Linux-Networking Digest #622, Volume #12 Fri, 17 Sep 99 18:13:39 EDT
Contents:
Remote server monitoring akin to AlertPage ("Stephen Perkins")
Re: Getting Linux and NetWare 5 to talk to each other (Gustin Kiffney)
Re: FTP problems (Cameron L. Spitzer)
DHCP problems after power failure - connection refused ("Dave Macolino")
Re: Recommendation for 100Mbps Switched Ethernet hardware (David C.)
Re: Anyone using Linksys Fast Ethernet 10/100 Network in a Box ?
([EMAIL PROTECTED])
ipmasqadm question (razoon)
Re: Recommendation for 100Mbps Switched Ethernet hardware ("Jim Matuska")
HELP!!! -- named down on RH5.2 ([EMAIL PROTECTED])
named killed telnet?!? Help! ([EMAIL PROTECTED])
Mismatched Read Page Pointers??? ("DrDAP!!!")
yppasswdd quits after one use ([EMAIL PROTECTED])
yppasswdd quits after one use ([EMAIL PROTECTED])
Re: Recommendation for 100Mbps Switched Ethernet hardware (Bryan)
named.conf error on bootup ([EMAIL PROTECTED])
Re: TCP/IP Programming Mystery (David C.)
Re: Router running NAT (David C.)
Re: linux box vs switched hub (David C.)
WfW/DOS w/TCP/IP can't use Samba share (Steve)
----------------------------------------------------------------------------
From: "Stephen Perkins" <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.misc
Subject: Remote server monitoring akin to AlertPage
Date: Fri, 17 Sep 1999 14:50:43 -0500
Hi all,
I'm looking for some a software package that will constantly monitor the
health of remote servers and services. Something along the lines of
AlertPage from Geneva Software or SiteScope. However, I'd like the utility
to run on Linux and, preferrably, not cost an arm and a leg.
Specifically, I need to automatically ping a set of servers every minute,
and if the ping fails x times in a row, send an email message to a
predetermed address. When/if the server comes back, I'd like a different
message sent.
It would be nice to be able to monitor services as well (Sendmail, DNS, WWW,
etc) but that is not required.
I'm sure there is something like this available, but searchs on dejanews etc
came up empty. I'm probably using the wrong keywords. Any thoughts
regarding this would be appreciated!
Thanks in advance,
- Steve
P.S. Running RedHat 6.0 and would love an RPM pointer...
------------------------------
From: Gustin Kiffney <[EMAIL PROTECTED]>
Subject: Re: Getting Linux and NetWare 5 to talk to each other
Date: Fri, 17 Sep 1999 19:46:14 GMT
[posted and mailed]
No, far as I know there is no IP-based Linux client for Netware 5, and
if there were one Caldera would have it. ncpfs uses IPX so you would
have to get that working to access the NW 5 server, assuming you set
IT up to use ipx (NW 5 I believe can run strictly IP, but with no
client that does you no good, of course).
If you can't get ipx to run from W95, which has pretty good ipx
support, then something is pretty wrong (perhaps you DIDN'T enable IPX
when you set up the Novell server???) Once you do connect with
w95 ipx, you should have pretty good luck connecting with ipx
and ncpmount.
[EMAIL PROTECTED] (Scott Alfter) wrote:
>SuSE 6.1
> includes ncpfs, but that only wants to work over IPX, and I have had
zero
> luck getting IPX to work (can't even use RCONSOLE from a Win9x box).
> Besides, I'd just as soon keep my home network 100% IP. I also
tracked down
> a version of Samba that had been put together for NetWare 4.1x, but
didn't
> have much luck getting it running either. Does anyone know of a
NetWare
> client for Linux that works over IP-based networks?
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (Cameron L. Spitzer)
Subject: Re: FTP problems
Date: 17 Sep 1999 19:47:30 GMT
In article <[EMAIL PROTECTED]>, Quiney, Philip
(EXCHANGE:HAL02:HM10) wrote:
>You should get something like
>
>anonftp-2.8-1
>ftp-0.10-22
>gftp-1.13-4
>ncftp-3.0beta18-3
>tftp-0.10-23
>wu-ftpd-2.4.2vr17-3
>
>wu-ftpd is the item of interest as it is the server end.
