Linux-Setup Digest #380, Volume #21 Tue, 5 Jun 01 22:13:07 EDT
Contents:
Re: booting Linux on hdb (KCmaniac)
Weird lilo.conf Re: Weird lilo problem, never seen before (Stanislaw Flatto)
Re: booting Linux on hdb ("Peter T. Breuer")
Re: named - fails to recognize service is already running?? (Hal Burgiss)
Re: linux box cannot be seen in networkneighbourhood of win2000. (M)
Re: Netscape - How Can I Install/Run? (KCmaniac)
Re: rc.local file. (Floyd Davidson)
Re: Building Kernel 2.4.5 (Steve)
Re: booting Linux on hdb ("Peter T. Breuer")
How to setup Downloader for X with Firewall? ("Huy Vu")
hylafax server ("cat")
Re: slooooooooooooooooooooowwwwwwwwwwwwwww!!!!!!!!!!! ("Joe Public")
Re: Red Hat Newbie -- Can I do this? (Dave Berntson)
Re: how to setup mail server... (Dataviz)
Re: XFree86 problems (Dave Uhring)
----------------------------------------------------------------------------
From: KCmaniac <[EMAIL PROTECTED]>
Subject: Re: booting Linux on hdb
Date: Tue, 05 Jun 2001 19:10:11 -0400
This might be an easy question for you to answer, Peter.
I am using some oddly named 2.2.19pre17-idepci kernal in Debian. Now, this
kernal was installed from the get go. Can I go get a newer kernal from
lets say kernal.org and successfully install it in place of this present
one or do I have to have a "Debian flavored version" of these kernals? Is
this pre17 kernal just for the installation process? I guess what I am
asking is can I install the same kernal in Debian that I could in RedHat?
Thanks,
RLH
------------------------------
From: Stanislaw Flatto <[EMAIL PROTECTED]>
Subject: Weird lilo.conf Re: Weird lilo problem, never seen before
Date: Wed, 06 Jun 2001 09:42:24 +1000
If what you posted is the whole lilo.conf then the waiting period of 15 seconds
is not needed, what else can it boot?
Is it your own box? Why password?
Trimm the lilo.conf to normality, rerun lilo and no more problems.
(man lilo and man lilo.conf as helpers)
Have fun
Stanislaw
Slack user from Ulladulla.
David Anderson wrote:
> Hi
>
> I have recently installed RedHat 6.0 on an old Compaq server, equiped with
> three 2Gb SCSI HDDs and an scsi cdrom drive.
>
> My problem is with lilo: After installing it with the lilo.conf file
> following, upon reboot the machine has a most bizarre attitude. If I wait
> the 15 seconds I have set for timeout for linux to boot, it boots fine and
> dandy. But if I dare type 'linux' myself or so much as press enter to speed
> things up, the kernel crashes out after detecting the scsi drives:
>
> Error: unable to access /dev/sda 00:80
> Kernel Panic: unable to mount /dev/sda1 on 00:80
>
> Or something like that... How can I fix this annoying bug?
>
> < Beginning of lilo.conf file >
> boot=/dev/sda
> map=/boot/map
> install=/boot/boot.b
> message=/boot/message
> prompt
> timeout=150
> image=/boot/vmlinuz-2.2.5-15
> label=linux
> root=/dev/sda1
> initrd=/boot/initrd-2.2.5-15.img
> read-only
> restricted
> password=********
> < End of lilo.conf file >
>
> I tried installing GRUB instead of lilo, but it can't load linux at all,
> boasting the same error...
>
> Thanks in advance
> David Anderson
------------------------------
From: "Peter T. Breuer" <[EMAIL PROTECTED]>
Subject: Re: booting Linux on hdb
Date: Wed, 6 Jun 2001 01:16:31 +0200
KCmaniac <[EMAIL PROTECTED]> wrote:
> By the way, Peter, I did get that WinModem fired up on Debian. The driver
Oh, it's you.
> didn't want to install with the particular kernal image I had. I don't
> understand it. The package maintainer said it was looking for another
> image of that version. I guess he was talking about me having to
Apparently the same version got packaged twice with the same or similar
name. I said there was a near-typo, I seem to recall?
