Linux-Setup Digest #445, Volume #21 Fri, 15 Jun 01 08:13:09 EDT
Contents:
Re: ide-tape.o errors. Thank you!! (Leonard Evens)
Re: rc.local part 2. (Floyd Davidson)
what is the way to download whole the directory.... (v.nagasrinivas)
Re: how can i delete the files with ?-------- permissions set.. (H.Bruijn)
Re: bad format line XX ("Eric")
Re: what is the way to download whole the directory.... (H.Bruijn)
Re: install linux on one harddrive and windows on another? ("Eric")
Re: GUI Login at Console (Robert Davies)
Re: init.d problem - starting program from home directory (Robert Davies)
Re: Linux Server overload problem during peak hours (Robert Davies)
Make user settings similar to root setting (Jesper Petersen)
prob with install LinuxRedHat ("Kristijan Schmidt")
Re: Make modules fails with RH 7.1 (glyn jones)
Re: ide-tape.o errors (Robert Davies)
Re: New H/W, Disk Tuning and risks of corruption from tweaking with hdparm (Robert
Davies)
Re: Cannot open master raw device '/dev/rawctl' (No such device) (Robert Davies)
Re: How do i reinstall Lilo? (Robert Davies)
Re: Kernel 2.2, IP masq pool, possible? (Robert Davies)
Re: Buying a CDR/RW (Julian Bordas)
Re: 2.4.2 Kernel Compiling Problem ("Gene Heskett")
USB mouse ("Tozzy")
----------------------------------------------------------------------------
From: Leonard Evens <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.hardware,comp.os.linux.questions,redhat.general,comp.os.linux.misc
Subject: Re: ide-tape.o errors. Thank you!!
Date: Fri, 15 Jun 2001 02:59:30 -0500
Denis Leroy wrote:
> The solution is to use the device in SCSI emulation mode, it'll work just fine.
>
> modprobe ide-scsi
> modprobe st
>
> drive is now available as /dev/st0 and /dev/nst0
>
> -denis
The first time I tried it, I got a "no such device response".
I noticed that the ide-tape module was loaded, so I removed
it, but that didn't help. Finally, I removed the ide-scsi
and st modules and reloaded them. Then it worked. Or at
least I was able to do mt status and tar tvzf
Now I have to figure out the most effective way of making sure
the ide-tape module is not loaded to start and then to load the
other modules.
Although this is an effective workaround, someone should fix
the ide-tape module. It worked fine up to kernel 2.2.14.
--
Leonard Evens [EMAIL PROTECTED] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
------------------------------
From: Floyd Davidson <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.misc
Subject: Re: rc.local part 2.
Date: 14 Jun 2001 23:46:50 -0800
"Liverpool_fc" <[EMAIL PROTECTED]> wrote:
>hello,
>
>i put this line at the bottom of rc.local.
>cp /etc/termcap /home/cpu1/filename
>when i rebooted it did not do the copy.
>this is rh7.0.
You probably need to use the full path name for cp, because
most likely PATH is undefined at that point for that user.
That is a basic lesson to learn for all automatic scripts,
whether they are boot scripts or cron jobs. Do not assume
that your login shell environment is available when the
script is run.
--
Floyd L. Davidson <http://www.ptialaska.net/~floyd>
Ukpeagvik (Barrow, Alaska) [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (v.nagasrinivas)
Subject: what is the way to download whole the directory....
Date: Fri, 15 Jun 2001 08:30:58 +0000 (UTC)
Hi,
What is the way to download whole the directory,
through ftp service..
I observed the ftp.netbsd.org is allowing me by
giving get the directoryname.tar But I tried through ftp.kde.org
its not allowing me to do that...
what is the difference..and what is the feature in ftp ...
to download the directory.......
thanks,
srinivas.
