Linux-Networking Digest #273, Volume #11 Tue, 25 May 99 01:13:45 EDT
Contents:
Re: Help: I've Lost Root Login (OldUncleMe)
Re: Will a SupraExpress 56i modem run under linux? ("esoteric")
All I want for Christmas is a ppp internet connection (Sandwalker)
Re: telnet'ing to same system (Jeffrey Bell)
Trouble detecting Intel EtherExpress Pro 10 with RH 6.0 ("Jack Richins")
Re: Netgear FA310TX/Slackware 3.6 ("Jim Scheffler")
Re: cdrom ounting trouble!! (Jeffrey Bell)
Re: token ring driver - performance questions? (bill davidsen)
Printing Question (Arthur Merar)
Dial On Demand options for RH6 ("Tim Gibson")
Re: Trouble detecting Intel EtherExpress Pro 10 with RH 6.0 ("Penguin")
Which EEPROMs for NICs, etherboot? ("R. Christopher Harshman")
clfowd data (Nathan Chen)
Re: Networking the internet ("Andrey Smirnov")
Re: no /dev/eth0 ("Andrey Smirnov")
TCP Very Slow w/kernel 2.2.7 (Steven Ihde)
NIC setup problem ("thinh")
Re: Dumb Telnet ? ("Dmitry V. Ketov")
Re: Intel Etherexpress Pro 10+/eepro ("Dmitry V. Ketov")
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (OldUncleMe)
Crossposted-To: linux.samba
Subject: Re: Help: I've Lost Root Login
Reply-To: [EMAIL PROTECTED]
Date: Tue, 25 May 1999 03:32:10 GMT
It was: Mon, 24 May 1999 09:33:16 +0200 and with STARTLING insight,
""Michael Ansley" <[EMAIL PROTECTED]>" posted "Re: Help: I've Lost Root
Login" to "linux.samba" :
-->The initial install of smb 1.9.18 would have put everything in /usr/bin (or
-->something). The make install puts everything under /usr/local/samba.
-->Unfortunately, the previous version of smb lies first in your path. If you
-->want to upgrade properly, first make a copy of your smb.conf, and anything
-->else that may be required after the upgrade, uninstall smb using rpm, and
-->then make install the new version of smb. Then copy your smb.conf to
-->/usr/local/samba/etc or /usr/local/etc (or somewhere, check the manual)
-->
-->
-->MikeA
Thanks a lot! I will try this out. I don't know what parts of the initial
samba installation are worth saving, or need to be kept, but I will back up
the smb.conf as it shared what I wanted how it should be. The new smb.conf
has a lot more options so I can just copy over the shares after adjusting
the initial settings, probably with swat.
This still doesn't answer the passwd issue. Maybe it will allow root login
locally after these changes, but it's still a maybe, AFAIK. I tried
working with shadow, passconv, anything I could find.
At one point, I copied a passwd file with only named entries and no
passwords over the current /etc/passwd. This disallowed any login at all.
To recover, I booted using this string to lilo:
linux init=/bin/ash
then remounting root drive:
mount -n -o remount,rw /
and then mounting hdd1 rw (it has many files and progs which are links from
their installed locations, such as editors.) All this required having ash
shell installed statically linked. I got this stuff from Michael Holve's
Rescue page, unfry_content.html from somewhere on eunuchs.org
tenox @ home dat com
/ts
------------------------------
From: "esoteric" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.hardware,comp.os.linux.help,comp.os.linux.misc,comp.os.linux.questions,comp.os.linux.setup
Subject: Re: Will a SupraExpress 56i modem run under linux?
Date: Mon, 24 May 1999 16:37:22 -0700
This modem will work fine. It is a true hardware based modem using the
RC56KDPF chipset. The only "winmodem" diamond makes is the SupraMax (stay
away from that piece of #$%#$^)
a good init for this modem is AT&F2 (set's autoreliable mode) to set v.34
just put +ms=11,1,300,33600 (or desired max speed)
Ozzy <[EMAIL PROTECTED]> wrote in message
news:2A523.14487$[EMAIL PROTECTED]...