>
>An 'rpm -ql wu-ftpd-2.4.2vr17-3'
>
>will reveal the executable /usr/sbin/in.ftpd which is the reference from
>/etc/inetd.conf
>
>If it is not present then you will need to install it from the CD.
WARNING! The wu-ftpd-2.4.2 that is widely distributed has
buffer overflow bugs that allow intruders to install files
on your system. There are unofficial patches floating around,
but wu-ftpd-2.4.2 is no longer supported or maintained.
If you like wu-ftpd you should get the current version from
ftp://ftp.wu-ftpd.org/
and read
http://www.landfield.com/wu-ftpd/
If you have linux-2.0 or -2.2 and gcc-2.7.2 or later the source
will build for you with no problems.
If your computer is exposed to the Internet, the FTP daemon will
attract intrusion attempts. Don't trust anybody to install and
configure an FTPd for you. Read the docs, build the daemon,
and write the config files yourself.
Debian and Red Hat use inetd and tcpd to invoke wu-ftpd.
Inetd is, in my experience, unreliable in the face of intrusion attempts
and/or denial of service attacks. If your server is important,
consider commenting ftp out of /etc/inetd.conf and
installing tcpserver to invoke wu-ftpd instead.
Obtain from ftp://koobera.math.uic.edu/www/ucspi-tcp.html
Tcpserver combines the functions of inetd and tcpd.
You run a separate instance of the daemon for each service.
Cameron
------------------------------
From: "Dave Macolino" <[EMAIL PROTECTED]>
Subject: DHCP problems after power failure - connection refused
Date: Fri, 17 Sep 1999 15:50:40 -0400
During the remnants of hurricane floyd yesterday, my house had a power spike
and everything
rebooted. After that my cable modem refuses to work with Linux/dhcpcd 0.7.
It has been working for
over a year now.
I am in Long Island NY and have Cablevision as a provider with a Lancity
LCPET-3 Cable Modem.
When I run:
/sbin/dhcpcd -d eth1
I get the following message:
recvfrom (rcvAndCheckDhcpMsg): Connection Refused. (repeated 3x)
then I get the standard "no DHCPOFFER messages". I've tried it with two
different ethernet cards, that work on my
internal lan. However, it does work with Windows 98 on a desktop and laptop
with no problem.
Has anyone seen this before ?
------------------------------
From: [EMAIL PROTECTED] (David C.)
Crossposted-To: comp.sys.ibm.pc.hardware.networking
Subject: Re: Recommendation for 100Mbps Switched Ethernet hardware
Date: 17 Sep 1999 16:03:03 -0400
Bryan <Bryan@[EMAIL PROTECTED]> writes:
> In comp.os.linux.networking David C. <[EMAIL PROTECTED]> wrote:
>
>> - Hub - An incredibly generic term. Any device that connects
>> multiple network devices together. Usually, when people refer to
>> "hubs" they are referring to dumb repeaters.
>
> depends on the context; for an office 'port expander' the size of
> which you can hold in your palm, it won't be managable (dumb). but
> alost any comms room hub will have snmp on it for management. maybe
> even an embedded webserver and telnet.
Yes. That's why I later mentioned that managed repeaters also exist. I
was, however, targetting the message at non-corporate people. When you
go into a computer store like CompUSA and look at the hubs, you find
mostly dumb repeaters and unmanaged layer-2 switches.
>> A repeater takes every packet that comes in and transmits it out
>> every port. Repeaters have no intelligence.
>
> they do know about when a port is 'bad' and will cut off the bad
> station from the rest of the network. too many collisions (sometimes)
> and definitely when a tolerance is horribly wrong. certainly when no
> link is at the far end, the port will partition ('shut off') that
> port.
Yes. I did oversimplify this a bit. I'll include this in my file.