--
v.nagasrinivas
YVL Software Consultancy
b4,q1,6th floor,CyberTowers,
Madhapur,Hi-Tech City,
Hyderabad-500033
AndhraPradeshState,
INDIA.
ph:091-040-3110200
visitme:http://www.nagasrinivasv.com
--
Posted from ns.stph.net [196.12.32.2]
via Mailgate.ORG Server - http://www.Mailgate.ORG
------------------------------
From: [EMAIL PROTECTED] (H.Bruijn)
Subject: Re: how can i delete the files with ?-------- permissions set..
Date: 15 Jun 2001 08:56:08 GMT
Reply-To: [EMAIL PROTECTED]
On Fri, 15 Jun 2001 05:58:12 +0000 (UTC), v.nagasrinivas allegedly wrote:
> Hi,
>
> while installing mountsmbfs, i forgot to copy mount.smbfs
> and tried to mount of windows share. It created a dummy
> mount.smbfs with ?--------- file permissions and not allowing me
> to delete.. How can i delete this one..
> Or how can i overwrite the correct file onto this...
Change the permissions with chmod and then do your thing.
chmod o+rw file.name gives the owner read+write permissions.
chmod 600 file.name also gives the owner read+write permissions.
--
If a trainstation is the place where trains stop, what is a workstation?
========================================================================
Herman Bruijn website: http://HermanBruijn.com
The Netherlands
------------------------------
From: "Eric" <[EMAIL PROTECTED]>
Subject: Re: bad format line XX
Date: Fri, 15 Jun 2001 11:06:21 +0200
> >after editing /etc/fstab I get a boot -up message that says -
> >'WARNING bad format line 29 /etc/fstab'
> >I'm using redhat 7.0 and kde. I've edited this file with vi and kedit
> >but still get this message. Computer is not effected. Please help.
>
> Thanks for your interest. There is nothing on line 29 /etc/fstab.
> I edited the file when adding another hard drive. The file just has the
usual
> lines (8) starting with:
> LABEL=/ .......
> /dev/cdrom ......
> /dev/fd0 .....
> none ......
> none......
> /dev/hdc5 ....
> /dev/hdc1 .....
> /dev/hda2 ....
If your fstab looks like this, it lacks several columns.
> The kernel seems to recquire some special type of text format that I don't
> know.
*show* the file, not some manually edited version of it.
You still have given only useless information.
And please, cut-and-paste, don't retype it in you message either.
Eric
------------------------------
From: [EMAIL PROTECTED] (H.Bruijn)
Subject: Re: what is the way to download whole the directory....
Date: 15 Jun 2001 09:09:18 GMT
Reply-To: [EMAIL PROTECTED]
On Fri, 15 Jun 2001 08:30:58 +0000 (UTC), v.nagasrinivas allegedly wrote:
> Hi,
> What is the way to download whole the directory,
> through ftp service..
>
> I observed the ftp.netbsd.org is allowing me by
> giving get the directoryname.tar But I tried through ftp.kde.org
> its not allowing me to do that...
That is a serverside feature, indepenadant of the ftp client. Generally
thats only possible when the file ftp://some.host/bin/tar exists, or
when it is announced in the login message. It is not a standard feature.
> what is the difference..and what is the feature in ftp ...
> to download the directory.......
Most ftp-clients support sometimes "get *" IIRC they have to support
the mget command to get multiple files, so "mget *" should work as
desired, although the most simple ftp clients need you to agree to
downloading each file.
Some clients support a -r flag to get directories recursively.
ncftp / > help (m)get
(m)get: fetches files from the remote host.
Usage: get [-flags] file1 [file2...]
Flags:
-R : Recursive. Useful for fetching whole directories.
-z : Get the remote file X, and name it to Y.
-a : Get files using ASCII mode.
-A : Append entire remote file to the local file.
-f : Force overwrite (do not try to auto-resume transfers).
Examples:
get README
get README.*
get "**Name with stars and spaces in it**"
get -R new-files-directory
get -z WIN.INI ~/junk/windows-init-file
--
If a trainstation is the place where trains stop, what is a workstation?