> X-Priority: 3
> X-MSMail-Priority: Normal
> X-Newsreader: Microsoft Outlook Express 5.00.2314.1300
> X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300
>
> Anybody running this modem under linux?
> successfully?
>
>
------------------------------
From: Sandwalker <[EMAIL PROTECTED]>
Subject: All I want for Christmas is a ppp internet connection
Date: Mon, 24 May 1999 21:39:45 -0500
Actually, I want it now.
I am having major troubles trying to get a simple internet connection
(rehat 5.2 on P120 with 16MB).
1. modem seems fine, I can initialize (I do have an acceptable modem
string) and dial out with xminicom but nothing happens after my username
and pasword are entered. (ifconfig and netstat reveal no ppp0
connection)
2. my ISP has what seems like very explicit instructions to set up a
linux connection (with PAP secruity) please see:
http://www.chorus.net/support/linuxhelp.html
to view what they say
3. i changed the ppp-on script info my ISP provides in the following
manner:
exec pppd /dev/modem 115200 debug modem crtscts defaultroute connect
$DIALER_SCRIPT
they have listed a +ua option and the pap-secrets file location as
an option. I don't know what they are for, but they are not in the pppd
man page. They caused errors in the execution, so I took em out.
4. the info listed in the /var/log/messages is as follows (the modem
does not even dial out):
Jan 2 03:50:47 sietch kernel: CSLIP: code copyright 1989 Regents of
the University of California
Jan 2 03:50:47 sietch kernel: PPP: version 2.2.0 (dynamic channel
allocation)
Jan 2 03:50:47 sietch kernel: PPP Dynamic channel allocation code
copyright 1995 Caldera, Inc.
Jan 2 03:50:47 sietch kernel: PPP line discipline registered.
Jan 2 03:50:47 sietch kernel: registered device ppp0
Jan 2 03:50:48 sietch pppd[1380]: pppd 2.3.5 started by root, uid 0
Jan 2 03:50:49 sietch pppd[1380]: Connect script failed
Jan 2 03:50:50 sietch pppd[1380]: Exit.
Jan 2 03:52:46 sietch kernel: PPP: ppp line discipline successfully
unregistered
It appears the connect script fails (the ppp-on-dialer chat script ?),
but I have no idea what is wrong, if that is actually what is wrong. I
also am not sure where the debug pppd option is storing any errors or
how to set a better error log (or whatever).
I have tried using redhat's netcfg tool to set up the connection, but it
seems the expect-send scripting info still doesn't work (it extracts it
from the ppp-on-dialer script).
If someone could please help me or reccommend a better way to get
online,
I would deeply appreciate it. I am extremely frustrated.
Sandwalker
------------------------------
From: Jeffrey Bell <[EMAIL PROTECTED]>
Subject: Re: telnet'ing to same system
Date: Tue, 25 May 1999 04:42:37 +0000
Reply-To: [EMAIL PROTECTED]
Did you remember to update your /etc//hosts.allow file?
You need to add the ip address of your network.
ie /etc/hosts.allow
ALL: 127.0.0.1
ALL: 192.168.1.
that should do it for you.
David Hinz wrote:
> On a RH 5.1 system, I'm having problems doing a telnet back to the same
> system I'm running on.
>
> If I start an xterm, then do a "telnet kenwig" it connects, then says
> "connection closed by remote system" (or something like that). This
> fails if I use the IP address also.
>
> What am I missing?
>
> david.
--
Jeffrey A. Bell
-------------------------------------------------------------------------
Research is what I'm doing when I don't know what I'm doing.