>> Layer-2 switches are often referred to as "bridges". They forward
>> packets between two network segments based on layer-2 information,
>> which is the 6-byte source and destination MAC addresses on Ethernet
>> networks. They are called bridges because the "bridge" between two
>> layer-2 segments, combining them into a single layer-3 network.
>
> also part of bridging is running a distributed algorithm called
> 'spanning tree'. which essentially exists to remove loops (cycles)
> from extended lans. if a lan has a loop, frames will travel thru and
> thru and never really expire. IP doesn't need this since the protocol
> has a 'time to live' field and if some lost packet circles thru too
> many times, it will auto-expire.
Again true. I didn't think this was very important for a typical user
to know. It's more important for anyone trying to design or build a
bridge (perhaps with software on a PC) to know.
>> layer-3 switches are often referred to as "routers". They forward
>> packets between two layer-2 network segments based on information in
>> each packet's layer-3 header. They will rewrite a packet's layer-2
>> information in doing so, because layer-2 information from one segment
>> is usually uselsss on another segment.
>
> slight problem with this. 'useless on the other segment'? ethernet
> addr's are (should be) globally unique so there's no problem in merely
> repeating the frame as-is on the other segment. but the destination
> address of the next-hop router has to be modified if that router is
> supposed to relay that packet.
I should have said "network" instead of "segment". Layer-2 information
is definitely valid when transmitted across bridges.
The addresses are not necessarily unique, however. The Ethernet spec
allows for host-assigned addresses, which may be used instead of the
factory-assigned addresses. If they are used, it is possible for two
hosts on two different networks to have the same layer-2 address. (It
wouldn't be a smart thing to do, but that's another subject.)
Either way, if you forward a packet from one network to another and
don't rewrite the layer-2 information, it won't go anywhere. That's
what I meant by "useless". I'll clarify this in my next revision.
>> Layer-3 switches can only work with the layer-3 protocols they're
>> designed for. This is usually IP, but many networks will use other
>> protocols as well (like NetBEUI, IPX and AppleTalk). If the router
>> doesn't have software to deal with those (and other) protocols, it
>> not be able to forward those kinds of packets to other networks.
>
> usually these days, routers can fallback to bridging mode and bridge
> what they can't natively route.
Maybe. It depends on how they're configured. If each port is connected
to a different network/subnet, this won't happen. If multiple ports are
connected to the same network/subnet, then the router may bridge packets
between those ports.
But that's not really a feature of a layer-3 switch. It is layer-2
functionality. As I said early on, there exist switches which provide
both layer-2 and layer-3 functionality. A router that also bridges some
packets is such a switch.
-- David
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Anyone using Linksys Fast Ethernet 10/100 Network in a Box ?
Date: Fri, 17 Sep 1999 20:00:33 GMT
I got mine working by using the tulip driver that came on the linksys
cards. Just follow the compile instructions as noted at the bottom of
the driver.
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: razoon <[EMAIL PROTECTED]>
Subject: ipmasqadm question
Date: Fri, 17 Sep 1999 21:48:26 +0200
Reply-To: [EMAIL PROTECTED]
I use kernel 229
When i do:
Ipmasqadm portfw -a -P udp -L <DHCP_IPADDRESS> 2213 -R <client> 2213
Ipmasqadm portfw -a -P udp -L <DHCP_IPADDRESS> 6666 -R <client> 6666
i get the following:
portfw: setsockopt failed: Invalid argument
What does this mean?
thnx
------------------------------
From: "Jim Matuska" <[EMAIL PROTECTED]>
Crossposted-To: comp.sys.ibm.pc.hardware.networking
Subject: Re: Recommendation for 100Mbps Switched Ethernet hardware
Date: Fri, 17 Sep 1999 13:23:25 -0700
I would recommend Cisco hardware, they make some really cool switches
ranging from about $700 to thousands of dollars depending on what speed,
port density and features you need. check them out online at www.cisco.com
you can get a local cisco representative to get in touch with you for more
advice. Cisco equipment is more expensive then some other brands, but you
get what you pay for. I beleive last I check their equipment makes up about
85% of the equipment that is connected to the internet.
sys8841 <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Specifically, I need recommendations on a switches that I can use to
> setup a 100Mbps Switched Ethernet hardware.
