Linux-Networking Digest #887, Volume #9 Fri, 15 Jan 99 10:15:14 EST
Contents:
Re: RJ-45 network needs hub ? (Stuart R. Fuller)
PLIP link SIOCADDRT: device or resource busy (Giovanni Russo)
Re: su won't let me shutdown/ifconfig ("Vikram.V.Asrani")
Re: IP Aliasing problem (Brian McCauley)
Need some ftp stat software (Denis Nuja)
tcp/ip & mars_nwe (long) ("J_Nick")
Re: New PPP Program (Oleg Letsinsky)
Re: Dialin from SCO OS5 fails (Ronald Duncan)
Re: How to send IP packet from a selected interface ? (Brian McCauley)
Re: Email Forwarding (Ronald Duncan)
PPPD (Roger Wernersson)
DHCP Lease Times ("Jamie Scott")
FTP for users ("Phil Massyn")
NFS linux server and an AIX client ("[EMAIL PROTECTED]")
Re: FTP for users (Raymond Doetjes)
Re: Setting up imapd: how to check? (Bruno Wolff III)
Need TCP/IP routing guru assistance (Sim)
Re: Remote backups as root? (Raymond Doetjes)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Stuart R. Fuller)
Subject: Re: RJ-45 network needs hub ?
Reply-To: [EMAIL PROTECTED]
Date: Thu, 14 Jan 1999 17:00:02 GMT
Drew Shelley ([EMAIL PROTECTED]) wrote:
: Lorenzo Delloni wrote:
: >
: > I reported before a problem of connection between my labtop and my Linux
: > PC. I did not mention that the Etherlink III 3 com card was provided
: > with a cable allowing direct connection of the computer to a RJ-45
: > socket and that I plug this cable to the RJ-45 out of my NE2000
: > compatible card. I have asked for advices in some of the tones of PC
: > shops close to Tottenham in London. If I exclude the stupids remarks
: > concerning the fact that I am not able to install the driver and they
: > can do it for me for �45 !!! or that it's not possible Linux with Win95
: > ! The most raisonable remark was that for a RJ-45 network, even between
: > two computers, you definitely need a hub. Is that true ? I am a bit
: > sceptic of what these computers vendors say !
: >
: > Thanks
:
: No, you do not need a hub. However, an ordinary patch cable will not do.
: You must use a cross-over cable. They are easily purchased, or, if you
: have access to the tools, you can make your own. Do a search for
: cross-over cable. When I did that, I found one site that actually had a
: picture showing the correct wiring method.
On the other hand, using a hub, even for 2 computers has its advantages:
- there are usually lotsa pretty lights to look at for troubleshooting
purposes.
- hubs aren't really that expensive.
- if you go to a 3rd computer, then you throw away the special
cross-over cable and end up buying a hub anyway.
- store-bought cross-over cables are more expensive than standard
cables (due to sales volume).
- making your own cross-over cable is expensive, since you have to buy
the cable, the connectors and crimper.
Stu
------------------------------
From: Giovanni Russo <[EMAIL PROTECTED]>
Subject: PLIP link SIOCADDRT: device or resource busy
Date: Thu, 14 Jan 1999 19:40:15 +0100
Hi
I tried to set up a PLIP link between two Linux computers but when
loading plip module
insmod plip io=0x378 irq=5
(adress and irq confirmed)
the message
SIOCADDRT: device or resource busy
appears :(
However no lp module is loaded or built into the kernel.
Where could be the problem? Please mail to SENDER as well !!!
tanks
giovanni :)
------------------------------
Crossposted-To: comp.os.linux,linux.redhat.misc
From: "Vikram.V.Asrani" <[EMAIL PROTECTED]>
Subject: Re: su won't let me shutdown/ifconfig
Reply-To: [EMAIL PROTECTED]
Date: Fri, 15 Jan 1999 08:53:13 GMT
In comp.os.linux YoYo <[EMAIL PROTECTED]> wrote:
>>When you 'su' you still have the PATH that you had in your user account.
>>If you're trying to run a command that lives in a directory that's not in
>>your path, you won't be able to find it. Try this: 'su' to root, and do
>>"locate command-I-want-to-find". That should give you the full pathname of
>>the command you're trying to invoke (e.g. "/sbin/ifconfig"). Then try
>>invoking it by its full pathname.
su does not change the environment. In order to do that use
su -. Then you need not worry about the settings. The default ones
will be used.
--
Vikram.V.Asrani.