-- Wernher von Braun --
------------------------------
From: "Jack Richins" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup,comp.os.linux.help
Subject: Trouble detecting Intel EtherExpress Pro 10 with RH 6.0
Date: Mon, 24 May 1999 16:07:41 -0700
Two questions. I'm installing RH 6.0 and I'm having trouble getting my
Intell EtherExpress
Pro 10 detected. Autodetection hangs. I copied IRQ and I/O port information
from Windows, which detects it fine. However, the RH manual shows a MEM
parameter that I should also pass the module. I can't find anything about a
memory address associated with my network card under Windows, so I've been
leaving it blank. This is what I've tried passing the module:
eepro=0x210,3
eepro=0x210,03
eepro=0x210,3,0
Any suggestions or guesses as to getting this card detected?
Second question is, could I just skip network setup and getting Linux up and
running and then go back and add the networking easily? Or will it be such a
pain I should just get the networking going during installation?
------------------------------
From: "Jim Scheffler" <[EMAIL PROTECTED]>
Subject: Re: Netgear FA310TX/Slackware 3.6
Date: Mon, 24 May 1999 14:48:22 -0400
Try getting the newest drivers
David Burlage <[EMAIL PROTECTED]> wrote in message
news:7ic64a$hbt$[EMAIL PROTECTED]...
>
> I had an ISA ne2000 clone working on my Slackware 3.6 (2.0.35)
> box. I want to now use the Netgear FA310TX PCI instead. My
> pre-purchase dejanews serahces implied this card would work.
>
> I got the module that came on the supplied floppy to compile and load.
> The card appears to be detected correcty. When I do a ifconfig,
> the eth0 shows up. But I can't ping anything except my own
> machine. I have not changed my IP address, etmask or anything.
> Can anyone offer clues?
------------------------------
From: Jeffrey Bell <[EMAIL PROTECTED]>
Subject: Re: cdrom ounting trouble!!
Date: Mon, 24 May 1999 04:41:14 +0000
Reply-To: [EMAIL PROTECTED]
Your're not trying to mount an audio cdrom as a regulare cdrom (computer data)
are you?
DeFCoN wrote:
> ok here's the deal...
>
> i can play cd's in X without even mounting the cdrom. i just put a cd in and
> and hit play and it works fine. but here's the problem. the first time i
> mounted my cdrom using, "mount -t iso9660 /dev/hdc /mnt/cdrom", it worked
> great! and then when i was done and wanted to put i different cd in i just
> typed, "umount /dev/cdrom" or i guess "umount /dev/hdc" would work too. but
> anyway, after i unmounted that cd i couldn't mount another one! i've try'd
> mounting it every way i've read to, with no luck. so i just gave up and
> completely reinstalled linux.
>
> BTW: i'm using redhat 6.0 if it matters... and i don't know exactly what
> kernel no. at the moment. (sorry)
>
> so anyway, i reinstalled linux and sure enough... it mounted and unmounted
> fine but after that i couldn't get it to mount anything!! not even after
> rebooting!! ANY IDEAS?? i'd greatly appreciate ANY help!! i'm new to this
> and i'm trying to lrean this crap the hard way. THANX
>
> ---> DeFCoN <---
--
Jeffrey A. Bell
-------------------------------------------------------------------------
Research is what I'm doing when I don't know what I'm doing.
-- Wernher von Braun --
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Subject: Re: token ring driver - performance questions?
Date: 25 May 1999 00:03:37 GMT
In article <7ib4r6$krk$[EMAIL PROTECTED]>,
<[EMAIL PROTECTED]> wrote:
| It doesn't appear that the standard token ring driver is sharable, so I
| can't do any load balancing using more than one card which is a pity.
It's been forever since I did anything fancy with TR on a 2.0 machine,
but I thought late versions of the driver worked for multiple cards.
