Linux-Networking Digest #626, Volume #11 Tue, 22 Jun 99 17:13:50 EDT
Contents:
vacation probs with procmail ? (Christophe Zwecker)
Re: AOL and Linux (Sim)
forcing 100 Mbps SMC card with EPIC driver into 10 mbps using epic-diag (Padma
Krishnaswamy)
Re: Which Subnet to use ? ([EMAIL PROTECTED])
Re: Two network cards under 2.2.x (Bob Tennent)
Qusetion about setting up the Internet Big Picture using Linux ("Shawn Pursley")
Re: Realtek NIC HELP??????? (root)
Re: sendmail and NFS... ([EMAIL PROTECTED])
disabling transmit (Frank Pinzin)
Re: patch panels ([EMAIL PROTECTED])
Help! ethernet keeps going down ([EMAIL PROTECTED])
Re: Oversized Ethernet Frames? HELP! ([EMAIL PROTECTED])
Identd -- necessary? (Mike Kozlowski)
Re: SuSE Linux 6.1 & PPPIOCGUNIT Operation not permitted (Dan Glover)
Using MS VPN Clients Behind IP Masquerade ("Steve Spar")
ping (Manuel Zabelt)
Re: Linux - Cable modem configuration ("Jeff Volckaert")
Re: FTP on Linux Redhat ("Farhad Farzaneh")
pptp tunneling (Jason Bright)
network/interrupt cpu accounting question (Fred Polizo)
Re: dont get a LAN started (Nicholas E Couchman)
Re: FTP Questions...again... ("David Means")
Re: Linux SMTP/W98 Setup (Nicholas E Couchman)
Linux SMTP/W98 Setup (unixman)
----------------------------------------------------------------------------
From: Christophe Zwecker <[EMAIL PROTECTED]>
Subject: vacation probs with procmail ?
Date: Tue, 22 Jun 1999 18:35:31 +0000
Hi,
Ich use vacation updated on 99/01/03 on Linux Redhat 6.0 Kernel 2.2.10,
Sendmail 8.9.3
When I send mail to a user that has vacation enabled (via .forward) this
is what happens:
Jun 22 20:13:01 dumbo sendmail[25144]: UAA25143: SYSERR(root): mailer
prog died with signal 13
Jun 22 20:13:01 dumbo sendmail[25144]: UAA25143: to="|vacation deas",
ctladdr=<deas> (501/100),
delay=00:00:01, xdelay=00:00:00, mailer=prog, stat=Deferred
the local mailer is procmail-3.13.1
Any ideas ?
thx for help
--
Christophe Zwecker mail: [EMAIL PROTECTED]
Hamburg, Germany fon: +49 179 3994867
UNIX is user-friendly. It's just not ignorant-friendly and
idiot-friendly.Build a system even a fool can use,and only
a fool will want to use it.
------------------------------
From: [EMAIL PROTECTED] (Sim)
Subject: Re: AOL and Linux
Date: Tue, 22 Jun 1999 18:33:25 GMT
On Thu, 17 Jun 1999 07:12:07 GMT, Nicholas E Couchman
<[EMAIL PROTECTED]> wrote:
>As the guy below said, you can get Netmail plugin, but I was thinking that if
>you can establish a PPP connection via Linux to AOL, you might be able to get to
>the mail servers. I know they aren't accessible from outside w/o Netmail.
>--Nick
It used to be a plug in or activex, neither of which worked under
Linux, but now they use a pure HTML format on a secure server for
netmail so there is no problem - see aol.com
Instant messenger is of course possible with GAIM for Linux.
I think the AOL connection is plip, but I've never tried to connect
under linux as with IM and mail covered I cool
Simon
------------------------------
From: Padma Krishnaswamy <[EMAIL PROTECTED]>
Subject: forcing 100 Mbps SMC card with EPIC driver into 10 mbps using epic-diag
Date: Tue, 22 Jun 1999 14:49:36 -0400
Hi
Im running an SMC card with an EPIC driver on a 2.0.36 kernel.Im trying
to get the SMC card to operate at 10 Mbps; the router it is connected to
has autonegotiation off and speed set to 10 Mbps.
I downloaded epic-diag.c and libmii.c from the NASA site, and the output
to the epic-diag -m cmd is shown below.