([EMAIL PROTECTED])
------------------------------
From: Brian McCauley <[EMAIL PROTECTED]>
Subject: Re: IP Aliasing problem
Date: 14 Jan 1999 18:42:20 +0000
"Alex Tsekhansky" <[EMAIL PROTECTED]> writes:
> I have a problem configuring aliasing on RedHat 5.1.
>
> I compiled kernel with aliasing included and when I try to use something
> like
>
> ifconfig eth0:0 192.168.0.1
>
> I get an error message
>
> SIOCSIFADDR: Invalid argument
>
> Also if I try to load ip_alias.o with insmod, I get
>
> ip_alias.o: unresolved symbol register_net_alias_type_R9eeab982
> ip_alias.o: unresolved symbol ip_rt_dev_Rde10f949
> ip_alias.o: unresolved symbol unregister_net_alias_type_R575fb95a
>
> I would appreciate any thoughts and/or comments.
Looks like the version of ip_alais you are using is not from the same
kernel build as your kernel.
Try:
grep register_net_alias_type </proc/ksyms
Do you see anything?
If you see nothing the you are running a kernel wihout network alias
support. You can't load the IP alias support without the network
alias support loaded.
--
\\ ( ) No male bovine | Email: [EMAIL PROTECTED]
. _\\__[oo faeces from | Phones: +44 121 471 3789 (home)
.__/ \\ /\@ /~) /~[ /\/[ | +44 121 627 2173 (voice) 2175 (fax)
. l___\\ /~~) /~~[ / [ | PGP-fp: D7 03 2A 4B D8 3A 05 37...
# ll l\\ ~~~~ ~ ~ ~ ~ | http://www.wcl.bham.ac.uk/~bam/
###LL LL\\ (Brian McCauley) |
------------------------------
From: Denis Nuja <[EMAIL PROTECTED]>
Subject: Need some ftp stat software
Date: 15 Jan 1999 12:14:12 GMT
I need some software for making statistics of our ftp site ?
any ideas
--
And on the Eight Day God said, "Ok Murphy, its all yours".
------------------------------
From: "J_Nick" <[EMAIL PROTECTED]>
Subject: tcp/ip & mars_nwe (long)
Date: Thu, 14 Jan 1999 14:03:37 -0500
As a recent convert to Linux I have some newbie questions I hope someone
can help with... I installed SuSE 5.3 on our network here at work.
Everything was fine, I set up TCP/IP, and I could reach Apache from Win95 IE
browsers on the network and see the web pages I put there. Telnet was also
OK. (tested from four test machines throughout the plant). I also had all
four people sending and receiving email with Exchange, Outlook & MSMail
clients with no problems.
Then I installed mars_nwe. I can see my Linux box as a Netware server on
*some* machines but not all<?> I can login to the mars_nwe server from Win95
Network Neighborhood, and see both as Netware servers with slist. (On the
boxes that have the icon in Network Neighborhood) Also from DOS on these
(true DOS, not a DOS box)But now I can't ping or telnet to the Linux box! Is
this an either/or proposition?
We have an all coax system with approx 80 machines, mostly for email.
One Netware 3.12 server, with an ethernet switch that splits the network
into approx. six lines to different areas of the building. I can see the
mars_nwe server icon from some machines on the other side of the server and
switch, but not all. And I can ping all four test machines *except* the
Linux box now. I can also ping the Linux box from itself. All services
appear to be starting OK.
Here's what I was attempting to do... maybe my logic is flawed? Our
email vendor told us they weren't Y2K and they are going to discontinue the
Noteworks product. I have approx. 18 diskless workstations I had hoped to
hook up to the mars_nwe server to retrieve email using the current DOS
Netware clients, using Pegasus for DOS. Upgrading these to Win95 would be
nice, but unlikely for financial reasons. Same probably goes for a new email
program (unless we have no choice!) I can live without the TCP/IP if I could
get the mars_nwe working. But I'd like to use both for future intranet
plans.
If anyone has any ideas, I'd love to hear them. Would love to have an
alternative to MS. Thanks!!
Joe
------------------------------
From: [EMAIL PROTECTED] (Oleg Letsinsky)
Crossposted-To:
alt.2600,comp.os.linux.setup,comp.os.linux.development.apps,comp.os.linux,alt.os.linux
Subject: Re: New PPP Program
Date: Fri, 15 Jan 1999 12:35:43 GMT
On Thu, 14 Jan 1999 13:47:44 +0000, Stuart Harris <[EMAIL PROTECTED]>
wrote:
>I'm about to start developing a ppp program (shell/perl)
>which will at first only work with REDNET,
>The ppp program (pppconnect) will hopefully grow very big!
