be able to have 2 computers log into the machine and have
> > them assigned DNS names. Both machines connect to the internet via modems
> > and get dynamic ip addresses.
>
> Dynamic IP's don't "work" with DNS. In a way they do, but not in a way you're
> probably going to find useful.. more in a bit.
>
> > They aren't in the same subnet as the server will be.
>
> That doesn't matter. So long as you have the authoritative name server(s) for
> your domain, you can point the IP addresses at anything you like (including
> other people's IP addresses, if you had the odd desire to do so).
>
> > [Example with bogus IP addresses deleted]
> > Is this possible?
>
> Yes and no.
>
> Here's a quick primer on how DNS works.
>
> 1) I type "skunx.reconnoiter.com" into my browser. 2) Netscape contact my
> primary DNS and asks it to look up "skunx.reconnoiter.com" 3) My primary DNS
> checks to see if it has "skunx.reconnoiter.com" in its cache -- this is what
> makes dynamic IP's mostly unusable, because what's in the cache may not be
> what IP you're using today. 4) My primary DNS then returns the IP (if it is
> in cache) or sends the request further "up" the chain (checking with root
> nameservers and whoever runs "reconnoiter.com").
>
> As you can see, the only step in the process that you'll have any control over
> is step 4 -- and that assumes no caching is going on, which isn't always true
> (in fact, it's almost never true).
>
> You could set the TTL (TimeToLive) for your DNS entries to something very
> low, but that's not a very nice thing to do at all, and in any event there
> are many many DNS machines out there that ignore the TTL anyway (mostly
> because of jerks that set it too low).
>
> If you happen to know what the *range* of IP addresses will be, you can
> assign a unique name to each of them. When you're assigned an IP address,
> you'll also end up being "assigned" a DNS name. This may or may not work for
> what you want to do (probably not), but it's about the closest solution you
> can get.
>
> Really, the best thing to do is to ask your ISP for more static IP's.
>
> -Bill Clark
>
> > Please respond via e-mail. Thank you.
>
> If dejanews had a feature to do so, I might consider it.
>
> Then again, asking for email replies in usenet is generally considered rude,
> so maybe I wouldn't. :)
>
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
>
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: Upali Weerasinghe <[EMAIL PROTECTED]>
Subject: Re: Network shuts down on a regular basis
Date: Fri, 19 Mar 1999 14:42:24 GMT
make sure you have disable you routed and gated, you be fine
dude wrote:
> Hi there. I just finished setting up a linux box, acting as a gateway for a
> small LAN (plus doing several other small services) and everything went
> smoothly and without pain. Except for a very peculiar problem: Every 2 hours
> or so, all network connections are broken, both intra- and internet, no
> pings, no routes, no nothing. Most of the time the network kicks in after ca.
> 5-8 minutes. Twice I've had to restart inetd, and once routed. I don't know
> if there's a connection here, but when I look at /var/log/messages, I see a
> cluster of error messages appearing every 8-10 minutes: Mar 18 14:35:40
> picard inetd[7230]: linuxconf/tcp: bind: Address already in use Mar 18
> 14:35:40 picard inetd[7230]: auth/tcp: bind: Address already in use Mar 18
> 14:35:40 picard inetd[7230]: time/tcp: bind: Address already in use
> ...followed by identical messages for each tcp service; finger,imap, pop,
> shell, telnet, etc.
>
> I'm running RedHat 5.2 (2.0.36) on a compaq prosignia 200 (nics eth0:TLan,
> eth1: 3c905b)
>
> I'm completely stuck and utterly baffled.
>
> Any suggestions, tips, hints, ideas, places_to_look_for_error_messages,
> solutions, pizzaz, or well, just anything to get me started would be SO
> greatly appreciated.
>
> My opinions are my own, and have nothing to do with my employer.
>
> If God had wanted Man to be multitasking,
> He would have given him two hands.