I compiled epic-diag using cc -O -Wall -o
epic-diag epic-diag.c -DLIBMII libmii.c;
the output of epic-diag -m is as follows:(it is for the correct i/f)
.
.
.
(registers)
Basic mode control register 0x1000: Auto-negotiation enabled.
Basic mode status register 0x782d ... 782d.
Link status: established.
Capable of 100baseTx-FD 100baseTx 10baseT-FD 10baseT.
Able to perform Auto-negotiation, negotiation complete.
Vendor ID is 00:60:51:--:--:--, model 0 rev. 1.
Vendor/Part: Quality Semiconductor (unknown type).
I'm advertising 01e1: 100baseTx-FD 100baseTx 10baseT-FD 10baseT
Advertising no additional info pages.
IEEE 802.3 CSMA/CD protocol.
Link partner capability is 0021: 10baseT.
Negotiation did not complete.
**********************************
Why does negotiation seem complete in the first statement after 'Link
established' and not in the last?
The card seems to recongnise that the othe side 's capability is 10baseT
How can I force it to 10baseT?the -F option does not seem to work.
thanks
Padma Krishnaswamy
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Which Subnet to use ?
Date: Tue, 22 Jun 1999 19:24:58 GMT
In article <S1wb3.2$[EMAIL PROTECTED]>,
"Richi" <[EMAIL PROTECTED]> wrote:
> I am trying to hook up an ASDL line to my Linux box. I have two
> ethernet cards. The first is for the LAN and is 10.0.0.10/255.0.0.0.
I assume this is the address you'll be having other machines on the LAN
use as their defaultgateway? Traditionally [and following the
"standard"], these are given a host address of "1", meaning you should
be using 10.0.0.1 as the address. It won't break anything to use
10.0.0.10, but it will make anyone looking at your network think you
don't know what you're doing.
> The eth1 card for the Cisco router is 10.0.0.99. Should it be on a
> different subnet?
This is the card connecting to USWest? If so, then you should be
configuring it to use DHCP [which is how I assume they're doing dynamic
IP]. Only the "internal" NIC needs a 10/8 address.... the external
should be assigned a routable IP address.
--
Bill Clark
Systems Architect
ISP Channel
http://neighborhood.ispchannel.com/
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (Bob Tennent)
Subject: Re: Two network cards under 2.2.x
Date: 22 Jun 1999 19:11:04 GMT
Reply-To: rdt(a)cs.queensu.ca
On Tue, 22 Jun 1999 17:45:29 GMT, [EMAIL PROTECTED] wrote:
>We loaded RH6.0 into our training room over the weekend.
>Most everything went well, except that we couldn't get both network
>cards in the instructor machine to work. The first thing we tried
>was to change /etc/conf.modules to include:
>
>alias eth0 eepro
>alias eth1 eepro
>options eepro io=0x300,0x240
>
>Has anyone successfully brought up two network cards under
>RH6.0 (or under any Linux 2.2.x, for that matter)?. What
>did you have to do to get do it?
>
Try eepro.o and an options line with an irq and io for each card,
as in
alias eth0 ne.o
options ne.o io=0x6000 irq=11
alias eth1 ne.o
options ne.o io=0x6100 irq=10
Bob T.
------------------------------
From: "Shawn Pursley" <[EMAIL PROTECTED]>
Subject: Qusetion about setting up the Internet Big Picture using Linux
Date: Tue, 22 Jun 1999 18:25:17 GMT
We are embarking on stamping Microsoft products out of our company. We are
primarily an AS/400 shop and I want to be able to serve web pages
externally, maintain an ethernet intranet, plus send and receive e-mail.
The web pages will require dynamic data pulled off the AS/400. We are going
to be using a T-1 for the connection to the net with a Cisco router. This
router will also be connecting other branches to us via frame relay.
Here's my questions:
Would I use a universal IP for intranet such as 192.168.0.0 -
192.168.255.255 for all my clients locally connected?
What would I want to use as the IP for the webserver if I used the above
address, realizing that our customers will want access to the web server?
Right now, Internic has our web address, which is hosted by Internic, as
207.82.89.37. Does this need to be changed when we pull away from Tabnet,
our virtual host?