>
> Thanks.
>
------------------------------
From: [EMAIL PROTECTED]
Subject: HELP!!! -- named down on RH5.2
Date: Fri, 17 Sep 1999 20:09:06 GMT
Somehow, overnight, my name server crashed which took down the entire
server altogether. I was able to reboot and "capture" stuff from the
lost+found directory, but now named isn't working.
It appears that there is a checksum error
in /etc/named.conf, /etc/named.boot and /var/named/named.ca .
Is there a way to 'repair' these files????
I can read and edit these files fine in pico. Have even rewritten them.
When doing a ndc restart, all seems to go well, but when reviewing the
messages log, this is what I get...
Sep 17 {time} {host}named[6820]: sysquery: findns error (SERVFAIL) on ?
({time}and {host} are my entries here)
httpd is running fine, all websites are working, email is working, just
workstations on the network aren't able to get out.
Is there a way to get the files noticed again?
What do I need to do here?
Also did a dig @rs.internic.net to rebuild the cache file and got this
error code.
;;->HEADER<<- opcode:QUERY, status:SERVFAIL, id:6
At a total loss and am in need of expert help.
If at all possible, please reply to my email address??
NOSPAM<<{michaelj}>>@<<{mgarts.net}>>
TIA
Michael
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED]
Subject: named killed telnet?!? Help!
Date: Fri, 17 Sep 1999 20:26:24 GMT
System: Intel Linux (Slackware) 2.0.30 running as a gateway for a tiny
private LAN
I've been able to telnet into my machine from around the net for
months and just recently decided to try and get named running for the
private LAN. I was having trouble with it and found an option in the
/usr/sbin/ndc script called 'stats'. I executed ndc with the stats
option while in a telnet session and the message "Dumping stats"
appeared briefly, immediately before I got disconnected with a
"connection terminated by remote system" message. I tried to telnet
back into the box but it won't accept my login any more (root or any
other).
I've taken named down and rebooted the system (named is not
autostarted). I've restored my /etc and /usr/sbin directories. I've
rebooted several times. I am still unable to telnet into that box (I
can't rlogin either) but I *can* login at the console just fine.
Incidentally, I can still telnet and rlogin FROM this machine to any
other. I can whois and nslookup just as before. It looks like only
my remote authentication has gotten screwed up (inetd does not spawn
in.telnetd after I enter my user name and password, as though identd
returned an error code or something... I don't use tcpd)
Anyone have any ideas?? I'm in deep crap here...
Thanks in advance,
Jon
jonb @ omsmotion . com
------------------------------
From: "DrDAP!!!" <[EMAIL PROTECTED]>
Subject: Mismatched Read Page Pointers???
Date: Fri, 17 Sep 1999 16:43:17 -0400
Hi,
I'm trying to set up a RedHat 5.2 Linux machine to connect with a Win98
machine via two ethernet cards. I've got everything setup as best as I
could understand from all the docs and how to's.
I can ping the Linux machine from itself and I can ping the Win98 machine
from itself. I cannot ping either from either. When I try to ping the
Linux machine from Win98 I get messages in the /var/log/messages and
sometimes on any open terminal window that says "Linux Kernel: eth0:
mismatched read page pointers 41 vs 49."
What the heck does that mean??? And How do I fix it??
Any help is appreciated - please keep it simple - I'm still quite new to
Linux.
DrDAP!!!
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED]
Subject: yppasswdd quits after one use
Date: Fri, 17 Sep 1999 20:42:52 GMT
I have an odd error. The yppasswdd stops working after one use. I can
restart the daemon and it will work fine, for one use. It is a never
ending loop. Does anyone have a solution or can someone point me to a
section of code where I can add a call to restart the yppaswdd?
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED]
Subject: yppasswdd quits after one use
Date: Fri, 17 Sep 1999 20:44:20 GMT
I have an odd error. The yppasswdd stops working after one use. The
error message says: RPC unable to receive.
