Linux-Networking Digest #740, Volume #9 Sat, 2 Jan 99 02:13:50 EST
Contents:
Shot networking capabilities to hell ([EMAIL PROTECTED])
Any MultiPPP development? (Chew Yu Chou)
dos client (Keith Hewitt)
Netatalk - setting default Appletalk zone?? (Cameron Moller)
Re: Looking for SAMBA book/help ("Corey M. Horton")
Re: How to build a Linux cluster? (Robert Clouse)
Re: Looking for SAMBA book/help ("Corey M. Horton")
Which is good standalone nntp server? (Constantin Riabitsev)
Re: pppd works, diald fails initializing modem. (Rod Pike)
Re: linux under MS proxy 2.0 ("compuware.com")
mismatched read page pointers (Jerry Dietrich)
Re: rc.local file with RedHat 5.0 ([EMAIL PROTECTED])
Poor performance with Redhat 5.1 ipfwadm (Francois FOURNEL)
Unknown NFS error message. (Anton Persson)
Re: Networking Problem (Neil Chambers)
tcp wrapper terribly slow ("Thomi Hug")
modem is sssslllloooowwwww under RH5.2 ("Daniel A. Jablonski")
any suggestions for a dual port nic? (Markus Schulte)
Re: NFSing a tape drive ? (Eric Jorgensen)
Re: dos client (patrick thempel)
Network behind PPP ("Tristan Braun")
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Subject: Shot networking capabilities to hell
Date: Sat, 02 Jan 1999 06:38:39 GMT
I was merrily typing along on a paper running netscape, x11amp, ncftp, and
some other apps when I got up to get some coffee. My foot caught my keyboard
cable sending it crashing to the ground. I picked it up and snapped the case
back together (this was at one time a frequent occurence) and got coffee.
When I looked back my screen was frozen. I'm assuming that this happened
immediatly upon the falling of my keyboard but there was about five min
inbetween checking the screen and picking up the keyboard. Onto what
matters... Everything was locked up so I tried everything I knew to do.
First Ctrl-Alt-Function key to get to another terminal, then
ctrl-alt-backspace to get out of X. Then I went up to my housmate's computer
and tried telneting in, then pinging. All of these gave me a fairly dead
computer, no different that as if it was unpluged and in the box in which it
arrived. I powered down after my first total crash in Linux in a year and a
half of running it and brought it back up again and resumed typing. I
evetually opened up ncftp to resume my downloads and found that I had no
network access. I can ping 127.0.0.1 and my real ip, but pinging anything
else gives me 100% packet loss, though network control panel is sure that
eth0 is active, as is /proc/. Running RedHat5.1, kernels 2.0.36 and 2.1.130,
not sure which was active on crash but niether works now. Logs give me
nothing as far as errors dealing with ehthernet are concerned. I can still
telnet, apache, ftp, etc into my own ip. None of my configs have new dates
or strange info in them that wasn't there before. I'm not sure how i'll go
about wasting time during finals week without net access. Back to
StarOffice..... -Lucas
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: Chew Yu Chou <[EMAIL PROTECTED]>
Subject: Any MultiPPP development?
Date: Sat, 02 Jan 1999 06:38:39 GMT
Does anyone knows whether there is any development going on the
implement a pppd that supports Multilink which allows
Win 95 to dial into a Linux server using Multilink PPP ? Anyone who have
any info on this can leave me an email? Thanks!
------------------------------
From: Keith Hewitt <[EMAIL PROTECTED]>
Subject: dos client
Date: Sat, 02 Jan 1999 06:39:59 GMT
Help
I'm looking to implement samba networking, trouble is I have a couple of
286 machines which have to be included (ie not win 3.11 / w95). How do I
get the old wfw for dos to talk to samba? where can I get a tcp/ip stack
for dos?