>the general idea is that its more a terminal based control
>panel rather than 'enter this then this and ooh what about this'
>
>Im currently looking for help todo this
>
>I need help in the following ways
> 1, Someone who is better than me at shell scripting! (good with loops
>for example)
> 2, People to submit modem init strings for different types of modems
> 3, People to harras modem manufacturers to send me/us a complete of
>init
> strings for all their modems
[skip]
Most modem vendors provide INF-files with their devices. They are
primary targetted for Windows, but actually they are just sets of init
strings. You can look in those files for init strings, they are simple
text files in Windows INI-file format. Or even better, you can add an
option "Load settings from INF-file" in your program (I'm wondering
why nobody implemented this yet - AFAIK).
--
This .sig is $10 shareware. Register now to get its full 10MB version!
------------------------------
From: [EMAIL PROTECTED] (Ronald Duncan)
Subject: Re: Dialin from SCO OS5 fails
Date: Fri, 15 Jan 1999 11:36:27 +0000
Hi Steven,
I *think* what is happening is that the successful connections with your
linux server are using pap/chap authentication which, your current
installation of mgetty recognises. However, when your sco system connects
it is using the older "unix" method of login/passwd/fire-pppd.
I am not sure if SCOopenserver can do pap/chap, but certainly mgetty can
be set up to work with either (if so complied), check the mgetty
documentation.
--
Ronald Duncan | http://www.duncans.co.uk
John Duncan & Son (est. 1824) | Email: [EMAIL PROTECTED]
On Mon, 11 Jan 1999 17:51:05 +1030, Stephen Davies <[EMAIL PROTECTED]> wrote:
>I have three SCO Open Server 5 systems which I want to configure to open PPP
>connections to my Linux 2.0.36 server.
>
>However, they all fail to login successfully using the SCO version of
>chat for UUCP/PPP.
>
>I am running mgetty 1.0.0 with the gettydefs option turned off (and
>cannot find any documentation as to just how it handles the line without
>gettydefs) and can successfully dialin and open PPP connections from
>Macs, W9x and other Linux boxes. Only the SCO boxes fail.
>
>What happens is that the login prompt is sent to the SCO box which
>recognises it and starts to reply with the username but at random points
>during that input, all communications cease and mgetty times out and
>hangs up.
>
>I guess that there is some incompatability between the way that SCO
>drives it's modem and the way that Linux/mgetty expects it be be driven
>but have no ideas left as to what it might be.
>
>Any suggestions gratefully received.
>
>Cheers and thanks,
>Stephen
>--
>========================================================================
>Stephen Davies Consulting [EMAIL PROTECTED]
>Adelaide, South Australia. Voice: 041-980 1417
>Computing & Network solutions. Fax : 08-8272 8863
------------------------------
From: Brian McCauley <[EMAIL PROTECTED]>
Subject: Re: How to send IP packet from a selected interface ?
Date: 14 Jan 1999 19:00:31 +0000
[EMAIL PROTECTED] writes:
> I define with IP-aliasing 10 ip address on the SAME subnet for one ethernet
> adatper:
>
> eth0 eth0:1 .... eth0:9
> 10.0.0.1 10.0.0.2 .... 10.0.0.10
>
> I want to send IP packets from these different Logical IP interface to a host
> on the network 10.0.0.100. The problem is How can I select the source IP
> address ?
bind()
Of course that would require you to hack the sources for telnet.
Alternatively manipulate the routing table between starting each
connection.
If the server is on the same LAN:
route add server dev eth0:n
If the server is not on the same LAN then you have to play about with
the default route.
As far as I can see there should be no reason why you couldn't do:
route add server gw gateway dev eth0:n
Where gateway is whatever gateway you need to route through. Brobablt
the same as your default gateway.
However this doesn't work.
--
\\ ( ) No male bovine | Email: [EMAIL PROTECTED]
. _\\__[oo faeces from | Phones: +44 121 471 3789 (home)
.__/ \\ /\@ /~) /~[ /\/[ | +44 121 627 2173 (voice) 2175 (fax)
. l___\\ /~~) /~~[ / [ | PGP-fp: D7 03 2A 4B D8 3A 05 37...
# ll l\\ ~~~~ ~ ~ ~ ~ | http://www.wcl.bham.ac.uk/~bam/
###LL LL\\ (Brian McCauley) |
------------------------------
From: [EMAIL PROTECTED] (Ronald Duncan)
Subject: Re: Email Forwarding
Date: Fri, 15 Jan 1999 11:43:23 +0000
On Thu, 14 Jan 1999 21:56:42 Garth ONeale <[EMAIL PROTECTED]> wrote:
>How do I enable my linux box to forward email to other servers? I just
>can't find the answer.