>
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: Enbugger <enbugger@usa./NO-SPAM/net>
Crossposted-To: hk.comp.os.linux,hk.comp.pc
Subject: Re: ne.o & 3c509.o compiled ???
Date: Fri, 19 Mar 1999 22:35:23 +0800
I am not sure if you are using RedHat.
If you are, please refer to this document for
version 2.2 kernel upgrade:
http://www.redhat.com/support/docs/rhl/kernel-2.2/kernel2.2-upgrade.html
Hints:
Have you done:
make modules
and
make modules_install
?
"��d�W" wrote:
>
> Before, I haven't build the kernel.
> I need to prepare the file /etc/conf.modules:
>
> alias eth0 ne
> options ne io=0x300
> alias eth1 3c59x
>
> And today, I try to build the kernel.
> I have chosen the [Y] in both 3com NIC support and NE2000 support.
> So, I won't got this two modules(ne.o and 3c59x.o) any more during building.
> Then, the problem happened !!!
> On starting:
> ...
> Enable IPv4 packet forwarding.
> modprobe: can't locate module net-pf-5
> ...
> eth1: unknown interface. (NE2000 NIC change from eth0 to eth1)
> modprobe: can't locate module net-pf-5
>
> What's more, I try to copy back the files:
> ne.o to /lib/modules/2.2.3/net/ne.o
> 3c59x.o to /lib/modules/2.2.3/net/3c59x.o
>
> Then the system report:
> ...
> modprobe: can't locate module net-pf-5
> insmod: /lib/modules/2.2.3/net/3c59x.o: kernel module version mismatch
> /lib/modules/2.2.3/net/3c59x.o was compiled for kernel version
> 2.0.36 while this kernel is version 2.2.3.
> eht1: unknown interface.
> modprobe: can't locate module net-pf-5
> ...
>
> What should I do now, is that supports been compiled to kernel yet ???
> I have chosen [Y] for these 2 NIC support already !!!
> Would you please offer your Kindly Help ?
--
Enbugger
p.s. Remove /NO-SPAM/ for e-mail.
------------------------------
From: Marco Danti <[EMAIL PROTECTED]>
Crossposted-To:
alt.os.linux,comp.os.linux,comp.os.linux.misc,comp.os.linux.setup,comp.os.linux.redhat,alt.os.linux.slackware
Subject: Re: What is the best Linux to install?
Date: Fri, 19 Mar 1999 13:19:01 -0600
George Csahanin wrote:
>
> I've tried REDHAT and Slackware, and while setup of a "pretty" Linux system
> is the goal of Redhat, try compiling a new kernel. You can't, even if you
> tell it in the install to load all of what you think you need. They seem to
> want you to use a "stock" kernel, kinda like the blue sky and white clouds
> people. Do it their way. The system setup stuff in /etc/rc.d is a laugh. Try
> to customize anything, it's tough, get ready to spend time with grep looking
> for where stuff is.
That was not true for me. I could recompile my RH5.2 kernel all right,
and
I didn't even install all the things from the CD. As for the setup
scripts,
they are just that: scripts. Anybody can read and change them as needed.
Marco
------------------------------
From: [EMAIL PROTECTED] (Villy Kruse)
Crossposted-To:
comp.os.linux.help,comp.os.linux.misc,comp.os.linux.questions,comp.os.linux.setup
Subject: Re: Delaying at the sendmail initialization
Date: 19 Mar 1999 20:32:36 +0100
In article <5ZwI2.9$[EMAIL PROTECTED]>,
George Csahanin <[EMAIL PROTECTED]> wrote:
>I give stuff on the local 192.168 nets a single name, make sure it is in its
>own hosts file and the others hosts files on the net.
>
>Sendmail, configured as you have it will choke if it can't resolve its own
>hostname. Somehow.
>
And remember sendmail will no be satisfied until these hostnames are fully
qualified. Thus, if your host name is define this way in /etc/hosts
123.123.123.123 myhost
change this into:
123.123.123.123 myhost myhost.mydomain
just as was done with the localhost entry.