Thanks for any help
Keith Hewitt
WENN London
<[EMAIL PROTECTED]>
------------------------------
From: Cameron Moller <[EMAIL PROTECTED]>
Subject: Netatalk - setting default Appletalk zone??
Date: Sat, 02 Jan 1999 06:38:40 GMT
Does anyone know how or where to set a default Appletalk Zone with
Netatalk.
I have RedHat Linux 5.1 and Netatalk 1.42b
We have a multi-zoned Appletalk Network, the system default is "UMB Net"
which is where my Linux box appears, by default.
I want to appear in a different zone i.e. "UMB Biology"
I am unsure which file I need to modify. I think it might be
/usr/local/atalk/etc/afpd, adding the line
afpd -n @"UMB Biology"
Thanks in advance to anyone who might shed light on this.
Cameron
--
****************************************************************************
Cameron Gray Moller e-mail:
[EMAIL PROTECTED]
Department of Biology
http://fisher.bio.umb.edu/index.htm
University of Massachusetts at Boston Phone:
617-287-6643
100 Morrissey Boulevard Fax:
617-287-6650
Boston, MA 02125-3393 Office:
W-3-036
****************************************************************************
------------------------------
From: "Corey M. Horton" <[EMAIL PROTECTED]>
Subject: Re: Looking for SAMBA book/help
Date: Sat, 02 Jan 1999 06:40:51 GMT
Samba:Integrating UNIX and Windows, by John D. Blair. ISBN#1-57831-006-7. 298
pages. Publisher is SSC (Specialized Systems Consultants, Inc.). Comes with a CD
ROM containing the most current version of Samba at the time of printing.
Talk about a thorough text...... I actually had to go the web to look for a more
generic, simplified set of directions to get Samba up and running. If you're
looking to do any tweaking, you'll find the info you need in this book.
James Knowles wrote:
> Scott,
>
> Your setup sounds similar to mine. Look at /etc/smb.conf (?), as this is the
> control point for Samba. There is also a book on Samba, which I received last
> Friday. I don't have it with me, but it's called "Samba: Integrating Unix and
> Windows." Sorry, I don't remember the author's name. He's one of the Samba team
> members.
>
> Samba has an overwhelming set of features, but works like a charm. I'm using it
> for a Win32 file and print server. Linux runs the rest of the system except for
> Win32 production work.
>
> I'm still learning how to use it. I still need to get my NT box to hook up to
> the domain, and figure out why smbmount is failin. I want to back up my windows
> machines from the DAT drive on Linux.
>
> If you get stuck, I may (or may not!) be able to to answer, but I'm willing to
> try.
>
> Good luck,
>
> James
------------------------------
From: Robert Clouse <[EMAIL PROTECTED]>
Subject: Re: How to build a Linux cluster?
Date: Sat, 02 Jan 1999 06:40:54 GMT
"K. K@FKA" wrote:
>
> Hi,
> I would like to build a Linux cluster from 3 computer (PII 300) to
> become a server for out department I have no idea where to find
> documentation on it (if any) or how to do it (if it's possible).
>
> Thanks for any help or advise you can give me,
> K.
Also, look around under PVM, Parrallel Virtual Machine
------------------------------
From: "Corey M. Horton" <[EMAIL PROTECTED]>
Subject: Re: Looking for SAMBA book/help
Date: Sat, 02 Jan 1999 06:41:03 GMT
Hello All,
Try this:
========
Samba:Integrating UNIX and Windows, by John D. Blair. ISBN#1-57831-006-7. 298
pages. Publisher is SSC (Specialized Systems Consultants, Inc.). Comes with a CD
ROM containing the most current version of Samba at the time of printing.
Talk about a thorough text...... I actually had to go the web to look for a more
generic, simplified set of directions to get Samba up and running. If you're
looking to do any tweaking, you'll find the info you need in this book.
========================================================================
Jose Antonio C. Baduria wrote:
> Hello Scott,
>
> Try the book "Linux Network Toolkit" by Paul G. Sery. It deals a lot
> about samba and setting up
> a small network. It helped me a lot in setting up my windows 95 machines as
> clients of a Linux
> server.
