Linux-Networking Digest #529, Volume #11 Mon, 14 Jun 99 11:15:32 EDT
Contents:
Re: Playing Audio CDs Over a LAN ("Stephan Beal")
NAT\Linux Questions (Kurt Rupprecht)
Server Problem.... (Evhen Loj)
how to use libpcap (cai yibo)
Re: setting up a second ethernet card (Gilford Wimbley)
Red Hat 5.2 --- postgresql /starting the postmaster ("James Gardner")
Setting put network - How would you do it? (Chris Hoover)
Re: problem with rh6 ("Harrington B. Laufman")
Re: Rules of thumb for posting (Was: afpfs can be found here...) (Rod Smith)
Re: DLink 530 *Tulip* HOW TO? (Rod Smith)
Re: Help! Networking & IP Masquerading & PPP, oh my! ("George Georgakis")
Re: Secure network-backup via nfs? (Frank Sweetser)
Re: Linksys LNE100TX (tulip) keeps going on and off ... conflict ?
([EMAIL PROTECTED])
----------------------------------------------------------------------------
From: "Stephan Beal" <[EMAIL PROTECTED]>
Subject: Re: Playing Audio CDs Over a LAN
Date: Tue, 8 Jun 1999 12:46:42 +0200
No chance on the remote audio. The CD audio is not piped through your CPU,
but it sent directly to the sound hardware on the source machine, meaning
that it can't be redirected. Look for a (normally) grey wire connecting your
CDROM to the sound card or main board (depends on the system).
Now you CAN play sound files from one system when the sound files reside on
another machine. But you can't be on machine A and play a sound to have it
be heard on machine B. Machine B has to pull the sound from machine A and
run it, since machine A doesn't have access to B's hardware (in this case,
the sound card).
Yes, you can use cron to start a CD if you use a program (like workman... or
is it workbone?) that can run CDs from the command line.
Desmond Coughlan wrote in message <[EMAIL PROTECTED]>...
>Ok, my question is (I suspect) quite simple. I have a 3-node LAN, but
>the server is buried away in a corner. How can I put an audio CD into
>the server, and play it (i.e., have the music heard) in another room ?
>
>If I try to export the CD, it needs to be mounted. If I don't mount
>it, then my host can't see it.
>
>Another question: I plan to put four speakers onto my sound card. How
>can I set cron to play a CD at a particular time?
>
>Thanks.
>
>--
>Desmond Coughlan |Restez Zen ... Linux peut le faire
>[EMAIL PROTECTED]
>[www site under construction]
>
------------------------------
From: [EMAIL PROTECTED] (Kurt Rupprecht)
Subject: NAT\Linux Questions
Reply-To: [EMAIL PROTECTED]
Date: Mon, 14 Jun 1999 05:30:57 GMT
I am connecting to the internet from the box in my kitchen (MDK 6.0)
via another box running NT 4.0 using the Sygate NAT software. I am
connected via @home cable service. The NT box is a dual boot that
also has Linux on it, and from there everything works fine. However
when I am logged into Linux on the box downstairs and it goes through
the NAT software it can not find the @home mail server. I solved the
problem by pinging the mail server to get it's IP address and entering
that in the Netscape dialog box for mail server name.
When I ping I see that the mail server name is not just "mail" (which
works when connected directly to the internet via any OS) but actually
"mail.sttln1.wa.home.com" I assume that there is no difference if I
enter this or enter the IP of the server, correct? No advantage to
either method?
Also, I have to enter the @home DNS numbers on the box behind the
Sygate NAT software. The Sygate doc's say to enter the the IP of the
Sygate server for the DNS, however this does not work. In order to go
anywhere I have to enter @home's DNS numbers.
Now, the idea of the Sygate NAT software is to allow more than one box
to connect to the network yet appear to the network as only one box.
Does my having to enter the DNS numbers on the second box (behind the
Sygate server) change this?
I am also rather perplexed by the whole NAT idea. I understand that
it changes the headers of the packets to reflect the MAC of the NAT
server (at least I think I understand that is the case) but how in the
world does the NAT server know when it gets a packet that it is
intended not ofr it, but for a different box? I assume that it it
adds information to the header, but then why is that information not
seen elsewhere? At what level is the info added? The network level?
The data link level? I would greatly appreciate any info on this,
maybe some pointers to some good info on the web.
Thanks Kurt
[EMAIL PROTECTED]
------------------------------
From: Evhen Loj <[EMAIL PROTECTED]>
Subject: Server Problem....