Early versions had warnings that it wouldn't work, though.
| Has anyone used this driver in anger though (the standard token ring one
| that ships with all kernels;
The TR driver changed a LOT in the 2.1 development process, so don't
assume old behaviour and new are the same. That said I have seen some
machines which talked very slowly to others on TR. I would start by
using ftp on a few MB file to see what the transfer rate really is. I
have one "picky" machine which will do 100kB to some machines and 2-3kB
to others. No clue at all why.
| I'm concerned about
| the performance of the driver itself. It does seem robust enough, but
| Squid seems suprisingly slow even given the less-than-ideal hardware
| I'm running this on (P166 PC, 2 x 2GB IDE disks with web cache on its
| own disk, 64MB memory) and so I'm leaning towards the token ring driver
| being the bottleneck at the moment...
No, but as I said, some machine just don't work well. My most used TR
machine will do ~200kB transfers, as opposed to (see above). The machine
with better performance is an overloaded P90, the slow machine is P200.
|
| Any comments welcome, please copy via email though... :)
Try ftp from several machines, get data, post back.
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
One common problem is mistyping an email address and creating another
valid, though unintended, recipient. Always check the recipient's
address carefully when sending personal information, such as credit
card numbers, death threats or offers of sexual services.
------------------------------
From: [EMAIL PROTECTED] (Arthur Merar)
Subject: Printing Question
Date: Tue, 25 May 1999 04:13:38 GMT
Hello,
I am fairly new to Linux. I just installed Red Hat 5.2. My network
consists of a Windows 98 box, my Linux box and an HP LJ 4M+ network
printer.
Everything is talking fine except the Linux box and the printer. The
printer does not seem to have an IP address. Rather it has a network
name and hardware address......since I think Windows 98 uses the
Client for Netware Networks and stuff.
Anyhow, I'm not sure how to install the printer so I can print to it
from either Windows or Linux. Can someone help me?
Thanks,
Please send response to e-mail.
Arthur
[EMAIL PROTECTED]
------------------------------
From: "Tim Gibson" <[EMAIL PROTECTED]>
Subject: Dial On Demand options for RH6
Date: Mon, 24 May 1999 09:41:30 -0700
What is the best solution for a dial-on-demand problem? I know about diald
but also know that I have tried it before and could never get it working. I
heard about the recent ppp version having dial-on-demand cpabilities but no
one seems to be using them. Help. I have a client that is currently loggin
into my server as root in order to bring the ISP connection when it goes
down.
------------------------------
From: "Penguin" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup,comp.os.linux.help
Subject: Re: Trouble detecting Intel EtherExpress Pro 10 with RH 6.0
Date: Mon, 24 May 1999 19:39:44 -0400
I'm a newbie to Linux but i think i can answer the second
question.......when you finally get linux setup with xwindows and all that
you can go to your konsole in gnome or KDE and type either "netconf" or
"linuxconf" all that you will need to set up your network is right
there,along with a quick and easy to use help section and it is fairly
straight forward as long as you know all the information that needs to go in
there which is where I'm stuck.........that is also where you can set up ftp
options and so on.....its endless and in a nice windowed format well thats
as far as my brain will let me go hope it helps out some
cheers
Jack Richins <[EMAIL PROTECTED]> wrote in message
news:7icm30$[EMAIL PROTECTED]...
> Two questions. I'm installing RH 6.0 and I'm having trouble getting my
> Intell EtherExpress
> Pro 10 detected. Autodetection hangs. I copied IRQ and I/O port
information
> from Windows, which detects it fine. However, the RH manual shows a MEM
> parameter that I should also pass the module. I can't find anything about
a
> memory address associated with my network card under Windows, so I've been
> leaving it blank. This is what I've tried passing the module:
>
> eepro=0x210,3
> eepro=0x210,03
> eepro=0x210,3,0
>
> Any suggestions or guesses as to getting this card detected?
>
> Second question is, could I just skip network setup and getting Linux up
and
> running and then go back and add the networking easily? Or will it be such
a
> pain I should just get the networking going during installation?
>
>
>
>
------------------------------
From: "R. Christopher Harshman" <[EMAIL PROTECTED]>
Subject: Which EEPROMs for NICs, etherboot?
