Linux-Setup Digest #170, Volume #20 Wed, 6 Dec 00 04:13:08 EST
Contents:
Re: IRQ Problems (James Richard Tyrer)
Sendmail question (root)
Offical Debian 2.1 CD - no SCSI (Ken)
I'm looking for info on bootscripts ("Anastasios Marmaras")
RH7 - can't get video configuration to work for X ("TSHACK")
Netscape installing ERROR? (Dean Kwak)
Re: How to fix bad sectors on HD? ("michael.fengler")
Re: Init.d stops after a given number of started services ("michael.fengler")
Re: Kernel won't mount raid0 (solved) ("Peter T. Breuer")
Re: control/alt/F2-F6 no longer work since installing helix gnome (Michael Perry)
Re: I'm screwed!!! Gnome menu fonts GONE! (Michael Perry)
Re: another newbie, trouble getting started (Michael Perry)
Re: Best Scsi card? (Michael Perry)
Re: windows VFAT partitions too fat?! (Eric)
Re: why can't one boot from /dev/hdb? (Eric)
What is this message??How to solve??thanks!! ("kitty")
----------------------------------------------------------------------------
From: James Richard Tyrer <[EMAIL PROTECTED]>
Subject: Re: IRQ Problems
Date: Wed, 06 Dec 2000 05:27:11 GMT
Matthew Alicton wrote:
>
> Hi there,
> I recently purchased a new linux compatible modem. Although I can get
> it to work under linux, there is one big problem, I always have to
> manually the irq. The modem is on com3 and the default irq for com3 is
> something other than what I have. So before I dial-up anywhere, I always
>
> have to 'setserial /dev/modem irq 5'. What I wonder about, is there any
> config file, or utility out there that will assign or set (as default)
> the irq for com3 as irq 5? Simple isn't it, but I simply cannot figure
> it out.
Do you have the serial script in your init scripts?
There is a file "rc.serial" in the documentation directory for the
setserial package:
setserial Version 2.15 (2-Nov-98)
Setserial is a program which allows you to look at and change various
attributes of a serial device, including its port, its IRQ, and other
serial port options.
Starting with Linux 0.99 pl10, only the COM1-4 ports are configured,
using the default IRQ of 4 and 3. So, if you have any other serial
ports provided by other boards (such as an AST Fourport), or if COM3-4
have been a non-standard IRQ so that you can use time simultaneously
with COM1-2, you *must* use this program in order to configure those
serial ports.
The simplest way to configure the serial ports is to copy the provided
rc.serial file to the appropriate /etc/rc.d directory. For example,
to install rc.serial on a RedHat system, copy rc.serial to
/etc/rc.d/init.d/serial, and then run the command "chkconfig -add
serial".
I would use the GUI script configuration tool, ksysv in KDE, to check
this, or to do it.
After you do this, your serial configuration will be saved when you
shutdown and restored when you reboot.
JRT
------------------------------
From: root <[EMAIL PROTECTED]>
Subject: Sendmail question
Date: Wed, 06 Dec 2000 05:34:22 GMT
Sendmail question
I am running RedHat Linux 6.2. As my system boots up it loads Sendmail
just fine and gives an [OK] status. When I log onto the system I can
send out e-mail to friends and family fine. But when they send e-mail
back to me they get the following bounce back:
"This Message was undeliverable due to the following reason:
Each of the following recipients was rejected by a remote mail server.
The reasons given by the server are included to help you determine why
each recipient was rejected.
Recipient: <[EMAIL PROTECTED]>
Reason: <[EMAIL PROTECTED]>... Relaying denied
Please reply to [EMAIL PROTECTED]
if you feel this message to be in error."
Can anyone help. What do I need to do to receive e-mail on my server to
my local domain?
Thanks for your help.
------------------------------
From: Ken <[EMAIL PROTECTED]>
Subject: Offical Debian 2.1 CD - no SCSI
Date: 6 Dec 2000 04:24:02 GMT
I am trying to load my "Official" Debian 2.1 CDs but when I boot to
it, it tells me that I have no hard drives installed. Anyone know why
it would do this? I have an Adaptec 2930CU with three SCSI drives
attached.