Date: 13 Jun 1999 16:30:44 GMT
I need help on a very annoying problem... I have a Celeron 400 w/128MB
of RAM running Linux 6.0 ... I want to setup the system to run as a web
server with Apache, which has been installed. Everything works fine until
I leave the machine running for 20+ minutes. Then my HTTP, Telnet, and FTP
services shut down. I can't even ping the machine from home, all I get is
a request timed out. I've even reinstalled the software with hopes that
that might solve the problem, but no luck. Any advice would be greatly
appreciated. Thx.
Also need some advice on setting up TCP wrapper, so I can admin. server
from a static IP.
================== Posted via SearchLinux ==================
http://www.searchlinux.com
------------------------------
From: cai yibo <[EMAIL PROTECTED]>
Subject: how to use libpcap
Date: Mon, 14 Jun 1999 19:53:09 +0800
where can I find libpcap programming documents?
------------------------------
From: [EMAIL PROTECTED] (Gilford Wimbley)
Subject: Re: setting up a second ethernet card
Date: Mon, 14 Jun 1999 05:06:08 GMT
On Sun, 13 Jun 1999 02:20:44 GMT, "James A. Robertson"
<[EMAIL PROTECTED]> wrote:
>I have an intel etherpro 100 (works fine) set up to a cable modem with a
>static ip. I have a second ethernet card configured to local LAN. I
>can't ping the machine (from other boxes) through the second IP.
>
>Route table:
>
>[root@cc36623-a log]# route
>Kernel IP routing table
>Destination Gateway Genmask Flags Metric Ref Use
>Iface
>24.3.20.0 * 255.255.255.0 U 0 0 354
>eth0
>192.168.1.0 * 255.255.255.0 U 0 0 2
>eth1
>127.0.0.0 * 255.0.0.0 U 0 0 13
>lo
>default cr2-hfc1.hwrd1. 0.0.0.0 UG 0 0 599
>eth0
>
>results of ifconfig:
>
>[root@cc36623-a log]# ifconfig
>lo Link encap:Local Loopback
> inet addr:127.0.0.1 Bcast:127.255.255.255 Mask:255.0.0.0
> UP BROADCAST LOOPBACK RUNNING MTU:3584 Metric:1
> RX packets:3938 errors:0 dropped:0 overruns:0 frame:0
> TX packets:3938 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0
>
>eth0 Link encap:Ethernet HWaddr 00:A0:C9:EF:48:75
> inet addr:24.3.20.102 Bcast:24.3.20.255 Mask:255.255.255.0
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:77813 errors:1 dropped:0 overruns:0 frame:0
> TX packets:22739 errors:0 dropped:0 overruns:0 carrier:0
> collisions:391
> Interrupt:9 Base address:0xe400
>
>eth1 Link encap:Ethernet HWaddr 00:10:5A:E4:FA:C9
> inet addr:192.168.1.1 Bcast:192.168.1.255 Mask:255.255.255.0
>
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:0 errors:0 dropped:0 overruns:0 frame:0
> TX packets:17 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0
> Interrupt:10 Base address:0x300
>
>I'm not even at the point where I can set up IP Masq; I can't get a
>local ping through. Anyone have an idea what I'm doing wrong ? eth0 is
>a PCI card, eth1 is an ISA 3c509
>
>Thanks
>
>
>
The only thing I can tell you is that I had a lot of problems with an
ISA 3c509b and I finally "fixed" it by putting in a different card.
Apparently there are things you can do to make the 509 work (something
about enabling support for legacy ISA devices).
The problems I had were more or less just like yours. The card seemed
to work, but I couldn't ping anything and nothing could ping me.
After I replaced the card with a newer PCI card (a 3c900b), everything
worked fine. My conclusion was that linux support for the 509b was
imperfect. I don't know if that conclusion was fair, but whatever.
good luck. When you resolve this situation, please post a note about
it for my edification. I'd like to know how to make the 3c509b work
if it really is possible.
regards,
GW
------------------------------
From: "James Gardner" <[EMAIL PROTECTED]>
Subject: Red Hat 5.2 --- postgresql /starting the postmaster
Date: Sun, 13 Jun 1999 21:06:57 -0700
This is a multi-part message in MIME format.
=======_NextPart_000_002F_01BEB5E0.AC5361C0
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
New to databasing on the Unix-Linux platform. Does anyone have any ideal =
on the following?
1.) How to setup the PGDATA environment variable?