Date: Mon, 24 May 1999 21:18:09 -0700
I'm needing to get some network cards (PCI and ISA NE2000 clones,
SMC Ultra16s) setup for network booting using Linux and Etherboot.
I've been to the various pages and read the introductory material,
but I'm still lost as to one critical aspect: which EEPROMs do I need,
and which burner will work with them? I need to order this stuff,
from JDR Microdevices or whomever else, and I have no idea what
to ask for.
Anyone? Thanks much!
- Chris
------------------------------
From: Nathan Chen <[EMAIL PROTECTED]>
Crossposted-To: comp.dcom.sys.cisco
Subject: clfowd data
Date: Mon, 24 May 1999 22:19:09 +0800
Hi,
I've installed cflowd 2.0.0b9 on my Linux(RedHat) server and receiving
NetFlow data from my Cisco router(7206).Everthing work fine and I can
use those arts++ utilities(e.g. artsases) to dump cfdcollect
information.But I hope can generate an ascii text file from my everyday
cfdcollect data for NT application to read it so that I can generate
report for my customer.Are there any tools or your valuable script
program can help me ?
Thanks in advacne for any help !
=============Nathan Chen
------------------------------
From: "Andrey Smirnov" <[EMAIL PROTECTED]>
Subject: Re: Networking the internet
Date: Mon, 24 May 1999 14:52:34 -0700
Hello!
You need to get familiar with IP Masquerading and PPP. Here are links to the
HOWTOs:
1) IP MASQUERADING: http://metalab.unc.edu/LDP/HOWTO/mini/IP-Masquerade.html
2) PPP: http://metalab.unc.edu/LDP/HOWTO/PPP-HOWTO.html
Good luck!
v4cal <[EMAIL PROTECTED]> wrote in message news:7ibt75$j45$[EMAIL PROTECTED]...
> I am trying to network the net through all my computer that are pluged
into
> linux
> the other computers are useing windows
>
> i would like a step by step help on haw to do this
>
> I want Linux to be the Internet server in my local network and i also want
> to dail into linux and also get access to the net and my files
>
> haw do i do this
>
> Norbert
>
>
------------------------------
From: "Andrey Smirnov" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.portable
Subject: Re: no /dev/eth0
Date: Mon, 24 May 1999 15:18:07 -0700
Hello!
There should not be /dev/eth0!
If your card is detected by boot process you should be able to use network
configuration files to set IP address and the rest of network info.
I'm not sure what is directory structure of SUSE, but in RH it's in
/etc/sysconfig/network-scripts and /etc/sysconfig (file named network).
Check out NET-3-HOWTO: http://metalab.unc.edu/LDP/HOWTO/NET-3-HOWTO.html
Good luck!
Neil <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> hi
>
> I've installed SUSE 6.1 on my dell lattitude cpi 300.
>
> Installed the PCMCIA package etc.
>
> Ethernet card detected at boot. But the initial installtion didn't create
> /dev/eth0 so I can't connect to my network.
>
> What's happened and any ideas how I can fix it pls. I've checked the
HOWTOs etc.
>
> attached is last dmesg
>
> TIA
>
> Neil
>
> -----------------------------
>
> Linux version 2.2.5 ([EMAIL PROTECTED]) (gcc version 2.7.2.3) #1 Tue
Apr
> 13 16:33:46 MEST 1999
> Detected 297791293 Hz processor.
> Console: colour VGA+ 80x25
> Calibrating delay loop... 297.37 BogoMIPS
> Memory: 127872k/131008k available (1156k kernel code, 408k reserved, 1532k
data,
> 40k init)
> VFS: Diskquotas version dquot_6.4.0 initialized
> CPU: Intel Celeron (Mendocino) stepping 0a
> Checking 386/387 coupling... OK, FPU using exception 16 error reporting.
> Checking 'hlt' instruction... OK.
> Checking for popad bug... OK.