I will also be generating invoice from AS/400 based data that will be sent
to customer via e-mail. Is there some type of script out on the net for
doing something like this? I've got to imagine there are people doing this
already. I would not only need to pull the data off, but then I would need
to format it somehow.
Finally, security will be an issue for all of this stuff so if you guys (and
gals) can help with the placement of things (such as inside and outside of
the routers, firewalls, etc) it will be a big help.
The final part of this that is screwing me up the most is the order in which
to do all this.
Thanks for any and all help,
Shawn Pursley
------------------------------
From: root <[EMAIL PROTECTED]>
Subject: Re: Realtek NIC HELP???????
Date: Thu, 17 Jun 1999 18:50:37 +1000
Kyle Hittle wrote:
> i have a reltek pci 10/100 nic model number RTL8131A
>
> is this device supported by linux, please help me anyone, i need to config
> this nic so that i can hook up my T1
>
> thanx
> kyle
If you are using RedHat - then use the Net2K PCI . I have 2 installed - auto
detected.
David
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: sendmail and NFS...
Date: Tue, 22 Jun 1999 19:35:18 GMT
In article <[EMAIL PROTECTED]>,
"Stefan Monnier <[EMAIL PROTECTED]>"
<[EMAIL PROTECTED]> wrote:
> No excuse for putting mqueue on NFS. This is a *really bad* idea.
Indeed it is. The queue should never, ever, *ever* be on anything other
than the local drive. To do otherwise is completely wrong.
Consider: What happens if the NFS server croaks? Your poor mail daemon
will have no idea what's going on, and will be completely unable to
deliver mail -- even local mail.
Using NFS [either exported by or mounted by the mail server] for the
mail spool is also a very bad idea, since it _greatly_ increases the
likelihood of corrupting your mailboxes.
However, if you absolutely insist on having your spool on a remote
drive, I highly suggest using qmail or some other mail daemon that
supports maildirs. This will minimize the chance of corruption, though
it will mean you'll have to use wrappers if clients are reading mail
local [qpopper, which comes with qmail, will support POP3 clients
seamlessly]. Qmail has the added benefit of being able to deliver local
mail even if the network goes down completely [sendmail will more often
than not die completely].
--
Bill Clark
Systems Architect
ISP Channel
http://neighborhood.ispchannel.com/
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: Frank Pinzin <[EMAIL PROTECTED]>
Subject: disabling transmit
Date: Tue, 22 Jun 1999 20:16:49 +0000
I'm running a system with 2 network cards attached to two different
subnets. Is it possible to set the one card (an SMC EtherPower) to only
receive and not transmit?
--
Frank Pinzin [EMAIL PROTECTED]
Sheridan College (905) 845-9430 x2194
Network Systems Group
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: patch panels
Date: Tue, 22 Jun 1999 19:36:47 GMT
In article <7kmkmq$pp8$[EMAIL PROTECTED]>,
I wrote:
> I am designing a network for a dorm-style building with a linux box
> router running dhcp. My plan is to run 24 cat5 cables into hubs, run
[cut]
I should have been more specific. I am doing everything in this
building, and the thing is one solid brick of concrete. The cabling
goes from a dorm room wallplate along a runway in the suite, to a runway
in the hall, to a hub bolted and locked in a steel box inthe stairwell.
This cabling is specifically designed not to be touched until the actual
wires need to be replaced. The hubs will need to be upgraded in the
future but if more than two computers were put in any of these rooms
there would be no room for a bed. Is it possible/acceptable to simple
run solid/stranded (which is thinner?) cat5 from faceplate to hub,
crimping it into the faceplate connectors on one end and rj45s on the
other? Thanks for all the replies.
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED]
Subject: Help! ethernet keeps going down
Date: Tue, 22 Jun 1999 19:42:43 GMT
Hi,
I was wondering if someone could please help me.
I have a PC running RedHat 6.0 with a 3Com 3c509
10mb ethernet card in it hooked up to a cable
modem. The problem is that it works for some time
then stops working. To get it back up i have to
run
/sbin/ifconfig down
/sbin/ifconfig eth0 up
/sbin/route add default gw xxx.xxx.xxx.xxx
Then it will start working again. I tried this
with two differnet IP addresses that I have been
assigned and it still happens.