2.) How to get the postmaster running?
3.) How to get started in general?
Thanks,
James
=======_NextPart_000_002F_01BEB5E0.AC5361C0
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN">
<HTML>
<HEAD>
<META content=3Dtext/html;charset=3Diso-8859-1 =
http-equiv=3DContent-Type>
<META content=3D'"MSHTML 4.72.3612.1700"' name=3DGENERATOR>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT size=3D2>New to databasing on the Unix-Linux platform. Does =
anyone have=20
any ideal on the following?</FONT></DIV>
<DIV><FONT size=3D2></FONT> </DIV>
<DIV><FONT size=3D2>1.) How </FONT><FONT size=3D2>to setup the PGDATA =
environment=20
variable?</FONT></DIV>
<DIV><FONT size=3D2>2.) How to get the postmaster running?</FONT></DIV>
<DIV><FONT size=3D2>3.) How to get started in general?</FONT></DIV>
<DIV><FONT size=3D2></FONT> </DIV>
<DIV><FONT size=3D2>Thanks,</FONT></DIV>
<DIV><FONT size=3D2>James</FONT></DIV></BODY></HTML>
=======_NextPart_000_002F_01BEB5E0.AC5361C0==
------------------------------
From: Chris Hoover <[EMAIL PROTECTED]>
Subject: Setting put network - How would you do it?
Date: Mon, 14 Jun 1999 00:14:42 -0400
I'm wanting to redo my personal network, and am looking for some
ideas/advice on how to best set it up.
Currently, I have 4 boxes on the network running 10baseT. However, they
are all on 10/100 cards, and I'll be picking up a new 10/100 hub in the
near future.
As for my current setup:
Machine 1 - a 486 with about a 400 meg hd and 24 megs of memory that I
use to dial up my isp and run my printer + samba. It is running Debian
2.1.
Machine 2 - a Pentium 100 with 64megs and a 4gig drive that is running
Suse 6.0 (this box is not really doing anything right now).
Machine 3 - a Pentium 200 with 128 megs and 2.5 gig drive, 4X4 ide
cdchanger running NT/win9X.
Machine 4. - a AMD K6-2 300 with 256 megs and ~21gigs of hd + scsi tape
backup (2.5 gigs) and scsi 4x6 cdchanger running RedHat 6.0. (2 - 8gigs
and 1 5gig drive)
All of the drives can be moved around and reused except the 5 gig which
is full of my mp3's.
Anyway, I want to keep one machine running Microsoft for games and so
forth. I'd also really like to setup a good linux server with one
(probably the K6-2).
So, how would you do it, and what distributions of linux would you run
(if you would pick a particular version over the others for a specific
reason - personal preference aside).
Thanks for the help,
Chris
------------------------------
From: "Harrington B. Laufman" <[EMAIL PROTECTED]>
Subject: Re: problem with rh6
Date: Sun, 13 Jun 1999 11:03:01 -0400
Colombier Pierre wrote:
> upgrading 5.2 to 6.0 or installing 6.0 put down my network, my computer
> became blind.
> My 3com 503c work very well with 5.2 but when I upgrade to 6.0 there's
> nothing to ping, all is unreacheable ???
> a solution ?
>
> thanks for helping me
>
> Pierre ;-)
Hello,
Not much information to go on, but I did the same upgrade and lost my ip
forwarding,
thereby breaking my little network. I was using some ipfwadm commands in
rc.local to help secure the network.
The ipfwadm command is gone. So /etc/rc.d/rc.local scripts fail-->no ip
forwarding. ipfwadm has been replaced by ipchains. A Howto tells us:
" Linux ipchains is a rewrite of the Linux IPv4 firewalling code (which
was
mainly stolen from BSD) and a rewrite of ipfwadm, which was a rewrite of
BSD's
ipfw, I believe. It is required to administer the IP packet filters in
Linux
kernel versions 2.1.102 and above.
"The older Linux firewalling code doesn't deal with fragments, has 32-bit
counters (on Intel at least), doesn't allow specification of protocols
other
than TCP, UDP or ICMP, can't make large changes atomically, can't specify
inverse rules, has some quirks, and can be tough to manage (making it
prone to
user error). "
But the life saver is ipfwadm-wrapper, which takes ipfwadm command
arguments and pipes them to ipchains in a correct format. Change all
script
references from /sbin/ipfwadm to /sbin/ipfwadm-wrapper and you are fixed.