> POSIX conformance testing by UNIFIX
> mtrr: v1.26 (19981001) Richard Gooch ([EMAIL PROTECTED])
> PCI: PCI BIOS revision 2.10 entry at 0xfc06e
> PCI: Using configuration type 1
> PCI: Probing PCI hardware
> Linux NET4.0 for Linux 2.2
> Based upon Swansea University Computer Society NET3.039
> NET4: Unix domain sockets 1.0 for Linux NET4.0.
> NET4: Linux TCP/IP 1.0 for NET4.0
> IP Protocols: ICMP, UDP, TCP, IGMP
> Initializing RT netlink socket
> Starting kswapd v 1.5
> Detected PS/2 Mouse Port.
> pty: 256 Unix98 ptys configured
> Real Time Clock Driver v1.09
> RAM disk driver initialized: 16 RAM disks of 20480K size
> PIIX4: IDE controller on PCI bus 00 dev 39
> PIIX4: not 100% native mode: will probe irqs later
> ide0: BM-DMA at 0x0860-0x0867, BIOS settings: hda:DMA, hdb:pio
> ide1: BM-DMA at 0x0868-0x086f, BIOS settings: hdc:pio, hdd:pio
> hda: FUJITSU MHE2064AT, ATA DISK drive
> hdc: SAMSUNG CD-ROM SCR-2438, ATAPI CDROM drive
> ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
> ide1 at 0x170-0x177,0x376 on irq 15
> hda: FUJITSU MHE2064AT, 6194MB w/0kB Cache, CHS=789/255/63
> hdc: ATAPI 24X CD-ROM drive, 128kB Cache
> Uniform CDROM driver Revision: 2.54
> Floppy drive(s): fd0 is 1.44M
> FDC 0 is a post-1991 82077
> md driver 0.36.6 MAX_MD_DEV=4, MAX_REAL=8
> linear personality registered
> raid0 personality registered
> raid1 personality registered
> scsi : 0 hosts.
> scsi : detected total.
> Partition check:
> hda: hda1 hda2 < hda5 hda6 >
> VFS: Mounted root (ext2 filesystem) readonly.
> Freeing unused kernel memory: 40k freed
> Adding Swap: 393556k swap-space (priority -1)
> Serial driver version 4.27 with HUB-6 MANY_PORTS MULTIPORT SHARE_IRQ
enabled
> ttyS00 at 0x03f8 (irq = 4) is a 16550A
> ttyS02 at 0x03e8 (irq = 4) is a 16550A
> Linux PCMCIA Card Services 3.0.9
> kernel build: 2.2.5 unknown
> options: [pci] [cardbus]
> Intel PCIC probe:
> Unknown [0x104c 0xac1c] PCI-to-CardBus at bus 0 slot 3, mem 0x68000000,
2
> sockets
> host opts [0]: [no pci irq] [lat 32/176] [bus 32/34]
> host opts [1]: [no pci irq] [lat 32/176] [bus 35/37]
> ISA irqs (scanned) = 3,4,5,7,9,10 status change on irq 10
> cs: IO port probe 0x1000-0x17ff: clean.
> cs: IO port probe 0x0100-0x04ff: excluding 0x290-0x297 0x378-0x37f
0x4d0-0x4d7
> cs: IO port probe 0x0a00-0x0aff: clean.
> 3c59x.c:v0.99H 11/17/98 Donald Becker
> http://cesdis.gsfc.nasa.gov/linux/drivers/vortex.html
> cs: cb_config(bus 32): vendor 0x10b7, device 0x5157
> fn 0 bar 1: io 0x200-0x27f
> fn 0 bar 2: mem 0xa0021000-0xa002107f
> fn 0 bar 3: mem 0xa0020000-0xa002007f
> fn 0 rom: mem 0xa0000000-0xa001ffff
> cs: cb_enable(bus 32)
> bridge io map 0 (flags 0x21): 0x200-0x27f
> bridge mem map 0 (flags 0x1): 0xa0000000-0xa0021fff
> vortex_attach(bus 32, function 0, device 5157)
> eth0: 3Com 3CCFE575 Cyclone CardBus at 0x200, 00:50:04:5b:dd:0e, IRQ 3
> eth0: CardBus functions mapped a0020000->c8067000
> 8K byte-wide RAM 5:3 Rx:Tx split, MII interface.