Thanks for any input.
-Ken
------------------------------
From: "Anastasios Marmaras" <[EMAIL PROTECTED]>
Subject: I'm looking for info on bootscripts
Date: Wed, 6 Dec 2000 08:35:22 +0200
I am looking for information about init and boot scripts. I have read the
init and inittab manpages, the power2bash and Linux From Scratch manuals and
O'Reilly's Running Linux but couldn't find enough information to be able to
actually tamper with these files. (without disastrous effects that is...:))
If there is anyone here who can point me to more online reading material on
the subject or suggest a good book I would really appreciate the help.
Anastasios Marmaras ( [EMAIL PROTECTED] )
------------------------------
From: "TSHACK" <[EMAIL PROTECTED]>
Subject: RH7 - can't get video configuration to work for X
Date: Tue, 5 Dec 2000 22:40:33 -0800
I've got Linux 7.0.
I've got an older 200mhz pentium computer.
My video card (on the motherboard) is SiS5597/5598 (as per autoprobe and
Windows boot screen), with 2mb video memory.
My monitor is an Impression 3 Plus 3428NG.
Both of these show up in the hardware database when trying to install, but I
can't find them on the Hardware Compatibility List on www.redhat.com.
Though I got this configuration working once, I did a reinstall and I can't
get it working again.
I keep getting "There is a problem with your configuration" and have to go
back.
I've tried every screen resolution one at a time, and no luck. I've tried
manually inputting the screen refresh rates. No luck.
What do I do to troubleshoot this? I'm an absolute newbie to Linux.
I'm even having problems logging into Redhat to get web support, just to
show how dumb/frustrated I feel with this.
Dwight
------------------------------
From: Dean Kwak <[EMAIL PROTECTED]>
Subject: Netscape installing ERROR?
Date: Tue, 5 Dec 2000 22:20:16 -0800
i just downloaded netscape 6 installer and ran the following command as
Netscape says:
gzip -dc netscape-i686-pc-linux-gnu-installer.tar.gz | tar -xvf -
is there something wrong with the command? because when ran the command i
get this error message
"tar: Skipping to next header"
"gzip -dc netscape.......: invalid compressed data--crc error"
"tar: Error exit delayed from previous errors"
I am totally lost here i need your input. thanks for your help in
advance.
please reply to "[EMAIL PROTECTED]"
------------------------------
From: "michael.fengler" <[EMAIL PROTECTED]>
Subject: Re: How to fix bad sectors on HD?
Date: Wed, 6 Dec 2000 07:41:38 +0100
Reply-To: Michael Fengler <[EMAIL PROTECTED]>
On Wed, 6 Dec 2000, Jc wrote:
>I need a fix, not a replacement drive. I'm doing this for someone else who can't
>afford to go and buy another drive.
>
Tough luck. Back everything up then fdisk + mkfs. Reinstall the backup.
However, be prepared to do this on a daily basis.
*There are no software solutions to hardware problems*
- mike
------------------------------
From: "michael.fengler" <[EMAIL PROTECTED]>
Subject: Re: Init.d stops after a given number of started services
Date: Wed, 6 Dec 2000 07:54:05 +0100
Reply-To: Michael Fengler <[EMAIL PROTECTED]>
On Tue, 5 Dec 2000, Pierre-Philippe Ravier wrote:
>I need to be a SMB client as well in order to print to a Windows NT machine,
>but I don't need to be a SMB server. Therefore the smb daemon is probably
>useless in my case, can you confirm that ?
That's true, smbclient works standalone. IIRC it *does* look into
smb.conf, however.
- mike
------------------------------
From: "Peter T. Breuer" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: Kernel won't mount raid0 (solved)
Date: Wed, 6 Dec 2000 07:47:25 +0100
In comp.os.linux.misc Stephan A Suerken <[EMAIL PROTECTED]> wrote:
[ of striped raid root ]]
> I don't think there are only very few people doing it, as it's a good
> thing (tm)!
What's the point? I don't see any. It's not more robust and it's not
faster if you mean / (though it has lower first-time latency).