Wrestle with the new ipchains command syntax later, get up and running
now.
ipfwadm-wrapper comes with Red Hat 6.0.
Regards,
Harrington
------------------------------
From: [EMAIL PROTECTED] (Rod Smith)
Crossposted-To: linux.appletalk
Subject: Re: Rules of thumb for posting (Was: afpfs can be found here...)
Date: Mon, 14 Jun 1999 13:50:04 GMT
Reply-To: [EMAIL PROTECTED]
In article <[EMAIL PROTECTED]>,
Bjørn Ruberg <[EMAIL PROTECTED]> writes:
>
> 2. Attaching a binary file to the posting; Personally I have no problem
> with that as I am now connected to a LAN, but when trying to read news
> connected through a modem, large attachments virtually killed the
> transfers. And there are still a lot of modem/ISDN users in the world.
> The rule of thumb is that binaries should never be posted, unless in
> dedicated newsgroups (usually something with .binaries in the NG name)
> so that people know what they might end up downloading. Instead, include
> a URL to where the file is to be found, so that those who want this file
> might download it themselves.
Some additional points that are closely related to this one:
1) Many people (including, I understand, most Europeans who use modems to
access the Internet) pay for net access per unit time. Binaries will
increase the total net access time, thus costing those people, in total,
substantial amounts of money, especially if they've got their systems
configured to auto-download the entire contents of a newsgroup each
night or some such.
2) Some smaller ISPs, or people who run news servers locally, may have
problems if large numbers of binaries (or even just one or two
particularly large binaries) suddenly appear in a text-only newsgroup.
Thus, posting binaries to a newsgroup could potentially cause some ISPs
to drop that newsgroup from their servers.
These two reasons are both pretty serious, and for this reason, tolerance
to violation of this rule does tend to be pretty low. I haven't checked
them out even remotely recently, but I'm sure the no-binaries rule is
discussed in regular postings to the news.newusers.questions and/or
news.announce.newusers newsgroups. Most newsreaders will automatically
subscribe a person to these or other similar newsgroups the first time
they're launched. I think most newbies to Usenet News ignore these
groups, but they shouldn't.
--
Rod Smith
[EMAIL PROTECTED]
http://www.channel1.com/users/rodsmith
NOTE: Remove the "uce" word from my address to mail me
------------------------------
From: [EMAIL PROTECTED] (Rod Smith)
Subject: Re: DLink 530 *Tulip* HOW TO?
Date: Mon, 14 Jun 1999 13:38:55 GMT
Reply-To: [EMAIL PROTECTED]
[Posted and mailed]
In article <[EMAIL PROTECTED]>,
Rich ''Doc'' Colley <[EMAIL PROTECTED]> writes:
>
> I have installed (successfully) my FIRST LINUX server,
> BUT cannot figure out how to use the DLink Driver I
> downloaded for the 530 NIC. It came from the 'official'
> LINUX nic driver site, linked to DLINK. But how
> do I use this 'module' (if that is what it is... called
> "via-rhine.c" ) I know that LINUX doesn't automatically
> find the NIC as it is not supported in 5.2 (?? Hmm wonder
> about 6.X??)
In brief, you need to recompile your kernel to include the new driver.
The easiest way to do this may actually be to download a recent 2.2.x
kernel, since the via-rhine.c file is standard with them. Then follow the
instructions for compiling, installing, and using a new kernel, reboot,
and it should work. I'm afraid a step-by-step procedure would be rather
involved for a single e-mail message, though; if you need more, either try
to find a web site on the subject or buy a good Linux book that includes a
chapter on compiling the kernel.
Red Hat 6.0 comes with kernel 2.2.5, but I don't know offhand if they've
actually compiled the via-rhine drivers. If they have, RH 6.0 might
support your NIC out of the box. If not, you'd have to recompile the
kernel and/or kernel modules to get it to work.
--
Rod Smith
[EMAIL PROTECTED]
http://www.channel1.com/users/rodsmith
NOTE: Remove the "uce" word from my address to mail me
------------------------------
From: "George Georgakis" <[EMAIL PROTECTED]>
Subject: Re: Help! Networking & IP Masquerading & PPP, oh my!
Date: Mon, 14 Jun 1999 06:12:21 GMT
My comments are below.
George
===========================================================================
I never reply by email as a) I don't give out my real email address freely,
and b) it stops other NG users from reading the solutions to problems
If necessary, however, I can be contacted thru geegs (a) linuxstart DOT com
==========================================================================
Gilford Wimbley <[EMAIL PROTECTED]> wrote in article
<[EMAIL PROTECTED]>...