>
> I hope this helps.
>
> Scott Palmer wrote:
>
> > I don't mind paying money to buy a good book that will teach me how to set
> > up Samba on Linux, especially if it will take me through in great detail,
> > step-by-step.
> >
> > Any ideas?
------------------------------
From: Constantin Riabitsev <[EMAIL PROTECTED]>
Subject: Which is good standalone nntp server?
Date: Sat, 02 Jan 1999 06:41:12 GMT
Hi, all:
I have configured INN-2.1 to work as a standalone server (just our
company's newsgroups), but I keep thinking that running INN for
these purposes is an overkill. Since it's not feeding anything
anywhere, is it rational to use it as a standalone? It doesn't seem
to devour so much memory (1.2% on a 128Mb machine), but maybe
someone has something more suitable to suggest?
I have tried nntp, but couldn't configure it to run under Linux.
Also, INN was a pain to configure, so there has to be a gooood
reason for me to want to switch... :))
Thanks!
--
Constantin Riabitsev,
Internet Development/Wide Area Networks
SEJ Administrative Council, U.M.C.
Running RedHat Linux-2.0.36 (Apollo)
------------------------------
From: Rod Pike <[EMAIL PROTECTED]>
Subject: Re: pppd works, diald fails initializing modem.
Date: Sat, 02 Jan 1999 06:41:20 GMT
It was suggested I turn on debug and look at the log. Added debug option to
/etc/ppp/options and /etc/diald.conf
I also removed proxyarp.
Here is the output from /var/log/debug
Dec 13 21:05:58 mycomputer diald[653]: Starting diald version 0.16.5
Dec 13 21:05:58 mycomputer diald[653]: Proxy device established on interface
sl0
Dec 13 21:05:58 mycomputer diald[653]: running '/sbin/ifconfig sl0
local.IP.address pointopoint 95.box.IP.address mtu 1500 up'
Dec 13 21:06:00 mycomputer diald[653]: SIGCHLD[1]: pid 654 system, status 0
Dec 13 21:06:00 mycomputer diald[653]: Setting pointopoint route for sl0
Dec 13 21:06:00 mycomputer diald[653]: running '/sbin/route add
95.box.IP.address metric 1 dev sl0'
Dec 13 21:06:00 mycomputer diald[653]: SIGCHLD[2]: pid 660 system, status 0
Dec 13 21:06:00 mycomputer diald[653]: Establishing routes for sl0
Dec 13 21:06:00 mycomputer diald[653]: running '/sbin/route add default
metric 1 netmask 0.0.0.0 dev sl0'
Dec 13 21:06:00 mycomputer diald[653]: SIGCHLD[3]: pid 661 system, status 0
Dec 13 21:06:00 mycomputer diald[653]: Changed snoop device to sl0
Dec 13 21:06:00 mycomputer diald[653]: Diald initial setup completed.
Dec 13 21:06:02 mycomputer diald[653]: new state CONNECT action 0x804ee14
timeout 60
Dec 13 21:06:03 mycomputer diald[653]: Running connect (pid = 663).
Dec 13 21:06:03 mycomputer chat[663]: report (CONNECT)
Dec 13 21:06:03 mycomputer chat[663]: abort on (BUSY)
Dec 13 21:06:03 mycomputer chat[663]: abort on (ERROR)
Dec 13 21:06:03 mycomputer chat[663]: abort on (NO CARRIER)
Dec 13 21:06:03 mycomputer chat[663]: abort on (NO DIALTONE)
Dec 13 21:06:03 mycomputer chat[663]: abort on (NO ANSWER)
Dec 13 21:06:03 mycomputer chat[663]: send (ATZ^M)
Dec 13 21:06:03 mycomputer diald[653]: SIGCHLD[4]: pid 663 dial, status 1024
Dec 13 21:06:03 mycomputer diald[653]: Connect script failed.