> re-compile another kernal to create the right imgage but the kernal I have
> now isn't any different according to him. He said he could drop the
If it's the same version, then it probably isn't significantly
different (compilation differences can be significant, bi wrt to the
module you want to compile and install).
> dependency but this method was "cleaner". He told me I should be able to
> just force the install without any lose. I did and it did and it works.
Yes, that's fine.
> So thats over with.
> I had already tried what you suggested in LILO
An important point is where the bootsector goes. I suggested /dev/hda.
(i.e. MBR of the first disk).
> image=/boot/<image-name>
> root=/dev/hdb5
This is more or less fine.
> It told me something about the absence of or an improper boot signature.
"proper" boot signature. That'll be a msdos partition in an
undealable-with format. Probably a win2002 FS.
> Something to do with boot signature.
> You said IF my BIOS was capable of booting the slave. How do I know if it
> is or not?
You can't tell except by disassembling it and/or trying. You could ask the
manufacturer.
> About the documentation and how I didn't seem to get the answers I was
> looking for. This is a case in point. If you are correct about the BIOS'
> cabability as being a limiting factor to boot hdb, I never saw that
> anywhere in the LILO documentation. I am not saying it isn't in there, I
Well, it's obvious. Lilo has to call the bios. There's nothing else
there to call when that 512bytes of lilo boot sector runs! And lilo
is pretty much a single-stage loader, so it doesn't load a which loads
b which loads c, etc. etc. If you want to play that game, you can load
dos/windows and then boot linux from there.
> just never saw it and I covered it pretty good. I would think I would have
> seen it if it was in there. It talks about where to put LILO in dual disks
> configurations but I didn't see anything about how to boot hdb with LILO.
There's no real difference with booting hda in terms of config files (it
should mention something about using chain.b sometimes in the docs, but
that's all). BTW, if all you want to do is boot and this isn't an
experiment to test the bios, you might as well put the kernel image and
/boot directory on the windows partition in /dev/hda. Then you don't
have to think about whether the bios can or can't boot from hdb. You
only have to worry about windows messing the files.
> RLH
--
=====================================================================
Peter T. Breuer MA CASM PhD. Ing., Asoc. Prof.
Area de Ingenieria Telematica E-mail: [EMAIL PROTECTED]
Dpto. Ingenieria Tel: +34 91 624 91 80
Universidad Carlos III de Madrid Fax: +34 91 624 94 30/65
Butarque 15, E-28911 Leganes URL: http://www.it.uc3m.es/~ptb
Spain
------------------------------
From: [EMAIL PROTECTED] (Hal Burgiss)
Crossposted-To:
comp.os.linux.misc,comp.unix.questions,linux.redhat,linux.redhat.install
Subject: Re: named - fails to recognize service is already running??
Reply-To: Hal Burgiss <[EMAIL PROTECTED]>
Date: 5 Jun 2001 19:32:53 -0400
On Tue, 05 Jun 2001 16:43:05 -0400, inetquestion <[EMAIL PROTECTED]> wrote:
>If I allow it to start named I end up with a ton of named processes
>running. If I choose do nothing, and exit the system works as normal.
>However I always get the same message the next time I run linuxconf.
>I looked in /etc/rc.d/rc3.d and found two files that began with S55*
>which were S55named and S55sshd. I renamed S55named to S58named and
>restarted the system. Upon entering linuxconf I got the same message
>on exit except this time it was S58named instead of S55named.
>
>I'm not an expert on the System V startup process, but originall thouht
>this was caused by two links begging with the same number "S55", but I
>noticed that this is not the only occurance of this, and the other two
>services seem to have no problem.
>
>also after bootup when I do "ps -ef | grep named"
>I see at lease 4 instances of named running.