Villy
------------------------------
From: "��d�W" <[EMAIL PROTECTED]>
Crossposted-To: hk.comp.os.linux,hk.comp.pc
Subject: Re: ne.o & 3c509.o compiled ???
Date: Sat, 20 Mar 1999 03:35:51 +0800
Thank you for giving me the reference Homepage:
http://www.redhat.com/support/docs/rhl/kernel-2.2/kernel2.2-upgrade-5.html#s
s5.2
Actually, I tried to follow a book's instruction to build the kernel.
After having a look on the Homepage,I find that I miss 3 procedures:
make mrproper
cp /usr/src/linux-2.2.3/System.map /boot/System.map-2.2.3
# put the kernel in the /boot partition/directory.
cd /boot
rm System.map
ln -s System.map-2.2.3 System.map
mkinitrd /boot/initrd-2.2.1.img 2.2.1
Also, I cannot find 1). System.map-2.2.3
2). initrd-2.2.3 by using the command:
find / -name System.map-2.2.3 -print
How can I do this 3 procedures now ?
Will it be too late for me the above procedure now ?
I have chosen [Y] for 1). NE2000 NIC support
2). 3Com NIC support in the make xconfig procedure !
P.S.
I use NE2000(ISA) and 3Com 3c905(PCI) as eth0 and eth1, respectively.
Do I need to use the file /etc/conf.modules now ?
How can I know if there are really be compiled in the kernel image ?
When I enter dmesg | grep eth
it report:
eth0: 3Com 3c905 Boomerang 100baseTx at 0x6000, 00:60:97:a1:f7:9d, IRQ
11
P.S.
3Com(PCI) can be detected and change from eth1 to eth0.
>Can you actually use the ethernet by running the
>new kernel?
No, I don't think so.
I can notice eth0 when I enter ifconfig -a.
But actually, I use NE2000 NIC for my ethernet LAN.
So, the LAN is completely dead now :.<
Please keep Helping !!!
Thank you !
Enbugger ���g��峹 <7ctuhj$kss$[EMAIL PROTECTED]>...
>Have you re-built the initrd image by using
>mkinitrd ?
>
>If you have set
>[Y] in both 3com NIC support and NE2000
>[Y] Enable IPv4 packet forwarding
>when building the kernel, they are already
>compiled into the kernel image.
>
>You can check the eth drivers by
>dmesg |grep eth
>
>Can you actually use the ethernet by running the
>new kernel?
>
>
>
>HTH
>
>"��d�W" wrote:
>>
>> I have done this la !
>>
>> Enbugger ���g��峹 <7ctncl$l52$[EMAIL PROTECTED]>...
>> >I am not sure if you are using RedHat.
>> >If you are, please refer to this document for
>> >version 2.2 kernel upgrade:
>> >
>> >http://www.redhat.com/support/docs/rhl/kernel-2.2/kernel2.2-upgrade.html
>> >
>> >Hints:
>> >Have you done:
>> > make modules
>> >and
>> > make modules_install
>> >?
>> >
>> >"��d�W" wrote:
>> >>
>> >> Before, I haven't build the kernel.
>> >> I need to prepare the file /etc/conf.modules:
>> >>
>> >> alias eth0 ne
>> >> options ne io=0x300
>> >> alias eth1 3c59x
>> >>
>> >> And today, I try to build the kernel.
>> >> I have chosen the [Y] in both 3com NIC support and NE2000 support.
>> >> So, I won't got this two modules(ne.o and 3c59x.o) any more during
>> building.
>> >> Then, the problem happened !!!
>> >> On starting:
>> >> ...
>> >> Enable IPv4 packet forwarding.
>> >> modprobe: can't locate module net-pf-5
>> >> ...