Dec 13 21:06:03 mycomputer diald[653]: new state CLOSE action 0x804f4a0
timeout -1
Dec 13 21:06:03 mycomputer diald[653]: Closing modem line.
Dec 13 21:06:03 mycomputer chat[663]: expect (OK)
Dec 13 21:06:03 mycomputer chat[663]: report (CONNECT)
Dec 13 21:06:03 mycomputer chat[663]: abort on (BUSY
Dec 13 21:06:03 mycomputer chat[663]: -- failed
Dec 13 21:06:03 mycomputer chat[663]: Failed (BUSY)
Dec 13 21:06:04 mycomputer diald[653]: Delaying 30 seconds before clear to
dial.
doesn't do anything for me but I'm hoping someone else finds it meaningful.
Rod Pike wrote:
> Any information that can help solve my problem would be greatly
> appreciated.
>
> Here are the details
>
> redhat ver 5.1
> diald ver 0.16.5
> ppp ver 2.3.3
>
> here are my files with user specific information changed to protect the
> niave:)
>
> /etc/ppp/chatscript
>
> REPORT CONNECT
> ABORT BUSY
> ABORT ERROR
> ABORT "NO CARRIER"
> ABORT "NO DIALTONE"
> ABORT "NO ANSWER"
> '' ATZ
> OK IPphoneNumber
> 'ipname :' myuserid
> Password: mypassword
>
> /etc/diald.conf
>
> mode ppp
> connect '/usr/sbin/chat -v -s -f /etc/ppp/chatscript'
> device /dev/cua1
> speed 115200
> modem
> lock
> crtscts
> local localIPaddress
> remote remoteIPaddress
> dynamic
> defaultroute
> proxyarp
> include /usr/lib/diald/standard.filter
> retry-count 2
> redial-timeout 30
>
> I start up ppp with the following script
>
> /root/startppp
>
> #!/bin/sh
> /usr/sbin/pppd /dev/cua1 115200 connect 'chat -v -s -f
> /etc/ppp/chatscript' defaultroute crtscts proxyarp lock
>
> When I try diald it fails initializing the modem and aborts the chat
> script. I'm using the exact same chatscript for ppp and diald
>
> here is what the /var/log/message looks like for a successful ppp
>
> Dec 13 20:32:26 mycomputer pppd[508]: pppd 2.3.3 started by root, uid 0
> Dec 13 20:32:27 mycomputer chat[509]: report (CONNECT)
> Dec 13 20:32:28 mycomputer chat[509]: abort on (BUSY)
> Dec 13 20:32:28 mycomputer chat[509]: abort on (ERROR)
> Dec 13 20:32:28 mycomputer chat[509]: abort on (NO CARRIER)
> Dec 13 20:32:28 mycomputer chat[509]: abort on (NO DIALTONE)
> Dec 13 20:32:28 mycomputer chat[509]: abort on (NO ANSWER)
> Dec 13 20:32:28 mycomputer chat[509]: send (ATZ^M)
> Dec 13 20:32:28 mycomputer chat[509]: expect (OK)
> Dec 13 20:32:28 mycomputer chat[509]: ATZ^M^M
> Dec 13 20:32:28 mycomputer chat[509]: OK
> Dec 13 20:32:28 mycomputer chat[509]: -- got it
> Dec 13 20:32:28 mycomputer chat[509]: send (ATDT5551212^M)
> Dec 13 20:32:28 mycomputer chat[509]: expect (myIPname :)
> Dec 13 20:32:28 mycomputer chat[509]: ^M
> Dec 13 20:32:43 mycomputer chat[509]: ATDT5551212^M^M
> Dec 13 20:32:43 mycomputer chat[509]: CONNECT
> 33600/ARQ/V34/LAPM/V42BIS^M
> Dec 13 20:32:44 mycomputer chat[509]: ^M^M
> Dec 13 20:32:44 mycomputer chat[509]: myIPname :
> Dec 13 20:32:44 mycomputer chat[509]: -- got it
> Dec 13 20:32:44 mycomputer chat[509]: send (myuserid^M)
> Dec 13 20:32:44 mycomputer chat[509]: expect (Password:)
> Dec 13 20:32:44 mycomputer chat[509]: mypassword^M
> Dec 13 20:32:44 mycomputer chat[509]: Password:
> Dec 13 20:32:44 mycomputer chat[509]: -- got it
> Dec 13 20:32:44 mycomputer chat[509]: send (mypasswrod^M)
> Dec 13 20:32:44 mycomputer pppd[508]: Serial connection established.