[hal@feenix hal]$ ps |grep named
hal 2823 0.0 0.1 1428 468 pts/0 S 19:28 0:00 grep named
named 8115 0.0 1.5 13880 3944 ? S 11:53 0:00 named -u named
named 8117 0.0 1.5 13880 3944 ? S 11:53 0:00 named -u named
named 8118 0.0 1.5 13880 3944 ? S 11:53 0:03 named -u named
named 8119 0.0 1.5 13880 3944 ? S 11:53 0:03 named -u named
named 8120 0.0 1.5 13880 3944 ? S 11:53 0:00 named -u named
named 8121 0.0 1.5 13880 3944 ? S 11:53 0:01 named -u named
I am guessing that BIND 9.x is multi threaded and this is 'normal'. Also
you probably have named listening on multip interfaces. By default I
believe it will listen on every interface it can, including lo.
Possibly the problem is linuxconf not dealing with correctly with some
aspect of this. I get this
[root@feenix tmp]# service named status
rndc: connect: connection refused
which I have no idea why, but named seems to start and run OK, although
it complains some during boot.
--
Hal B
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
Spamtrap: [EMAIL PROTECTED] and [EMAIL PROTECTED]
--
------------------------------
From: M <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.misc
Subject: Re: linux box cannot be seen in networkneighbourhood of win2000.
Date: Tue, 05 Jun 2001 19:44:49 -0400
Can your 2000 box see the Linux box at the network layer (ping the IP address)?
Also, what happens if the ping the hostname of the linux box? If it has
connectivity, it should still be able to respond to the netbios broadcast
request. If 2000 does not see a host belonging to a specific workgroup, it may
have trouble navigating it net neighbourhood.
mf
Liverpool_fc wrote:
> thank you.
>
> all the pc's and server are in the same subnet.
> 192.168.27.1 - 192.168.27.100
> there are no wins, dhcp or dns servers.
>
> simple lan with one rh6.2 box, 5 win98/95 boxes and one win2000 workstation.
> only the win2000 cannot see the rh6.2 server in network neighbourhood.
>
> thank you.
------------------------------
From: KCmaniac <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.install
Subject: Re: Netscape - How Can I Install/Run?
Date: Tue, 05 Jun 2001 19:47:55 -0400
I am having the same problem with v4.77 in that I can't get the executable to do
anything.
When I use "ldd <path/to/executable/filename>" it tells me that
libstdc++-libc6.1.1.so.2 is not found.
>From what Gregory D. Horne wrote:
> Netscape v4.77 uses an older version of libstdc++-libc6, so you might need to
> create a symbolic link from libstdc++-libc6.1.1.so.2 to libstdc++-libc6.1-2.so.3
> (or whatever newer version you may have installed).
>
it appears that I indeed need that lib file or evidently a newer version of it
anyway. I don't understand why it wouldn't be part of the installation package if
it needs it that much in the first place. But where do I get this library file??
Is it already in compiled form? Where do I put it on my system? In other words
where is Netscape expecting it to be?
Any help on this would be appreciated.
RLH
------------------------------
From: Floyd Davidson <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.misc
Subject: Re: rc.local file.
Date: 05 Jun 2001 15:28:56 -0800
"Liverpool_fc" <[EMAIL PROTECTED]> wrote:
>hello,
>
>it appears that in rh7.0 the rc.local file is the new
>autoexec.bat (dos) equivalent.
New? You realize that UNIX was around before DOS, right? The
concept of rc files is older than DOS's braindead way of doing
things.
>however this file looks like a script. and i need to add a few services/
>modules that start on boot.
>example virtport (command) (options).
>from my experience if you fiddle with a script your syntax needs to be
>correct.
>adding a command line like the example above will cause the script to give
>errors.
>
>any suggestions?
Yep. Learn how to write shell scripts, and learn how to test
them. Also learn how to boot you system to single user
runlevel, or from a rescue disk, *before* you start screwing
with boot scripts. Because while it is exactly true that the
syntax needs to be correct, it is also true that the results
need to be valid too (running the wrong program is just as bad
as mispelling the name of the right one!), and worst of all...
even the best of system admins makes a mistake now and then.
Hence be prepared to recover from your mistakes.
You might also consider a cheap second computer that is used
just to play with such things as learning systems admin.