I can restart the daemon and it will work fine, for one use. It is a
never ending loop. Does anyone have a solution or can someone point me
to a section of code where I can add a call to restart the yppaswdd
after every passwd update?
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: Bryan <Bryan@[EMAIL PROTECTED]>
Subject: Re: Recommendation for 100Mbps Switched Ethernet hardware
Crossposted-To: comp.sys.ibm.pc.hardware.networking
Date: Fri, 17 Sep 1999 20:54:20 GMT
In comp.os.linux.networking Jim Matuska <[EMAIL PROTECTED]> wrote:
: I would recommend Cisco hardware, they make some really cool switches
: ranging from about $700 to thousands of dollars depending on what speed,
: port density and features you need. check them out online at www.cisco.com
: you can get a local cisco representative to get in touch with you for more
: advice. Cisco equipment is more expensive then some other brands, but you
: get what you pay for.
well, you get some of what you pay for; the rest is based on the
famous name. arguably for low-end things like lan switches, you don't
need a sledgehammer to swat a fly. lan switching in layer2 is
becoming so non-specialized that you don't need the 'smartest' company
to get such mundane tasks implemented.
--
Bryan, http://www.Grateful.Net - Linux/Web-based Network Management
->->-> to email me, you must hunt the WUMPUS and kill it.
------------------------------
From: [EMAIL PROTECTED]
Subject: named.conf error on bootup
Date: Fri, 17 Sep 1999 20:58:12 GMT
When I reboot my Linux box (running RH 5.2)
everything loads fine until about
3/4th of the way.
I then get an error dialog box saying
"Error: Missing zone type (master or slave)
Line 10"
That window will go away after about 20
seconds.
Then I have to go through the entire
named.conf file and hit "OK" for every error
that it finds, such as:
"Invalid keyword {
file /etc/named.conf line 18"
This goes on for about 70 times as
it finds an several errosr for every
entry
Obviously there is something wrong
with my named.conf file but I can't see
any errors. After I do all of this the
machine finishes booting and everything
works fine!
All the domains I am hosting work
without a problem.
I've included my named.conf file
for anyone who might be able to point
out where I have gone wrong.
TIA.
Jim
==============
// generated by named-bootconf.pl
options {
directory "/var/named";
};
//
// a caching only nameserver config
//
zone "." in {
type hint;
file "named.ca";
};
zone "0.0.127.in-addr.arpa" in {
type master;
file "named.local";
};
zone "tribecentral.com" in {
type master;
file "db.tribecentral";
};
zone "brownscentral.com" in {
type master;
file "db.brownscentral";
};
zone "akronhostlions.com" in {
type master;
file "db.akronhostlions";
};
zone "akronrubbergroup.org" in {
type master;
file "db.akronrubbergroup";
};
zone "bjnie.com" in {
type master;
file "db.bjnie";
};
zone "hogan.ohio.com" in {
type master;
file "db.hogan";
};
zone "neohiocars.com" in {
type master;
file "db.neohiocars";
};
zone "akroncars.com" in {
type master;
file "db.akroncars";
};
zone "realparenting.com" in {
type master;
file "db.realparenting";
};
zone "cavscentral.com" in {
type master;
file "db.cavscentral";
};
zone "oaea.org" in {
type master;
file "db.oaea";
};
zone "211ohio.net" in {
type master;
file "db.211ohio";
};
zone "akronjaycees.org" in {
type master;
file "db.akronjaycees_org";
};
zone "akronjaycees.com" in {
type master;
file "db.akronjaycees_com";
};
zone "akronsymphony.org" in {
type master;
file "db.akronsymphony";
};
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (David C.)
Subject: Re: TCP/IP Programming Mystery
Date: 17 Sep 1999 16:52:51 -0400
John Gardner <[EMAIL PROTECTED]> writes:
>>
>> It has to do with the TCP stack on the sender side.
>>
>> A properly written TCP stack will send each TCP packet at a size that
>> will not get fragmented. This is 1460 bytes of data:
>>
>> ...
>
> Verbosely put!