>> >> eth1: unknown interface. (NE2000 NIC change from eth0 to eth1)
>> >> modprobe: can't locate module net-pf-5
>> >>
>> >> What's more, I try to copy back the files:
>> >> ne.o to /lib/modules/2.2.3/net/ne.o
>> >> 3c59x.o to /lib/modules/2.2.3/net/3c59x.o
>> >>
>> >> Then the system report:
>> >> ...
>> >> modprobe: can't locate module net-pf-5
>> >> insmod: /lib/modules/2.2.3/net/3c59x.o: kernel module version mismatch
>> >> /lib/modules/2.2.3/net/3c59x.o was compiled for kernel
>> version
>> >> 2.0.36 while this kernel is version 2.2.3.
>> >> eht1: unknown interface.
>> >> modprobe: can't locate module net-pf-5
>> >> ...
>> >>
>> >> What should I do now, is that supports been compiled to kernel yet ???
>> >> I have chosen [Y] for these 2 NIC support already !!!
>> >> Would you please offer your Kindly Help ?
>> >
>> >--
>> >Enbugger
>> >
>> >p.s. Remove /NO-SPAM/ for e-mail.
>
>--
>Enbugger
>
>p.s. Remove /NO-SPAM/ for e-mail.
------------------------------
From: "Phantom" <[EMAIL PROTECTED]>
Subject: Re: Too Frequent Dial ( DNS/Ipfwadm/Diald)
Date: Fri, 19 Mar 1999 15:30:21 -0000
I'm still looking at this one myself! I have it fixed in NT but am still
playing with 9x
It has to do with network neighbourhood browsing!
you will have to add lines such as:
ignore tcp tcp.dest=tcp.netbios-ssn
ignore udp udp.dest=udp.netbios-dgm
into your diald.conf & possibly also comment out/change
some entries in /usr/lib/diald/standard-filter
these tell diald to ignore file & printer sharing / microsoft networking
packets
or you could try:
ipfwadm -F -i deny -S 192.168.1.0/24 -D 0.0.0.0/0 [incorrect syntax next as
I'm not on my linux box] port 138
ipfwadm -F -i deny -S 192.168.1.0/24 -D 0.0.0.0/0 port 139
to stop smb packets from being forwarded - so they never get to diald.
Also try using:
tcpdump -i sl0
as this will tell you what packets are causing diald to connect.
craig somerville wrote in message
<01be7218$1d9f18e0$[EMAIL PROTECTED]>...
>e.g. at every boot the link comes up
>at frequently throught the day the link comes up and down after only around
>40 - 50 seconds of traffic. No Windows user is activly requesting access,
>no windows users are shown in the dialmon queue either ?
>
>I belive that it is related to DNS but cannot quantify it further.
------------------------------
From: "Quantum Leaper" <[EMAIL PROTECTED]>
Crossposted-To:
comp.lang.perl.misc,comp.lang.python,comp.lang.tcl,comp.mail.sendmail,comp.os.linux.setup,comp.os.ms-windows.advocacy
Subject: Re: The truth about the Pentium III chip and ID --- **boycott info**
Date: 19 Mar 1999 19:57:44 GMT
Curt Steger wrote in message <[EMAIL PROTECTED]>...
>
>
>Quantum Leaper wrote:
>
>> Michael Barnes wrote in message <[EMAIL PROTECTED]>...
>> >Every MODEM has a MAC address also...so your friends pc is nicless, but
not
>> >macless
>>
>> Interesting is over 15 years of using modems, 300 baud to 56K modems
>> (hopefully a Cable or DSL modem in about a year or so), I have NEVER
heard
>> that they have a MAC address? So what command or how do you get the MAC
>> address of a modem? Does this only apply to Mac modems or all modems?
One
>> other question, why would a modem need a MAC address?
>
>Only if it is running on an Apple/Mac!
>
Why would it need a MAC address? MAC address is a unique number so that
NIC card can be found on the network. Unless the modem had a dual propose,
I see no reason why it would ever need a MAC address.