> I needed 2.2.18(prewhatever) as I needed USB. That just to
> explain. Basically, its not the kernel causing the "failure" here. I
> actually tested it with 2.2.17 stable AND 2.4.0-test<something> as
> well.
Better report the failure.
> Still not finding any info about how to exactly specify the "chunk size"
> in the kernel parms, I made a exhaustive test. The results for
> my two raid0s are:
> mdtab : /dev/md0 raid0,16k,0,802eab69 /dev/hda3 /dev/hdc2
> kernel args: md=0,0,2,0,/dev/hda3,/dev/hdc2
> ^--"chunk size": 2 works
> mdtab : /dev/md0 raid0,8k,0,08716985 /dev/sda3 /dev/sdb1 /dev/sdc1
> kernel args: md=0,0,1,0,/dev/sda3,/dev/sdb1,/dev/sdc1
> ^--"chunk size": 1 works
> Other values for the chunk size (factor) (tested 1-16) do NOT
> work. Obviously, "mdtab-value == kargs-value x 8k" must be met. This
> really should be in kernel's md.txt, imho.
How would that work with 4K stripes? (;-).
> My normal disk / fs layout is to only have a small /boot partition on
> a bios-secure place and some swap space partition. All the rest is
> root-raid. This obviously has two advantages:
> 1. Speed (striping).
No speed increase on / as it is cached. It only contains libraries
(cached) and utilities needed at boot (not needed after boot).
Oh .. you are saying that you are striping /usr? Not /? Yes, striping
/usr has advantages. But you lose heavily in robustness by striping
/! And you only gain in speed if you stripe over two controllers.
Well, I suppose you can sync / to a different partition daily.
> 2. No fixed space mounted partitions (which is REALLY annoying).
??How does raid enable you to resize partitions? Yes, it enables you to
combine partitions to taste (which is what I use it for).
> I don't have a clean "old-style" root partition any more to boot
> from. I get security by other means. Previously, I had that, using the
> raid via symlinks, but that solution has some disadvantages, besides
> the fact that it is ugly !
Peter
------------------------------
From: [EMAIL PROTECTED] (Michael Perry)
Crossposted-To: comp.os.linux.x
Subject: Re: control/alt/F2-F6 no longer work since installing helix gnome
Date: Tue, 5 Dec 2000 22:30:52 -0800
On Mon, 04 Dec 2000 02:55:55 GMT, The.Central.Scrutinizer.wakawaka@invalid.
pobox.com <[EMAIL PROTECTED]> wrote:
>On Sun, 03 Dec 2000 18:33:17 -0500, Norman Levin <[EMAIL PROTECTED]> wrote:
>>[EMAIL PROTECTED] wrote:
>>>
>>> System is an IBM netvista w/ the dreaded intel I815 video.
>>>
>>> I installed helix gnome on it and now when I try control/alt/F2,etc., I get
>>> a blank screen. Hitting c/a/F7 doesn't restore the desktop. Control/alt/BS
>>> doesn't work either. The only way to restore the screen is c/a/d/.
>>>
>>> I'd like some hints and/or a clue how to debug this.
>>** the xserver owns the hardware at the xstation which is your keyboard, mouse,
>tablet
>>and screen. X can handle CTL/ATL keys or ignore them. Check config file
>
>Can you be more specific? I see no mention of them in any config file. It
>isn't just that they don't work; they switch the display to an invalid mode
>making it impossible to even return to the desktop. The only thing I can do at
>that point is control/alt/del to reboot. Even control/alt/backspace doesn't
>restore the display if I've hit c/a/F2-6.
Are you using gdm to boot helixcode? I have seen this behavior with a
mis-configured gdm on a workstation I have. I ceased using a graphical
display manager after this type of incident and when I mangled a video card
display driver once and pretty much lost the ability for awhile to get X
running.
You can remove gdm or change the default runlevel to console mode and see if
that helps.
--
Michael Perry
[EMAIL PROTECTED]
==================
------------------------------
From: [EMAIL PROTECTED] (Michael Perry)
Subject: Re: I'm screwed!!! Gnome menu fonts GONE!