> Dec 13 20:32:45 mycomputer pppd[508]: Using interface ppp0
> Dec 13 20:32:45 mycomputer pppd[508]: Connect: ppp0 <--> /dev/cua1
>
> and the failed diald log
>
> Dec 13 20:09:15 mycomputer diald[2207]: found interface eth0 for proxy
> arp
> Dec 13 20:09:25 mycomputer diald[2207]: Running connect (pid = 2220).
> Dec 13 20:09:26 mycomputer chat[2220]: report (CONNECT)
> Dec 13 20:09:26 mycomputer chat[2220]: abort on (BUSY)
> Dec 13 20:09:26 mycomputer chat[2220]: abort on (ERROR)
> Dec 13 20:09:26 mycomputer chat[2220]: abort on (NO CARRIER)
> Dec 13 20:09:26 mycomputer chat[2220]: abort on (NO DIALTONE)
> Dec 13 20:09:26 mycomputer chat[2220]: abort on (NO ANSWER)
> Dec 13 20:09:26 mycomputer chat[2220]: send (ATZ^M)
> Dec 13 20:09:26 mycomputer diald[2207]: Connect script failed.
> Dec 13 20:09:26 mycomputer chat[2220]: expect (OK)
> Dec 13 20:09:26 mycomputer chat[2220]: report (CONNECT)
> Dec 13 20:09:26 mycomputer chat[2220]: abort on (BUSY
> Dec 13 20:09:26 mycomputer chat[2220]: -- failed
> Dec 13 20:09:26 mycomputer chat[2220]: Failed (BUSY)
> Dec 13 20:09:27 mycomputer diald[2207]: Delaying 30 seconds before clear
> to dial.
>
> Any ideas???
>
> Thanks in advance
>
> Cheers,
>
> Rod
>
> [EMAIL PROTECTED]
------------------------------
From: "compuware.com" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux
Subject: Re: linux under MS proxy 2.0
Date: Sat, 02 Jan 1999 06:41:28 GMT
It actually does more than just registry changes. It copies a modified
version of the winsock.dll toyour computer.
news.netway.at <[EMAIL PROTECTED]> wrote in message
news:74oh9d$ccj$[EMAIL PROTECTED]...
>
>
>>The last one naturally only works with MS Windows based winsock.dll OS's
>>such as Win 3.1 and Win 9x/NT. I'm willing to bet that thats what is
being
>>used since after you run the \\proxy\mspclnt\setup.exe it installs it.
>
>That is absolutly correct. This setup makes some registry-entries reffering
>to the proxy.
>
>
------------------------------
From: Jerry Dietrich <[EMAIL PROTECTED]>
Subject: mismatched read page pointers
Date: Sat, 02 Jan 1999 06:35:58 GMT
I am getting "eth0: mismatched page pointers" when I mount my cdu31a
cdrom. My system is the following.
DX2 66, 16Megs, 2 IDE hd, ne2000 adapter, cdu31a cdrom with sun moon
interface card I am running linux version 2.0.34
I have tried all the following.
1) Changing the io port address of both the adapter and the interface
card(Many, many different combos, sometimes making the adapter having
the lower address space and sometimes higher. I always keep the
difference between the two io ports at least 0x20.