I was seriously considering running a business on
my Linux Box but if I can't get the ethernet
working I'll have to drop it. What can I do to fix
it? PS my motherboard has 5 pci slots and my
ethernet card is in a ISA slot. Would that matter?
Thanks
Desperate
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Oversized Ethernet Frames? HELP!
Date: Tue, 22 Jun 1999 19:16:21 GMT
In article <7kmsi3$224$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> [Description of boxes snipped]
> Can anyone shed any light for me?
> [Syslog? error messages snipped]
Post the output of an `ifconfig -a` [block out the IP if you're paranoid
about posting it]. In particular, I'd like to see what the MTU [Maximum
Transmission Unit] is.
--
Bill Clark
Systems Architect
ISP Channel
http://neighborhood.ispchannel.com/
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (Mike Kozlowski)
Subject: Identd -- necessary?
Date: 22 Jun 1999 19:23:19 GMT
I'm trying to tighten up a Linux box, so I'm removing non-essential
services from inetd.conf. However, I'm unsure if identd is essential or
not; reading the man page failed to illuminate.
What do I lose if I turn off identd?
--
Michael Kozlowski
http://www.cs.wisc.edu/~mlk/
------------------------------
From: Dan Glover <[EMAIL PROTECTED]>
Crossposted-To: uk.comp.os.linux,comp.os.linux.setup
Subject: Re: SuSE Linux 6.1 & PPPIOCGUNIT Operation not permitted
Date: Tue, 22 Jun 1999 19:24:57 GMT
Reply-To: [EMAIL PROTECTED]
In article <7knlrj$4ll$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
[oh, glad you got it working, shame that SuSE v6.1 doesn't seem to work
"out of the box"...]
>In article <[EMAIL PROTECTED]>,
> Malware <[EMAIL PROTECTED]> wrote:
>> > Jun 20 22:48:04 xenonsoft pppd[451]: sent [IPCP ConfReq id=0x3 <addr
>> > 192.237.75.1> <compress VJ 0f 01>]
>> > Jun 20 22:48:04 xenonsoft pppd[451]: rcvd [IPCP ConfNak id=0x3 <addr
>> > 193.237.75.1>]
>> Why do you insist on getting assigned the address 192.237.75.1 [...]
>AH ha
>That is because I reconfigured the system in the style of my Slackware
>distribution.
>I set the /etc/hosts
>127.0.0.1 localhost
>193.237.75.1 xenonsoft
>and of course /etc/HOSTNAME to `xenonsoft', because
>that what I thought you had to do to connect to Demon as
>`xenonsoft.demon.co.uk'. I am thinking of getting a second
>machine and the two machines cant obviously be called
>`xenonsoft' at the same time, even if I share the
>same dial-up connection betwen them.
Actually it doesn't matter whether your machine thinks it's
xenonsoft.d.c.u, though it can make things easier.
>Are you saying that my ISP is offering me an IP address
>and that `pppd' will create a `ppp[0-9]' device
>with this IP address?
Yes.
># These are the parameters. Change as needed.
>LOCAL_IP=0.0.0.0 # Local IP address if known. Dynamic = 0.0.0.0
>REMOTE_IP=0.0.0.0 # Remote IP address if desired. Normally 0.0.0.0
It will work perfectly if you leave both of these as 0.0.0.0, or don't
set them at all...
>LOCAL_IP=192.237.75.1
Oops - as has already been pointed out.
>REMOTE_IP=158.152.1.222
You can let Demon negotiate this. It may cause problems in the long run
if you hard code it (what happens if they change their configuration?).
The general rule for setting up PPP these days: unless you've been given
a good reason to set a value for a particular option, use the defaults
and let the protocol negotiate when you connect.
Dan
--
Dan Glover ([EMAIL PROTECTED])
Today's Excuse:
Hard drive sleeping. Let it wake up on it's own...
------------------------------
From: "Steve Spar" <[EMAIL PROTECTED]>
Subject: Using MS VPN Clients Behind IP Masquerade
Date: Mon, 21 Jun 1999 20:20:14 -0500
I have a Linux box dialed in to the Interent, running IP Masquerade which I
use to allow multiple MS (98/NT) boxes to be on the Internet at once.
I want to use the MS boxes with the MS PTPTP to connec tto a VPN over the
Net.