Date: Tue, 5 Dec 2000 22:35:34 -0800
On Tue, 05 Dec 2000 18:08:37 -0700, Adam J Boyle <[EMAIL PROTECTED]> wrote:
>hi all,
>
>I desperately need some help. This morning when I booted X the gnome
>main menu was very skinny and only showed the icons but no fonts! Same
>thing with the menus when I right-click on the desktop, etc. Also, with
>the Control Center I am missing many fonts in that dialog as well. I've
>restarted, reinstalled helix-gnome, rebooted, etc. but to no avail.
>
>But note that all of my fonts in Netscape (I'm writing this from the
>same machine right now) and the gnome-terminal windows show up fine
>(except for the menu at the top of the gnome-terminal window).
>
>I am running the following:
>
>RH 6.2
>Helix Gnome (yes, I reinstalled the whole thing again today after I had
>this problem, but it didn't help!)
>256 MB memory
>PIII 733
>
>
>Has anyone seen or heard of this before? I have exhausted almost every
>possible option, but nothing seems to help.
>
>Thanks in advance.
>
>ajboyle
>
I have not heard of this specifically; but I have had problems with varying
things in the .gnome directory and the .gnome-desktop directory. Sometimes
cleaning these directories out helps. If there is some bit of corruption in
the settings files that are preserved in your user's home directory,
uninstalling and reinstalling won't do much since they do not erase the
hidden .gnome system directories. If you ace .gnome, you start all over
again with a new setup, I believe. Also check out .gnome-desktop for the
icons on your desktop, etc.
--
Michael Perry
[EMAIL PROTECTED]
==================
------------------------------
From: [EMAIL PROTECTED] (Michael Perry)
Subject: Re: another newbie, trouble getting started
Date: Tue, 5 Dec 2000 22:26:12 -0800
On Tue, 5 Dec 2000 15:29:05 -0800, Mark
<[EMAIL PROTECTED]> wrote:
>I downloaded the two RH7.0 iso images off of a redhat mirror. I burned the
>first one to CD using EasyCD creator 4.0... i was under the impression that
>it was a bootable. That .iso file is the only thing on the cd, however no
>system will boot off the CD. If I put my Win2K Pro CD in, those systems
>boot fine off the CD.
>
>What did I miss?
>
>
If the cd's wont boot, you can download boot floppies and build them off of
the cd using rawrite. I don't use windows stuff to burn cd's but I also
don't use redhat so I don't know the particulars about the mirror. I tend
to use xcdroast on a debian box I have for cd burning. The boot images are
usually in dosimages or dosutils or something like that.
--
Michael Perry
[EMAIL PROTECTED]
==================
------------------------------
From: [EMAIL PROTECTED] (Michael Perry)
Subject: Re: Best Scsi card?
Date: Tue, 5 Dec 2000 22:27:37 -0800
On Mon, 04 Dec 2000 07:17:25 GMT, The Chief <[EMAIL PROTECTED]> wrote:
>Hi: Iam having trouble setting up my scsi card adaptec ava-1505ae for my
>scanner/zip disk! I decided to buy a different one!which is best? my linux
>setup is mandrake 7.1 on my PC Clone!
>
Well, I use adaptec 2940s with pretty good results overall. I have a
adaptec 2940 AU that does not boot hard disks. It only manages a zip drive
for me. My other system is a scsi system with twin 17g IBM drives and a
adaptec 2940U2W. This system also has a tecram tape drive in it and runs
flawlessly.
--
Michael Perry
[EMAIL PROTECTED]
==================
------------------------------
From: Eric <[EMAIL PROTECTED]>
Subject: Re: windows VFAT partitions too fat?!
Date: Wed, 06 Dec 2000 09:53:42 +0100
Reply-To: [EMAIL PROTECTED]
> >> >disk /dev/hda: 255 heads 63 sectors 2498 cylinders
> >> >units = cylinders of 16065 * 512 bytes
> >> >
> >> >device boot start end blocks id system
> >> >
> >> >/dev/hda1 * 1 255 2048224+ b win95 fat32
> >> >partition 1 does not end on cylinder boundary:
> >> >phys=(1023,15,63) should be (1023,254,63)
> >> >
>
> The above tables are made using a 16 heads translation (end head is
> 15), while the usual BIOS setting for a disk this size would be LBA
> mode with a 255 heads setting.