>
Using sendmail, if the forwarding is to be permanent, edit the "aliases"
into /etc/aliases and restart sendmail. If temporary, investigate the
".forward" mechanism in the users home directory.
Hope this helps...
--
Ronald Duncan | http://www.duncans.co.uk
John Duncan & Son (est. 1824) | Email: [EMAIL PROTECTED]
------------------------------
From: Roger Wernersson <[EMAIL PROTECTED]>
Subject: PPPD
Date: Fri, 15 Jan 1999 14:14:47 +0100
Hi!
I have ISDN at home. I also have one RH 4.2 and one RH 5.2
installations. RH4.2 connects to the internet without problems but RH5.2
can't receive packets. Both connect using PPP and get a local
IP-address. IFCONFIG on RH5.2 reports the following after an
unsuccessful PING:
RX packets:10 errors:8 dropped:8
TX packets:15 errors:0 dropped:0
It seems as if though RH5.2 can send but not receive packets. RH4.2 uses
PPPD 2.2.0 and RH5.2 uses PPPD 2.3.5. I have tried installing PPPD 2.2.0
on RH5.2 without results.
When I try to load PPP as a module using "insmod ppp" I get the
following:
/lib/modules/2.0.36/net/ppp.o: unresolved symbol slhc_free_R3787e5b9
/lib/modules/2.0.36/net/ppp.o: unresolved symbol slhc_remember_Rbc0f8a5e
/lib/modules/2.0.36/net/ppp.o: unresolved symbol slhc_toss_R89ec95b6
/lib/modules/2.0.36/net/ppp.o: unresolved symbol
slhc_uncompress_Ra2ca7e04
/lib/modules/2.0.36/net/ppp.o: unresolved symbol slhc_compress_R5d6838a9
/lib/modules/2.0.36/net/ppp.o: unresolved symbol slhc_init_R20741a64
Is this normal? What am I doing wrong?
PS. I don't think ISDN has anything to do with it. DS.
\\\|///
\\ ~ ~ //
( @ @ )
====================-oOOo-(_)-oOOo-===================
Roger
============================-Oooo.-===================
.oooO ( )
( ) ) /
\ ( (_/
\_)
------------------------------
From: "Jamie Scott" <[EMAIL PROTECTED]>
Subject: DHCP Lease Times
Date: 15 Jan 1999 13:07:24 GMT
I am setting up my Linux box to provide DHCP services for my LAN and have
some questions about lease times:
1] If a lease expires for a Win95 client does the client have to restart
the machine or is it an automatic type of re-lease?
2] What settings of default-lease-time and max-lease-time would be put
into /etc/dhcpd.conf to have a non-expiring lease?
Thanks for the help,
Jamie Scott
[EMAIL PROTECTED]
------------------------------
From: "Phil Massyn" <[EMAIL PROTECTED]>
Subject: FTP for users
Date: 14 Jan 1999 18:04:27 GMT
Hi all,
I'm running WU-FTP 2.4.2 and I would like to setup a system where I can
have users FTP into the system and only have access to the directory I've
given them access to. Right now, my users can see the whole file system,
and I don't want it this way.
Any help will be much appreciated.
Phil Massyn
[EMAIL PROTECTED]
------------------------------
From: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
Subject: NFS linux server and an AIX client
Date: Thu, 14 Jan 1999 13:33:26 -0600
when i try to mount a NFS mount on an aixv3 client i recieve:
mount: 1831-010: server LINUX_server not responding: RPC: sucess
mount retrying
LINUX_server:/filesystem
mount: 1831-011: access denied for linux_server: /filesystem
mount: 1831-008 giving up on
LINUX_server:/filesystem
The file acces permissions do not allow the specified action.
my /etc/exports read
/filesystem -rw=AIX_server_hostname
/filesystem -rw=AIX_server_ipaddress
the exports file was exported.
NFS server was stopped and restarted
in messages on the linux server i get:
Kernel: Warning: possible SYN Flood from AIX_server on Linux_server:653
mountd: NFS mount of filesystem attemted from AIX_server
mountd: NFS client <anon client> tried to access filesystem
mountd: Blocked attempt attempt of AIX_server to mount filesystem
???????
where do i start looking? any ideas please reply via e-mail
[EMAIL PROTECTED]
TIA
------------------------------
From: Raymond Doetjes <[EMAIL PROTECTED]>
Subject: Re: FTP for users
Date: Thu, 14 Jan 1999 21:02:58 +0100
Well they can see your root but they can't download anything if they don't
have the rights to it.