Can anybody tell me how to do this?
email response is appreciated
[EMAIL PROTECTED]
------------------------------
From: Manuel Zabelt <[EMAIL PROTECTED]>
Subject: ping
Date: Tue, 22 Jun 1999 21:23:07 +0200
Hello!
If I type �ping 127.0.0.1� everything works fine.
But if I want to �ping anyhost� almost nothing is shown, but there is
something going through the network(xosview tells me).
Only the first line �PING anyhost (IP): 56 data bytes� is shown.
If I quit (Ctrl-C) the following is shown:
--- anyhost ping statistic ---
�n packets transmitted, 0 packets received, 100% loss�
Everything else than ping (http,ftp,traceroute) works fine.
Does anyone know, how, I can solve this problem?
------------------------------
From: "Jeff Volckaert" <[EMAIL PROTECTED]>
Subject: Re: Linux - Cable modem configuration
Date: Tue, 22 Jun 1999 14:18:21 -0400
Just a quck thought... did you turn the cable modem off and on? I had alot
of grief with mine which happened to be that the cable modem was remembering
my old MAC address. I now cycle mine whenever I put a different box on it.
Jeff Volckaert
Geets <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Hi. I am currently up and running with a Motorola cable modem (provided
> by Comcast using @home service) attached to an ethernet hub and then out
> to 2 PC's, one WinNT Server, the other Win98, In the interest of not
> leaving well enough alone (and for other not nearly as important reasons,
> like getting fired if I don't), I need to use a third PC running Red Hat
> 5.2 as an internet server.
>
> Before I network the three machines, I thought I'd attach the cable modem
> directly to the linux box, and get that connection working. No luck so
> far. I think I'm doing something stupid, so if you will humor me a bit
> I'll describe how I'm correctly configured on the Win NT box, and how I'm
> incorrectly configured on the Linux.
>
> WinNT:
>
> TCP/IP Properties:I
> IP Address tab: "Obtain an IP address from a DHCP server" is checked.
> Also on the "Advanced" button --> "DHCP enabled" is checked.
>
> DNS tab: Host bame: cc262221-a
>
> WINS Address tab: "Enable LMHOSTS Lookup" checked
>
> DHCP relay tab: Seconds threshold: 4; Maximum hops: 4
>
> Routing tab: "Enable IP forwarding" is checked
>
> And that's pretty much all there is
>
> ----------------
> Linux:
>
> For the Linux machine, in control panel I'm configured as follows:
>
> Names tab:
>
> Hostname: localhost
> Domain: localhost.localdomain
>
>
> Hosts tab:
>
> IP address Name Nicknames:
> 127.0.0.1 cc262221-a Linux
> 192.168.0.4 PIII-1 WinNT Box
> 192.168.0.9 PII-1 Win98 Box
> 127.0.0.1 cc262221-a Linux
> 192.168.0.7 foo.bar.com LinuxBox foo
>
> Note, on the above I was not able to configure the first 127.0.0.1 IP
> address to use localhost as its name. Nor was I able to leave the second
> 127.0.0.1 IP address blank (DHCP protocol is enabled in the Broadcast
> tab. I tried to leave the IP address blank so my ISP could provide one
> dynamically, but linux won't let me)
>
>
> Interfaces Tab:
>
> IP proto at boot active:
> lo 127.0.0.1 none yes active
> ppp0 none no inactive
> eth0 dhcp yes inactive
>
> When I press "Edit" button for eth0, I get:
> IP: field is blank
> Netmask: field is blank
> Network: 192.168.0.0
> Broadcast: 192.168.0.255
> "activate interface at boot time" is checked
> Interface configuration protocol: DHCP
>
> Also, when I run ifconfig I'm informed that only lo is running, even
> after setting eth0 to active and at boot. In fact, when I save and close
> control panel, and re-open it, eth0 is listed as inactive again.
>
>
>
> So, there you have it. Any kind soul willing to help will be the
> beneficiary of my good wishes. Thanks.
>
> Vince
>
> PS email to [EMAIL PROTECTED] but remove those "XXX's"
> first.