>Any here I thought that a MAC address was a special address book that
Billy-bum
>kept for retaliation if he ever took over the world. ?:^)
>
You do realize you only anwsered one of my questions.
------------------------------
From: [EMAIL PROTECTED] (Tim Roberts)
Crossposted-To:
comp.lang.perl.misc,comp.lang.python,comp.lang.tcl,comp.mail.sendmail,comp.os.linux.setup,comp.os.ms-windows.advocacy
Subject: Re: The truth about the Pentium III chip and ID --- **boycott info**
Date: 19 Mar 1999 19:58:16 GMT
John Lehmann <[EMAIL PROTECTED]> wrote:
>Kano wrote:
>>
>> Oh come on. Many other standard devices (ethernet cards, for example)
>> have their
>> own unique serial numbers that software can use at will. Get over it.
>>
>
>Yes - but ethernet mac cards are only broadcast over the lan, not over
>the net (except by Office98, of course).
The POINT here is that it is exactly as easy to embed my unique MAC address
in an Internet transaction as it is to embed my unique Pentium-III CPU ID
in an Internet transaction. Any software which will go to the trouble to
fetch the unique CPU identifier and send it over the network to identify me
could just as easily be written TODAY to use a MAC address for exactly the
same purpose. The CPU identifier is not significantly different in concept
from the MAC address, and yet there has not been a hue and cry to boycott
NIC manufacturers.
>And what do you mean get over it??? This kind of personal information
>is valuable. Business are willing to pay for it. This kind of
>behavior IS DOWNRIGHT THEFT!!!
But whatever they could do with a CPU ID, they are probably already doing
with a MAC address. It is just as good as a unique identifier. "Get over
it" is exactly the right attitude.
--
- Tim Roberts, [EMAIL PROTECTED]
Providenza & Boekelheide, Inc.
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: DNS entry for dynamic IP address
Date: Fri, 19 Mar 1999 19:53:24 GMT
In article <[EMAIL PROTECTED]>,
Raymond Doetjes <[EMAIL PROTECTED]> wrote:
> The rumor goed, that there is a ddns server available. I haven't found it yet.
> They should be a series of scripts working towards bind version8.
> Check the web for ddns!
No, don't.
DNS is not meant to work that way, for very good reasons. If you plan on
pissing off every DNS admin on earth by ruining their cache, go right ahead.
Personally, I prefer to not to have my domain blocked.
DDNS is to DNS as SPAM is to email.
If you're going to set up a domain, do it right. Get a static IP.
-Bill Clark
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: [EMAIL PROTECTED] (W. thomas)
Subject: Urgent! NFS woes...
Date: Fri, 19 Mar 1999 18:45:17 GMT
Hello--
We have a production server (SERVER 1) which mounts its /home
directory to 3 other local servers via NFS, the server is running
RH5.2, kernel 2.2.2, running nfs-server 2.2beta37.1
other systems are 5.2 running 2.2.3 with the following mount options:
rsize=8192,wsize=8192,rw,bg,hard,intr,nolock
>>>>>>>>>>
Problem: SERVER 1 locks up occasionally (that's another problem for
me to worry about), after such a lockup it brings the other systems to
their knees with errors like the following:
>server SERVER 1 not responding, still trying
>nfs: task 20266 can't get a request slot
>(several hundred of these)
>kernel: Unable to load interpreter
When the load interpreter message starts, the system starts spewing
'too many files open' to the console, with load averages in the
hundreds and the systems become effectively useless and need to be
manually shutdown in some cases.
Question: How can I mount the SERVER 1 filesystem so that when SERVER
1 crashes, the other servers automatically 'release' SERVER 1 from its
NFS duties. These systems will function normally without mounting
from SERVER 1, and i can manually remount when SERVER 1 becomes stable
anew.
Any/All help appreciated...
Regards,
W. Thomas
[EMAIL PROTECTED]
WizardNet
------------------------------
** 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
******************************