2) Tried changing the the irqs of both the adapter and interface card.
3) Installed RedHat 5.1 and slackware 3.5 same results.
4) Putting the cards in different slots.(like the good old days??)
As soon as I mount the cdrom and the next ethernet packet is recieved
the messages start
flowing. If I unmount the cdrom or disconnect the network the messages
stop. The cdrom works just find because I install linux using it. It
also works after the network is disconnected. It is probably working
with the network connected but it is impossible to tell with all the
mismatched messages being displayed.
I would like to submit a bug report on this if it truly is one. Any
info/experience with this type of problem would greatly be appreciated.
Thanks
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: rc.local file with RedHat 5.0
Date: Sat, 02 Jan 1999 06:41:48 GMT
In article <[EMAIL PROTECTED]>,
Simon Andre Simonsen <[EMAIL PROTECTED]> wrote:
> Why doesn't the commands I add in my rc.local file get runed. Here is
> the commands
[snip]
Perhaps you should try running rc.local manually to see if there are any
syntax errors or whatnot. I once had a problem with a stray quote mark.
>
> If you do have any suggestions feel free to E-mail me.
>
>
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: Francois FOURNEL <[EMAIL PROTECTED]>
Subject: Poor performance with Redhat 5.1 ipfwadm
Date: Sat, 02 Jan 1999 06:39:20 GMT
I'm using a Linux Redhat 5.1 box as an ip filter between a lan and a wan
My kernel version is 2.0.35 and i configure ipfwadm for filter purpose.
My linux box is running with two Olicom 3137 Token Ring network card
Without the ipfwadm up, all works fine and i get from/to lan/wan with
good performances
As soon as I configure ipfwadm, the ip filter works as i want, but it's
a very
long time for getting an accept session, i.e. a telnet session from the
wan to
a host on the Lan
I used ipfwadm with "-F" feature (forwarding)
One rule for telnet wan/lan acceptance:
ipfwadm -F -a accept -P tcp -S 0.0.0.0/0 -D 192.168.1.0/24 23
Any ideas ?
Note that I don't use DNS requests at this time but just the IP
addresses
--
Francois FOURNEL
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
NCTech - 8 rue Hermann Frenkel - 69007 LYON
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Phone: +33 (0) 478.61.46.28 - Fax: +33 (0) 478.61.46.99
Email: [EMAIL PROTECTED] - WWW: http://www.nctech.fr
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
------------------------------
From: Anton Persson <[EMAIL PROTECTED]>
Subject: Unknown NFS error message.
Date: Sat, 02 Jan 1999 06:41:59 GMT
Program: NFS-server-2.2beta37.(Slackware 3.6)
Error: (/var/adm/messages)
Dec 14 16:52:22 pc1 nfsd[77]: strange write req from amiga1.holiday.com:
count 511418268 len 311
---
I was trying to copy some files from my Amiga to my Linux system. I can
create the
file but no data is written to the file(size 0). What does the error
message really mean?
Note: It's a serial link running 38400 bps.
//Anton Persson, Holiday Studios International.
------------------------------
From: Neil Chambers <[EMAIL PROTECTED]>
Subject: Re: Networking Problem
Date: Sat, 02 Jan 1999 06:39:48 GMT
Neil Chambers wrote:
> My 2 computers still won't ping
>
> where do i put route add -net 192.168.251.0 eth0
>
> i typed this, and it didn't give an error, but still wouldn't ping the
> other machine?