>
>
------------------------------
From: "Farhad Farzaneh" <[EMAIL PROTECTED]>
Subject: Re: FTP on Linux Redhat
Date: Tue, 22 Jun 1999 13:06:50 -0700
On my linux RH5.1 there was no ftp server configured. I had to download and
build the wu-ftpd program. This is available as an rpm on the redhat site.
--Farhad
==========
In article <7ko2ea$2ur5$[EMAIL PROTECTED]>, "A. de Vos" <[EMAIL PROTECTED]>
wrote:
> I cannot FTP from Windows to Linux through the LAN (ethernet). But I am able
> to connect to the Apache server through Explorer.
>
> How is FTP server started on Linux? And how can I reach it?
>
> Arne de Vos
>
>
>
------------------------------
From: [EMAIL PROTECTED] (Jason Bright)
Subject: pptp tunneling
Date: 22 Jun 1999 18:05:42 GMT
Looking for info from anyone who has successfully set up a linux pptp client
to tunnel into an NT network. The connection will be done over an already
established TCP/IP connection (not over a ppp link). About to try out
C.S. Ananian's pptp client but doesn't have a lot of docs. Hoping to cut
down on the trial and error if anyone has already covered this ground....
thx
j
------------------------------
From: Fred Polizo <[EMAIL PROTECTED]>
Subject: network/interrupt cpu accounting question
Date: Tue, 22 Jun 1999 12:03:39 -0700
I've got a uniprocessor PC with RedHat6.0 setup to test simple packet
forwarding between two ethernet interfaces. No user applications are
running. The cpu is 100% idle according to vmstat. What is odd is that
when the PC is forwarding many thousands of packets/sec, vmstat still
reports the cpu virtually 100% idle. That seems wrong. Why is there
no reported overhead for processing packets?
In looking at the 2.2.5 kernel code, (redhat 6.0), I noticed that if
task 0 is running when the system clock interrupt occurs, then the time
is considered "idle" time. This seems to imply that indeed your linux
box can be very busy, yet appear to be idle, when looking at system
(kstat) statistics. CPU accounting should be more accurate when you
have a CPU bound user process that keeps the kernel idle thread from
running. Have I missed something here?
How do folks out there running linux as a router monitor cpu usage?
Does anyone have a work-around or kernel patch for seeing accurate cpu
usage for network processing in such a case?
Thanks,
---Fred P.
------------------------------
From: Nicholas E Couchman <[EMAIL PROTECTED]>
Subject: Re: dont get a LAN started
Date: Tue, 22 Jun 1999 20:34:49 GMT
Make sure you are pinging using IP addresses. I assume you know this
because you were able to setup an IP on Linux and you didn't use a DNS
server. Also, just for luck, you might want to add some or all of the PCs
on your network to /etc/hosts. It helps a lot sometimes.
--Nick
Saraal wrote:
> Hi
> I've tried to connect my linux-pc to a small LAN (� 10 pc's, ethernet)
>
> There was a problem however: the other pc's could ping me (they were
> all windows systems) but i wasn't able to ping the other pc's
> i configured the following things:
> - my IP
> - netmask
> -host
> -domein
>
> We didn't have a gateway, nor did i configure a dns server.
> The hubs (and the rest of the netwerk) were running at 10Mbps, but my
> networkcard (3com ethlnk3) was running at 100Mbps, altough i
> configured it for 10.
>
> What did I do wrong, and what can i do to make it all work?
> I used SuSE 6.0 linux. (and i'm not really an experienced linux user
> :)
>
> greetings, Mathijs
------------------------------
From: "David Means" <[EMAIL PROTECTED]>
Subject: Re: FTP Questions...again...
Date: 22 Jun 1999 18:17:36 GMT
Tony C <[EMAIL PROTECTED]> wrote in message
news:7ko7vs$aeb$[EMAIL PROTECTED]...
> Greetings,
>
> I apologize for the redundant question but I am in need of some help in
> getting ftp between a Win98 and Linux PC working. I am running RH5.2 on
the
> Linux PC along with ipfwadm for IP masquerading. I have a local private
net
> of 2 Win PCs connected through as 10/100 hub to the Linux PC. The Linux PC
> has 2 ethernet cards, one attached to the private net and the other
attached
> to my DSL modem. Everything is working as far a accessing the Internet. I
> can even use my Web browser to ftp items from various sites on the net.