I figured that this was what you meant
> The partition table CHS values however may be correct, if the BIOS is
> set to NORMAL mode. This can be the case, since the fdisk information
> may not be correct.
>
> Another way to verify that the tables are made using a 16 heads
> translation is to look at the number 2048224+ for hda1. This is
> 2*2048224+1 sectors = 4096449 sectors. In front of hda1 we usually
> have 63 sectors. Then (4096449+63)/16/63 = 4064 is a full number of
> cylinders.
I don't get this part, using 256 instead of 16 also gives an integer
> If the partition tables match a BIOS setting, but the BIOS setting
> currently is wrong, the solution is to redetect the disk in BIOS, and
> then manually set the correct mode if the mode still is wrong.
In the BIOS you mean?
Or with fdisk -> expert -> set CHS values?
> For programs that use the BIOS for reading and writing the disk, such
> as DOS and Lilo, the result of having wrong partition tables, or
> primary partitions that do not end at a cylinder boundary, is
> unpredictable. The reason for this is some subtle BIOS features.
Again I don't really understand this.
I would understand that starting boundaries must be cylinder aligned,
But ending cylinders I don't understand. Even if DOS fdisk would make
the mistake and
make a FS that extends beyond the partition boundary ( IIRC DOS always
assumes that partition boundaries are cylinder aligned), no severe error
would be made. As long as the next partition still starts at the cyl.
boundary, the FS on that partition would not be touched. Maybe DOS would
be unable to access the last few data sectors though?
Thanks for your time,
Eric
------------------------------
From: Eric <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux.mandrake,gnu.utils.help
Subject: Re: why can't one boot from /dev/hdb?
Date: Wed, 06 Dec 2000 10:01:21 +0100
Reply-To: [EMAIL PROTECTED]
> err, sorry, I just thought I might reduce the noise pollution. Thanks for
> caring.
> Mainly my discovories are [which might be hints for newbie helping docs]:
> A boot disk has two types:
> one that just contains the MBR
> one that has more --- a rescue disk
> one that has even more: a linux complete on a floppy... ok 3rd type.
>
> also I never expected it to be true that one could just
> edit menu.lst and not even run a command and it would affect things---true
> in some cases. This is diffrerent than lilo it seems. one still needs to run
> lilo.
> All these newbie things aren't really mentioned often.
>
> A question I have is "installing lilo/grub on windows hda or hda1"
> there are apparantly seveal degrees to such an installation.
> one would be just writing to the MBR of hda.
> another degree would entail more space on hda or hda1, but is there
> such space waiting or will a windows thingy get ruined in the process.
>
> I'm thinking the best thing would be if the boot process would not invold both
> hda hdb abd floppy. i.e., revove the windows disk and still can boot linux,
> remove linux disk and still boot windows----all with the minimal amount of
> reconfiguration.
>
>
> indeed, i'm sure if I finally give in and overwrite MBR of hda, all will be
> simple...
> I just thought I could get away with overwriting MBR of hdb for the same
> effect..
Now I see what you want to do :-)
Svend olaf mentioned the solution to that:
1) force hdb to 0x80 and hda to 0x81
2) Install LILO in the MBR of hdb and configure your BIOS to boot from
hdb
If that doesn't solve all the problems, (You still may have that
geometry mismatch)
report back to us.
Eric
------------------------------
From: "kitty" <[EMAIL PROTECTED]>
Subject: What is this message??How to solve??thanks!!
Date: Wed, 6 Dec 2000 16:50:32 +0800
HELLO
I HAVE A QUESTION.
MY LINUX(MAIL-SERVER) TO APPEAR : "Unable to load
interpreter".
WHAT IS THIS ERROR-MESSAGE ? HOW TO SOLVE IT THAT I DO'T KNOW ?
SO IF ANYONE COULD HELP PLEASE COULD YOU E-MAIL ME AT
[EMAIL PROTECTED]
THANKYOU VERY MUCH FOR ALL YOUR HELP!!!
------------------------------
** 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.setup) 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-Setup Digest
******************************