>
> --
> Later
>
> Neil
> [EMAIL PROTECTED]
> www.hoser.net
Anothing thing,
if i uncomment the gateway, and gateway dev in /etc/sysconfig/network it
gives a SIOCADDRT error
--
Later
Neil
[EMAIL PROTECTED]
www.hoser.net
------------------------------
From: "Thomi Hug" <[EMAIL PROTECTED]>
Subject: tcp wrapper terribly slow
Date: Sat, 02 Jan 1999 06:42:00 GMT
hello
im running a redhat 5.1 linux on my pc.
every daemon who is running with tcp (wuftp, telnet, ...) is terribly slow. every
other daemon like apache or
samba work perfectly.
when i connect to telnet via network, i have to wait about 2 minutes for the connect...
help!!!
thanks
tom
------------------------------
From: "Daniel A. Jablonski" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: modem is sssslllloooowwwww under RH5.2
Date: Sat, 02 Jan 1999 06:42:02 GMT
==============42E199371E3ACFB3610F02FA
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Okay,
I have all the classic symptoms of an irq conflict with my modem under
my installed-as-of-yesterday RH 5.2 system. It will dial, and will
connect, but is tortuously slow. This goes for ppp and minicom.
I was previously running RH 5.1, upgraded from RH 4.2 (still the best
release), under which all of my hardware - none of which has changed -
worked great, including the modem. I initialized it in
/etc/rc.d/rc.local with setserial to use /dev/cua2 0x3e8, and IRQ 7, and
I pass it the 'spd_vhi' option as well - this is how the modem is
jumpered. Now the serial mouse is on /dev/cua0, which means 0x3f8
irq4. The norm for /dev/cua2 is to use 0x3e8, but also irq4, and
I think this is probably exactly what's happening, despite my explicit
instructions to the kernel (via setserial) that /dev/cua2 should use irq
7. /proc/interrupts says that irq 4 is used by 'serial', and so is 7
when the modem is active. The curious thing is that xosview shows that
irq 4 is used whenever i move the mouse, but reflects no irq activity as
a result of the modem being active. Hmmmm. I'm going to move my mouse
to /dev/cua1, and hopefully that will make the system work, but it's
very distressing to me that the exact same hardware setup I used under
RH5.1 suddenly doesn't work under 5.2! Could linuxconf have anything to
do with this?? It really seems like the modem is trying to use irq 4,
despite both hardware and software instructions to use 7, just because
it is on /dev/cua2, where irq4 is the default! I have found no other
'setserial' commands under /etc/rc.d/ aside from my own in rc.local, and
no other references to any serial ports.
Does anyone have a solution or a reason behind this? Please help.
-DJ
--
=============================
Daniel A. Jablonski
[EMAIL PROTECTED]
=============================
==============42E199371E3ACFB3610F02FA
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit
<HTML>
Okay,
<P> I have all the classic symptoms of an irq conflict with my modem
under my installed-as-of-yesterday RH 5.2 system. It will dial, and
will connect, but is tortuously slow. This goes for ppp and minicom.
<P> I was previously running RH 5.1, upgraded from RH 4.2 (<B>still</B>
the best release), under which all of my hardware - none of which has changed
- worked great, including the modem. I initialized it in /etc/rc.d/rc.local
with setserial to use /dev/cua2 0x3e8, and IRQ 7, and I pass it the 'spd_vhi'
option as well - this is how the modem is jumpered. Now the serial
mouse is on /dev/cua0, which means 0x3f8 irq4. The norm for /dev/cua2
is to use 0x3e8, but also irq4, and I think this is probably exactly
what's happening, despite my explicit instructions to the kernel (via
setserial) that
/dev/cua2 should use irq 7. /proc/interrupts says that irq 4 is used
by 'serial', and so is 7 when the modem is active. The curious
thing is that xosview shows that irq 4 is used whenever i move the mouse,
but reflects no irq activity as a result of the modem being active.
Hmmmm. I'm going to move my mouse to /dev/cua1, and hopefully that
will make the system work, but it's very distressing to me that the <B>exact</B>
same hardware setup I used under RH5.1 suddenly doesn't work under 5.2!
Could linuxconf have anything to do with this?? It really seems like
the modem is trying to use irq 4, despite both hardware and software instructions
to use 7, just because it is on /dev/cua2, where irq4 is the default!