>
> What I can't seem to do right now is ftp from the Windows PC to the Linux
> box. When I invoke ftp from a DOS shell I get the following message:
>
> C:\ftp 192.168.1.1
> > ftp: connect: 10061
> ftp>
>
> At which point if I try to provide a user login or anything else for that
> matter ftp responds with a 'not connected' message. If I try to ftp to the
> Lan interface that is connected to the DSL line I get nothing. It just
hangs
> until I Ctrl-C or close the DOS window.
>
> I am sure this is a simple problem to fix and the solution is surely
staring
> me in the face. If you can help me 'see' the answer I would greatly
> appreciate it.
Have you checked that your FTP daemon is set up to run when the
attempt is made to connect to the Linux box ? Check in /etc/inet.conf,
where there should be a line under standard services saying:
ftp stream tcp nowait root /usr/sbin/tcpd in.ftpd -l -a
or words to that effect. If this is commented out, you might get this
problem
Failing that, is your masquerading set up to accept all kinds of packets
on the
private port of the Linux machine ? That also could cause the failure to
even get
connected.
------------------------------
From: Nicholas E Couchman <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Linux SMTP/W98 Setup
Date: Tue, 22 Jun 1999 20:30:37 GMT
It isn't a Wintel problem on this one, I don't think, so don't totaly trash
Win98 (although 98 is just a cheap scam by Microsoft to get you to use
Internet Explorer, which is why I'm sticking with 95 or NT 4). It sounds
like your SMTP server (I would guess sendmail?) is having a problem. There
is some line you need to add to the syslog.conf (whatever its called) to
make syslog do logging for certain daemons. I can't remembe what the line
is, but there is some way to forward all messages from inetd to your
console. It sound more like Linux is acting up than 98.
--Nick
unixman wrote:
> Ok, I am about to pull my damn hair out on this one.
>
> I have a RH6.0 box up and running with a live connection to the net.
> Everything seems to work fine, and rightly so, as I consider myself a
> fairly advanced UNIX developer. Of course, then there's friggin'
> Windows 98 - the bane of my existence right now.
>
> I connect to the POP3 server on the Linux box for mail. No problem. I
> can receive mail just fine (using a variety of clients including
> Communicator and Eudora). The problem is when I try to send mail out.
> It works sometimes - but after a while, things just sort of "hang".
>
> In fact, while I can perform regular telnets to the Linux box (even on
> custom user ports where I have some other apps running), when I try to
> telnet to the SMTP port, it just sits there like a damn knot on a log.
> Nothing in syslog, nothing reported on netstat. Of course, everything
> works fine from the console, just remote clients. I can hit our httpd
> daemon, regular telnet, ftp, everything. It just seems to have problems
> connecting to the SMTP port (sendmail running on 25).
>
> My Windows setup is typical. My outgoing SMTP hostname is correct (I've
> even tried using the IP).
>
> Has ANYONE seen this? I'm about to throw this damn thing out the window
> and dive out after it!
------------------------------
From: unixman <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Linux SMTP/W98 Setup
Date: Tue, 22 Jun 1999 16:19:55 -0700
Ok, I am about to pull my damn hair out on this one.
I have a RH6.0 box up and running with a live connection to the net.
Everything seems to work fine, and rightly so, as I consider myself a
fairly advanced UNIX developer. Of course, then there's friggin'
Windows 98 - the bane of my existence right now.
I connect to the POP3 server on the Linux box for mail. No problem. I
can receive mail just fine (using a variety of clients including
Communicator and Eudora). The problem is when I try to send mail out.
It works sometimes - but after a while, things just sort of "hang".
In fact, while I can perform regular telnets to the Linux box (even on
custom user ports where I have some other apps running), when I try to
telnet to the SMTP port, it just sits there like a damn knot on a log.
Nothing in syslog, nothing reported on netstat. Of course, everything
works fine from the console, just remote clients. I can hit our httpd
daemon, regular telnet, ftp, everything. It just seems to have problems
connecting to the SMTP port (sendmail running on 25).
My Windows setup is typical. My outgoing SMTP hostname is correct (I've
even tried using the IP).
Has ANYONE seen this? I'm about to throw this damn thing out the window
and dive out after it!
------------------------------
** 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
******************************