--
Floyd L. Davidson <http://www.ptialaska.net/~floyd>
Ukpeagvik (Barrow, Alaska) [EMAIL PROTECTED]
------------------------------
From: Steve<[EMAIL PROTECTED]>
Subject: Re: Building Kernel 2.4.5
Date: Tue, 5 Jun 2001 20:29:52 -0400
> > Provided that everything is enabled correctly in the system bios, I have a
> > somewhat unusual question, "what kind of mouse are you using"?
>
> Just a simple Logitech PS/2 mouse. Nothing special about it really. I've
> checked my XF86Config to see what type of mouse it's running as under
> 2.2, and yep, it's a PS/2. As an update, even with kernel 2.2 where the
> mouse does work, catting /dev/psaux and /dev/mouse doesn't have any
> effect, so maybe that wasn't such a terrific diagnostic.
>
> Joal Heagney/AncientHart
>
I ran into problems with a PS/2 mouse (pointing stick on my laptop) and
the 2.4.3 kernel, but building the 2.4.5 kernel from scratch cured it.
My basic install is Mandrake 8.0 (kernel 2.4.3).
I've heard of a few people with other problems with PS/2 mice and 2.4.x
kernels, but most of those seem to have been interrupt conflicts (see if
/proc/interrupts has IRQ 12 assigned to something else besides psaux - my
apologies if you've already tried this). The most common conflict is a
PCI soundcard.
-- Steve
------------------------------
From: "Peter T. Breuer" <[EMAIL PROTECTED]>
Subject: Re: booting Linux on hdb
Date: Wed, 6 Jun 2001 02:32:27 +0200
KCmaniac <[EMAIL PROTECTED]> wrote:
> I am using some oddly named 2.2.19pre17-idepci kernal in Debian. Now, this
> kernal was installed from the get go. Can I go get a newer kernal from
> lets say kernal.org and successfully install it in place of this present
Yes.
> one or do I have to have a "Debian flavored version" of these kernals? Is
No.
> this pre17 kernal just for the installation process? I guess what I am
No.
> asking is can I install the same kernal in Debian that I could in RedHat?
Yes. Provided you have the tools that match it (correct modutils,
pppd, psutils, etc.).
(but redhat kernels are horrible .. please compile your own from
pristine source).
Peter
------------------------------
From: "Huy Vu" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.security
Subject: How to setup Downloader for X with Firewall?
Date: Wed, 06 Jun 2001 01:05:24 GMT
Hi all,
I'm trying to setup Downloader for X under Redhat 7.1 to get through our
Firewall but still cannot.
When I do ftp ftp.site.com, our firewall will ask me login/password of the
firewall before connect to ftp.site.com, after that I have to type USER and
PASS to login to ftp.site.com.
With Downloader for X, our firewall takes the value username/password
setting to connect to ftp.site.com for firewall's authentication and after
that it has no more value for authentication to ftp.site.com
Does someone can use Downloader for X under firewall?
PS: I'm successful to get ncftp get through our firewall in setup
.ncftp/firewall file using authentication number 6
Thanks in advance for any help
DH.
------------------------------
From: "cat" <[EMAIL PROTECTED]>
Subject: hylafax server
Date: Wed, 6 Jun 2001 09:10:41 +0800
My fax server is hylafax. When I send a fax, the messages file log :
OPEN /dev/ttyS1
/dev/ttyS1: Can not initialize modem
the modem can work on suse linux 6.3
but not on suse linux 7.1
Actually, I try to recieve the fax signal by Minicom. It seems successful.
However, when I try it again using Hylafax. It fails!!!!
------------------------------
From: "Joe Public" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.misc
Subject: Re: slooooooooooooooooooooowwwwwwwwwwwwwww!!!!!!!!!!!
Date: Wed, 6 Jun 2001 02:37:34 -0000
> > I have 128mb of memory. 250 mb swap. Pentium 3 - 700mh cpu.
> > correction even telnet is extremely slow getting in.
Do you find that it takes a minute or two to connect via telnet, but then
telnet is fast? If so, it is simple problem.