I have found no other 'setserial' commands under /etc/rc.d/ aside from
my own in rc.local, and no other references to any serial ports.
<P> Does anyone have a solution or a reason behind this? Please
help.
<P>-DJ
<PRE>--
=============================
Daniel A. Jablonski
[EMAIL PROTECTED]
=============================</PRE>
</HTML>
==============42E199371E3ACFB3610F02FA==
------------------------------
From: Markus Schulte <[EMAIL PROTECTED]>
Subject: any suggestions for a dual port nic?
Date: Sat, 02 Jan 1999 06:39:49 GMT
hi,
we want to set up a raid 5 fileserver with some discless webservers
attached to it.
to assure the webservers don't encounter bandwidth bottlenecs we want to
have eth0 and eth1 inside. due to the limited number of pci ports on the
motherbnoard and the requirement for as much as possible scsi
controllers and a grafix board we would like to use a dual port 100MBit
nic.
but i can't work out wich dual port nic are around and which are
supported.
any suggestions?
markus
--
Digital Online Media GmbH
(we like to be called DOM)
http://www.dom.de
mailto: [EMAIL PROTECTED]
phone: +49 221 951680
fax: +49 221 951688
------------------------------
From: Eric Jorgensen <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,linux.redhat.install,linux.redhat.misc
Subject: Re: NFSing a tape drive ?
Date: Sat, 02 Jan 1999 06:42:29 GMT
Alan Sparks wrote:
>
> NFS will not work for this. Granted, you can NFS mount the remote
> machine's /dev directory, but you're only getting copies of its special
> files... the references to the devices are interpreted by the local
> machine's kernel.
>
> You can check into autofs, which supposedly _will_ allow remote device
> access (although as yet I've simply read the man page, not actually set
> it up).
It might be nice if there was a tar filesystem driver - I mean, yes, it
would have to use the loopback device, but it would be handy for a
variety of tasks.
Although, I'm fairly certian it would be a painful thing to use with a
tape drive.
- Eric
------------------------------
From: patrick thempel <[EMAIL PROTECTED]>
Subject: Re: dos client
Date: Sat, 02 Jan 1999 06:40:18 GMT
Keith Hewitt wrote:
> Help
>
> I'm looking to implement samba networking, trouble is I have a couple of
> 286 machines which have to be included (ie not win 3.11 / w95). How do I
> get the old wfw for dos to talk to samba? where can I get a tcp/ip stack
> for dos?
>
> Thanks for any help
>
> Keith Hewitt
> WENN London
> <[EMAIL PROTECTED]>
this is just an idea ... i have NOT tried it myself !!!
there is a nt/client for dos out, enabling you to connect to a nt/domain
and (net use) access shares.
comes with nt-server. might be downloadable from microsoft-ftp.
might be usable for your purpose
------------------------------
From: "Tristan Braun" <[EMAIL PROTECTED]>
Subject: Network behind PPP
Date: Sat, 02 Jan 1999 06:40:03 GMT
I have build up a network with 1 Linux an d 1 Win98 PC. They have the IP
numbers 192.168.0.1 (= Linux = defaultGW) and 192.168.0.2 with Subnet
255.255.255.0.
The Linux PC is connected to my ISProvider (T-Online) with PPP with no
problems. He gets dynamic IP-address. This IP is the default route.
Now I want the Win98 PC to connect to the Internet (HTTP, POP, SMTP ...)
over the Linux as a gw. Here is the default route the Linux PC. When I try
to PING it is not reachable and TRACERT he find the first hop (my Linux PC)
but this is the last one.
Do I have to tell the Win98 the addresses for my ISP, too (but they are
dynamic) ?
Might it be, that my Provider does not allow requests from other IP ?
If so, must I masquerade my Win98 host (and how does it work) ?
Thanks in advance for your help.
Regards,
Tristan
------------------------------
** 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
******************************