> MII transceiver found at address 0, status 7809.
> Enabling bus-master transmits and whole-frame receives.
> cs: memory probe 0xa0000000-0xa0ffffff: excluding 0xa0000000-0xa00fffff
> tty01 at 0x02f8 (irq = 5) is a 16550A
> cs: IO port probe 0x1000-0x17ff: clean.
> cs: IO port probe 0x0100-0x04ff: excluding 0x290-0x297 0x378-0x37f
0x4d0-0x4d7
> cs: IO port probe 0x0a00-0x0aff: clean.
>
> ----------------------------------
> Neil
> ICQ 11875525 - AOL IM "imheifetz"
------------------------------
From: Steven Ihde <[EMAIL PROTECTED]>
Subject: TCP Very Slow w/kernel 2.2.7
Date: Mon, 24 May 1999 20:53:38 -0700
Hi all,
I'm running kernel 2.2.7 and have been seeing a very strange
phenomonon. I'm connected to IBM's intranet via an ISDN line. This
should give me about 8kB/s, or 16kB/s if it brings up both B-channels.
When I ftp to a number of different systems inside, I get just about
exactly that. But on certain systems, I get terrible performance,
around 1.5kB/s. I was wondering what could be wrong, so I did a little
tracing with tcpdump. I found that my box doesn't appear to be sending
TCP ACKs correctly.
Here's a short sample from a working (fast) ftp:
15:17:36.763571 EXPLORE3.ALMADEN.IBM.COM.3453 > ihdehome..3649: .
2049:2561(512) ack 1 win 16384
15:17:36.771015 EXPLORE3.ALMADEN.IBM.COM.3453 > ihdehome..3649: .
2561:3073(512) ack 1 win 16384
15:17:36.771085 ihdehome..3649 > EXPLORE3.ALMADEN.IBM.COM.3453: . ack
3073 win 32256 (DF) [tos 0x8]
15:17:36.837844 EXPLORE3.ALMADEN.IBM.COM.3453 > ihdehome..3649: .
3073:3585(512) ack 1 win 16384
15:17:36.845451 EXPLORE3.ALMADEN.IBM.COM.3453 > ihdehome..3649: .
3585:4097(512) ack 1 win 16384
15:17:36.845515 ihdehome..3649 > EXPLORE3.ALMADEN.IBM.COM.3453: . ack
4097 win 32256 (DF) [tos 0x8]
15:17:36.912280 EXPLORE3.ALMADEN.IBM.COM.3453 > ihdehome..3649: .
4097:4609(512) ack 1 win 16384
15:17:36.919784 EXPLORE3.ALMADEN.IBM.COM.3453 > ihdehome..3649: .
4609:5121(512) ack 1 win 16384
15:17:36.919848 ihdehome..3649 > EXPLORE3.ALMADEN.IBM.COM.3453: . ack
5121 win 32256 (DF) [tos 0x8]
Note especially the third, sixth, and ninth lines, where my box
(ihdehome) ACKs everything up to and including the last byte received.
Now here's a sample from a slow (apparently broken) ftp:
23:11:04.256561 socks2.almaden.ibm.com.1080 > ihdehome..3005: .
18824:20272(1448) ack 1 win 15928 <nop,nop,timestamp 931434408
133834089>
23:11:04.440755 socks2.almaden.ibm.com.1080 > ihdehome..3005: .
20272:21720(1448) ack 1 win 15928 <nop,nop,timestamp 931434189
133822936>
23:11:04.755789 ihdehome..3005 > socks2.almaden.ibm.com.1080: . ack
20272 win 31856 <nop,nop,timestamp 133834163 931434408> (DF) [tos 0x8]
23:11:04.986819 socks2.almaden.ibm.com.1080 > ihdehome..3005: .