> On Sat, 12 Jun 1999 23:31:22 -0700, "David Gallo"
> <[EMAIL PROTECTED]> wrote:
>
> >Hmmm..... I guess I was being a little vague on what exactly I wanted
help
> >with.
> >
> >First things first.
> >
> >1) The kernel already has the appropriate bits installed, so there's one
egg
> >in the basket. Lucas Bradstreet: Will the info about IPchains really
help me
> >if I'm running RH5.2? I was under the impression that 5.2 only had
ipfwadm?
> >Please correct me if I'm wrong.
> >
> I'm glad you weren't asking me, because I don't know ;-)
Yes, RH 5.2 uses ipfwadm not ipchains
>
> >2) I have issued forwarding policies (I also have a few books that seem
to
> >be quite excellent as well as the How-To. It's still a bit confusing,
but I
> >think I understand this part). Egg isn't quite in the basket, but this
isn't
> >the point I'm at yet (I think).
> >
> >3) Yes. This is easy. However, the part I want to know about is this:
> >
> >If I do not have a separate firewall server, then how do I use ipfwadm
(how
> >do I set the interfaces/source & destination) and how do I configure the
> >gateway? All documentation I've been able to find has been for a
dual-server
> >setup, and that's not what I want to do right now ( I just want to get
it
> >working this way so I can use it. Down the road I'll end up setting up a
> >second server for security's sake.)
> I don't have a second server either. My home lan isn't an ultra
> secure zone. Oh, well...
>
> The commands I put in /etc/rc.d/rc.local are as follows:
>
> ipfwadm -F -p deny
> ipfwadm -F -a m -S 192.168.1.0/24 -D 0.0.0.0/0
You will also need to put in another line to enable masquerading. On my
Slackware system the command is
echo "1" > /proc/sys/net/ipv4/ip_forward
BUT!! I believe RH may have a different location. Check it out first.
<rest of this stuff snipped as I have nothing more to add>
------------------------------
From: Frank Sweetser <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: Secure network-backup via nfs?
Date: 14 Jun 1999 09:31:54 -0400
[EMAIL PROTECTED] writes:
> I work for a company who runs a number of publicly accessible servers.
> That's five servers running some version of RedHat ( kernel 2.2.x ) and
> various services like sendmail, apache, mysql.
> For backups we have a HP SureStore 24eU DAT tapedrive, which is
> connected to one of the servers. BRU is the preferred software.
>
> My question is: How can I backup all the servers in a secure way, by
> using the host to which the DAT is connected?
'secure' and 'nfs' don't belong in the same sentence.
> Ideally I would like to nfs-mount all servers on the DAT-host, but I
> have avoided nfs for perceived lack of security and performance issues.
> How vulnerable does it make our servers if we use nfs ( assuming we
> configure it properly )?
>
> Another possibility would be using the rmt device. I would still need to
> enable rpc for that, again opening up the system.
>
> What's the best way of handling backup for a similar setup?
my preference is ssh. you can set it up to log in to a particular machine
w/out any password. you can then have the clients do something like this
tar cvzf - / | ssh backupaccount@backupserver dd of=/dev/tape_device
--
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
>Ever heard of .cshrc?
That's a city in Bosnia. Right?
(Discussion in comp.os.linux.misc on the intuitiveness of commands.)
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Linksys LNE100TX (tulip) keeps going on and off ... conflict ?
Date: Mon, 14 Jun 1999 06:13:09 GMT
Okay, so I answered my own question. Just FYI for those interested:
Created tulip driver as a module. (downloaded the latest and greatest)
Compiled Kernel and did make modules_install
Added to conf.modules
options tulip io=0x300 irq=11
alias eth0 tulip
io address was the problem, it was autoconfiguring for 0x100,
apparently a conflict with something.
Then added the following to my /etc/rc.d/rc.sysinit
Under the #load sound modules section
#Load Tulip Driver
if [ -n "$USEMODULES" ]; then
if grep -s -q "^alias eth0 /etc/conf.modules ; then
action "Load TulipMod" insmod /lib/modules/2.2.5-15/net/tulip.o
options=0,16
ifdown eth0
ifup eth0
fi
fi
Saved and reboot the system. WOrking fine now.
Note: the options=0,16 supposedly forces it to use full-duplex
100Mbit.
------------------------------
** 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
******************************