I find its easier for the novice to understand if the mechanism is
spelled out in great detail. (And since I can type fairly quickly, it
doesn't take much time to write.)
-- David
------------------------------
From: [EMAIL PROTECTED] (David C.)
Subject: Re: Router running NAT
Date: 17 Sep 1999 17:02:40 -0400
[EMAIL PROTECTED] (Lew Pitcher) writes:
>
> First, you have to enable IP forwarding and IP Masquarading in your
> kernel. This is usually done by altering the kernel source
> configuration files and recompiling your kernel. I'd suggest that you
> read the "Linux Kernel Howto" before attempting this. You should also
> read the "Linux IP Masquerade mini-Howto" to get the proper
> parameters, and a better understanding of Linux NAT.
Many distributions already have this compiled in. You may be able to
simply set some run-time parameters.
For instance, on RedHat 5.2 and 6.0, you can set the "FORWARD_IPV4" flag
in /etc/sysconfig/network, and the scripts that come with the
distribution will do the rest after a reboot.
> Second, acquire and install the ipfwadm (for Linux 2.0 kernels) or the
> ipchains (for Linux 2.1 and better kernels) tools from xos.nl . These
> tools allow you to configure your IP masquerading rules (which are
> part of the IP Firewall implementation, another howto, I'm afraid -
> the "Firewalling and Proxy Server Howto").
Again, these may be compiled and ready to go with your distribution.
Check your manual (if you have one) to see what packages need to be
installed.
> Third, configure your masquerade rules (the above Howto docs have the
> instructions, as does the xos.nl website).
Definitely read all the relevant HOWTO files, though. Even if you think
you know what you're doing, they're useful.
-- David
------------------------------
From: [EMAIL PROTECTED] (David C.)
Subject: Re: linux box vs switched hub
Date: 17 Sep 1999 16:48:29 -0400
Mark Evans <[EMAIL PROTECTED]> writes:
> H.Bruijn <[EMAIL PROTECTED]> wrote:
>> On Fri, 27 Aug 1999 16:28:45 GMT, chas allegedly wrote:
>>>
>>> Preparing for a discussion with our network administrator...
>>>
>>> What are the advantages (if any) of a switched hub over a linux
>>> router/firewall?
>>
>> they are totally different.
>> A switched hub is used to build the network, that the device is where
>> the cables from the NICs come together and the network traffic get's
>> routed to the correct computers. A switch nows which ip-number is on
>> which port (unlike a regular hub) so the packets get send directly
>
> Nope, it uses the MAC address and couldn't care less about ip
> addresses.
Depends on your definition of "switch". For a layer-2 switch, yes. For
a layer-3 switch, or a switch that has layer-3 features, no.
Blame the hub- and switch-vendors' marketing departments for the
confusion. (If it was up to me, the only "switches" around would be for
connection- oriented networks, like ATM or frame relay, and the things
we're talking about would only be called repeaters, bridges and
routers.)
-- David
------------------------------
From: Steve <[EMAIL PROTECTED]>
Subject: WfW/DOS w/TCP/IP can't use Samba share
Date: Fri, 17 Sep 1999 20:55:25 GMT
I want to use the MS DOS networking to access a Samba share to download
ghost images from the Linux server to PC's being set up. To experiment
I setup a Win 95 machine (which can see the Samba share and use it) with
alternate autoexec.bat/config.sys files and setup MS DOS
(WfW)networking on it - run strictly from a "DOS" command prompt not
under the GUI. I have netbui and tcp/ip setup to run on it using the
exact same IP as the Win 95 GUI uses on the same machine. I can ping to
the server and ping back but the "net use g: \\pc1\c password"
fails and asks for a password again. The valid password fails as well
as an all UPPER case try too. I do not have Samba setup to use encrypted
passwords. I can use the same setup to access a share on another Win 95
machine just fine. Workgroup is identical and a lmhosts and hosts file
are present with the right entries, indeed "ping pc1" works. Even using
the machines IP "net use g: \\192.168.3.56\c password" - fails the
same way. The share name is c.
Any ideas?
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
** 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
******************************