21720:23168(1448) ack 1 win 15928 <nop,nop,timestamp 931434409
133834163>
23:11:04.986910 ihdehome..3005 > socks2.almaden.ibm.com.1080: . ack
20272 win 31856 <nop,nop,timestamp 133834186 931434409> (DF) [tos 0x8]
23:11:05.171812 socks2.almaden.ibm.com.1080 > ihdehome..3005: .
23168:24616(1448) ack 1 win 15928 <nop,nop,timestamp 931434189
133822936>
23:11:06.578067 socks2.almaden.ibm.com.1080 > ihdehome..3005: .
20272:21720(1448) ack 1 win 15928 <nop,nop,timestamp 931434412
133834186>
23:11:06.578184 ihdehome..3005 > socks2.almaden.ibm.com.1080: . ack
23168 win 30408 <nop,nop,timestamp 133834345 931434412> (DF) [tos 0x8]
You can see the ACKs don't bear too much resemblance to the last byte
transmitted, though the delay (in time) before the ACK is sent is
substantial, over one quarter second. The result is that many packets
wind up being retransmitted, and throughput is cut to about one-sixth of
what I observe on "working" connections. This appears to be the fault
of my box, though I don't know why.
The machines that I've had slow connections with are running AIX 4.3,
or, (surprisingly) Linux 2.2.7. I've gotten good performance to Windows
NT 4.0 SP3, and AIX 4.2. I haven't tested enough machines to say if
there is really a good correlation between the remote OS and the speed
of the connection.
Does anyone have any ideas what could be going on? I can mail more
complete logs if anyone wants them.
Thanks,
Steve Ihde
------------------------------
From: "thinh" <[EMAIL PROTECTED]>
Subject: NIC setup problem
Date: Mon, 24 May 1999 22:56:35 -0500
I am getting some weird errors when try to configure my network card. I am
running Linux 5.2 with 2.0.36 kernel. I get the following errors when I try
an "ifconfig eth0 <ip addr> netmask <mask> up"
SIOCSIFADDR: Operation not supported by device
eth0: unknown interface.
SIOCSIFNETMASK: Operation not supported by device
eth0: unknown interface.
However when I go into netcfg, I can enter the information I want. But the
interface will not be active when I try a "ifconfig -a". I only get the
loopback interface.
Any help provided will be appreciated. Thanks in advances.
Thinh
[EMAIL PROTECTED]
------------------------------
From: "Dmitry V. Ketov" <[EMAIL PROTECTED]>
Subject: Re: Dumb Telnet ?
Date: Mon, 24 May 1999 15:54:18 +0400
Hi
c:\>telnet x.x.x.x
DB7654321 ����� � ��������� <[EMAIL PROTECTED]>
...
>How do I telnet from my windoze computers to my linux server by entering an
IP
>address?
>
>David Bell
>
>Please don't email me just reply on the board.
------------------------------
From: "Dmitry V. Ketov" <[EMAIL PROTECTED]>
Subject: Re: Intel Etherexpress Pro 10+/eepro
Date: Mon, 24 May 1999 16:06:39 +0400
Hi!!!
>I also tried adding an append ="ether=11,0x230,0,0,eth0" to lilo.conf
>and ran lilo.
check order of IRQ,IO in "ether=" options
>I also tried recompiling the kernel w/ eepro compiled in (not as a
>module).
>
>The IRQ and the IO port are free.
>ifconfig doesn't show eth0 and doing an
>ifconfig -add for the card locks up the machine.
If you have a module for a card, try /etc/conf.modules:
==================================================
alias eth0 eepro
options ne io=0x230 irq=11
==================================================
then
ifconfig eth0 <ip> netmask <mask> ........
------------------------------
** 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
******************************