But if you want it the way you've described it, that take alook at RedHats
anonynous FTP, there you can setup a series of directorys for persons that
have passwords but they can't crawl up to your root anymore. Most FTP
servers can do these things try examining your man on your ftpd.
Raymond
Phil Massyn wrote:
> Hi all,
>
> I'm running WU-FTP 2.4.2 and I would like to setup a system where I can
> have users FTP into the system and only have access to the directory I've
> given them access to. Right now, my users can see the whole file system,
> and I don't want it this way.
>
> Any help will be much appreciated.
>
> Phil Massyn
> [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Bruno Wolff III)
Crossposted-To: comp.mail.imap,comp.os.linux.setup,comp.mail.pine,linux.redhat.install
Subject: Re: Setting up imapd: how to check?
Date: 15 Jan 1999 13:49:22 GMT
Reply-To: [EMAIL PROTECTED]
At least some version of netscape URL encode passwords that have a small
number of special characters in them. That may be your problem.
Otherwise to test IMAP I suggest you get a copy of RFC 2060 and telnet
to your IMAP port and try things out.
>From article <[EMAIL PROTECTED]>, by S P
>Arif Sahari Wibowo <[EMAIL PROTECTED]>:
> Hi!
>
> I just setting up the imapd from pine 4.05 package in a RedHat linux
> system, however when I try to login using Netscape from other machine, it
> always fail to login (as if it was from wrong password).
>
> Any idea on this? How I check what actually happened?
>
> Thanks.
------------------------------
From: Sim <[EMAIL PROTECTED]>
Subject: Need TCP/IP routing guru assistance
Date: Thu, 14 Jan 1999 20:52:49 +0100
Reply-To: [EMAIL PROTECTED]
Because of my own stupidity, I wrecked a perfectly operating Linux
(RH5.2 2.1.127) environment. Now I need help getting it back together
again.
I have two network cards:
- one connected to a cable modem and calling dhcp successfully at start
up
- the other connects to my second computer and has a fixed IP address
(192.168.0.254)
The two computers can ping each other successfully, but my second
computer can no longer access the samba server on the first, and the ip
masquerading is no longer working. In addition, from my main computer I
get the following phonemena:
[root@P350 /root]# traceroute www.netscape.com
traceroute: Warning: Multiple interfaces found; using 192.168.0.254 @
eth0
traceroute to www-uk2.netscape.com (195.44.254.3), 30 hops max, 40 byte
packets
1 * * *
2 * * *
3 * * *
4 * * *
5 * * *
6 * * *
7 * * *
8 * * *
9 * * * ad infinitum
[root@P350 /root]# netstat -nr
Kernel IP routing table
Destination Gateway Genmask Flags MSS Window irtt
Iface
195.162.208.0 0.0.0.0 255.255.255.0 U 0 0 0
eth1
192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0
eth0
192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0
eth0
127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0
lo
0.0.0.0 195.162.208.254 0.0.0.0 UG 0 0 0
eth1
I'm sure this is a routing problem, but don't know how it crept in. I
can't find any obvious additional routes being added fromanything
controlled by linuxconf. I need help.
Simon
------------------------------
From: Raymond Doetjes <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Remote backups as root?
Date: Thu, 14 Jan 1999 21:07:56 +0100
You could mount the file system of the other server to a directory on the
server with tapestreamer and you back up it as if it is your own drive. You
can do this by using smabs and smbmount (wich is pretty fast) or doing it
with NFS (wich is not that fast).
Or you could take alook at BRU 2000 server edition.
Raymond
Raymond
Sjoerd Krol wrote:
> Haai there,
>
> I've 2 Linux machines and one off them has a tapedrive attached to it
> (HP Surestore 5000). Currently I'm using the dump and restore commands
> on the one with the tape attached. I want to do the same with the
> remote machine, but I can't do it from the root account. root can't do
> any remote commands. For a normal user it's no problem (.rhosts file),
> but when backing up the whole system you need to be root no?
> Anyway does anyone know how to backup the remote system?
> Or an other way to do it?
>
> 2nd : what are the best parameters to put in for the dump command
> for a 4mm 2G 90m DDS tape?
>
> Thanks in advance!
>
> Greetings,
> Sjoerd
> +-------------------------------------------------------+
> | E-mail : [EMAIL PROTECTED] |
> | [EMAIL PROTECTED] |
> | World Wide Web : http://web.inter.NL.net/users/S.Krol |
> +-------------------------------------------------------+
> "What are you doing?"
> "Examining the world's major religions. I'm looking for something
> that's light on morals, has lots of holidays, and with a short initiation
> period."
------------------------------
** 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
******************************