When you telnet to a box the telnet server needs to do a reverse look up on
the IP address of the telnet client (get the host name from the IP address)
so that it can log the access in /var/log/messages. This can be done by a
hosts file entry (/etc/hosts) or by a DNS. If there is no way to so the
reverse lookup, then the telnet server will wait a minute or two for the
resolver to time out (the resolver does the reverse lookup).
You need to check the following
/etc/resolv.conf - DNS entris go here
/etc/hosts - host file
/etc/nsswitch.conf - sets the method for looking up name services, eg hosts
first, then DNS then stop. Tip. Don't use NIS if you don't have a NIS
server!
John Auld
------------------------------
From: Dave Berntson <[EMAIL PROTECTED]>
Subject: Re: Red Hat Newbie -- Can I do this?
Date: Tue, 05 Jun 2001 20:14:37 -0500
On 5 Jun 2001 13:12:16 -0700, [EMAIL PROTECTED] (Michael Ballai) wrote:
>> How big is your hard drive and how much of it is taken up with win98?
>> db
>Without looking, I think it's 5.5-6 gigs and probably less than about 2 gigs are used
Great!! Make sure you defrag your hard drive before you use fips (First
non-destructive partition splitting program). It's a small program that
usually is included in most distro's, but easily downloaded from the net.
You'll have to decide how much room to leave on the windows partition for
future "growth".
db
------------------------------
From: [EMAIL PROTECTED] (Dataviz)
Crossposted-To:
comp.os.linux.networking,comp.os.linux.questions,comp.os.linux.redhat,linux.redhat
Subject: Re: how to setup mail server...
Date: 5 Jun 2001 18:52:36 -0700
http://www.stalker.com/CommuniGatePro/default.html#Current
"MD" <[EMAIL PROTECTED]> wrote in message news:<[EMAIL PROTECTED]>...
> Hi,
> Is there any step by step information to setup a mail server in Redhat 7 ?,
> which is sendmail or postfix configuration for a static IP which I will need
> to setup smtp/pop. Pls help..
>
> Thanks in advance,
> mrcopter
------------------------------
From: Dave Uhring <[EMAIL PROTECTED]>
Subject: Re: XFree86 problems
Date: Tue, 5 Jun 2001 21:00:34 -0500
Michael David Pedersen wrote:
> Hej all.
> I have now been trying to get XFree86 4 to work for days, but without
> luck. My system is a AMD Athlon 1200 on a soltek 75-KAV motherboard and a
> Palit Daytona GeForce 2 Pro graphics card. My monitor is an old Sony
> Multiscan 200GS.
> I startet the SuSE 7.0 installation (with Yast2), and when all the
> packages had been copied, the screen turned weird and the computer froze.
> I then started over again, installing a base system with Yast1, and linux
> is now working. I have downloaded the Detonator 3 drivers from Nvidia, and
> installed them. But I cannot get either SaX or SaX2 running, and I have
> tried with a lot of different command line arguments.
> As far as I can see, the /var/log/XFree86.0.log does'nt reveal much except
> for "Fatal server error: Caught signal 4. Server aborting" at the end of
> the log. Some times it said "Caught signal 11" instead. There are a lot of
> warnings complaining about fonts not being avalible, but I also had these
> warnings on a different system which worked perfectly.
> I have also tried creating the /etc/XF86Config with xf86config, but with
> the same result.
>
> So now I am getting desperate for this to work, and any help or
> suggestions would be greatly apreciated!
> Thanks a lot!
>
> - Michael.
>
>
>
>
Try the method recommended in the install notes from www.xfree86.org
# XFree86 -configure
which will leave a config file in your /root directory. You may need to
edit that file for your mouse protocol and device, also for the default
depth. Then test the config file as the screen display has instructed. If
it works you will get the herringbone pattern screen with a movable mouse
pointer. Copy the config file to /etc/X11/XF86Config-4. Fine tune it with
xf86cfg.
------------------------------
** 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 by posting to comp.os.linux.setup.
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-Setup Digest
******************************