========================================================================
Herman Bruijn website: http://HermanBruijn.com
The Netherlands
------------------------------
From: "Eric" <[EMAIL PROTECTED]>
Subject: Re: install linux on one harddrive and windows on another?
Date: Fri, 15 Jun 2001 11:09:03 +0200
> : Install Mandrake and very carefully read the install questions
especially
> : at the time of configuring lilo. Put the Windoze disk as the primary
> slave
> : (/dev/hdb) and make the Linux disk the primary master (/dev/hda).
> : Mandrake's installer will configure a proper lilo.conf for you and you
> will
> : be able to boot either Linux or Windoze.
>
> I had a bit of a problem there. You should first try if this setup
described
> above works or not.
You should have used the map-drive command.
(See posting of Robert Davies in this thread)
Eric
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: GUI Login at Console
Date: Fri, 15 Jun 2001 10:32:02 +0100
Paul Crabb wrote:
> I am currently running Redhat 7.1 and have the X server configured and
> running properly. I am able to login via a text login and do a startx
> to get the GUI to come up.
>
> My question: How do I configure the init scripts to start the GUI
> login manager and bring the GUI up as a default?
In /etc/inittab the default runlevel is selected. You probably need to
change the 3 to a 5 in there, there should be comments to tell you what
runlevel does what.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: init.d problem - starting program from home directory
Date: Fri, 15 Jun 2001 10:27:50 +0100
Mark Grimshaw wrote:
> Is there any way to get the init.d script to start to start sc_serv
> from within its home directory?
Yes, cd(1).
I'm wondering why you're not using the standard start daemon function
though, that most of the init.d scripts use.
Rob
> #######################
> #!/bin/sh
> #
> # Startup script for the Shoutcast mp3 Server
> #
> # chkconfig: 345 85 15
> # description: Shoutcast mp3 server
> # processname: sc_serv
> ## pidfile: /var/run/sc_serv.pid
> # config: /usr/local/shoutcast/sc_serv.conf
>
>
> # Source function library.
> . /etc/rc.d/init.d/functions
>
> # See how we were called.
> case "$1" in
> start)
> echo -n "Starting shoutcast: "
> /usr/local/shoutcast/sc_serv /usr/local/shoutcast/sc_serv.conf
(cd /usr/local/shoutcast; ./sc_serv /usr/local/shoutcast/sc_serv.conf)
&
> echo
> touch /var/lock/subsys/sc_serv
> ;;
> stop)
> echo -n "Shutting down shoutcast: "
> killproc sc_serv
> echo
> rm -f /var/lock/subsys/sc_serv
> # rm -f /var/run/sc_serv.pid
> ;;
> status)
> status sc_serv
> ;;
> restart)
> $0 stop
> $0 start
> ;;
> reload)
> echo -n "Reloading shoutcast: "
> killproc sc_serv -HUP
> echo
> ;;
> *)
> echo "Usage: $0 {start|stop|restart|reload|status}"
> exit 1
> esac
>
> exit 0
>
> ################################
>
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: Linux Server overload problem during peak hours
Date: Fri, 15 Jun 2001 11:01:30 +0100
Michael Heiming wrote:
> Ken Fung wrote:
>>
>> I got a Linux server with kernel 2.0.36, 128MB RAM, 128MB Swap,
>> PII 450. It's running Apache, sendmail, ipopd, telnetd, ftpd,
>> MySQL, PHP4 and Perl, mainly Apache, MySQL and PHP only.
I had a web server with 2.0.36 kernel, that used to suck mud like this. In
it's case I thought it was because of the low simultaneous process limit,
with a front end webserver generating more proxy pass thru connections than
the older machine could handle.
So what is going on when it starts to struggle?
Can you run vmstat(8) every 10 seconds with 'vmstat 10' and have it go to a
log on a different machine? Then you can see if it's failing due to
thrashing.
What about the number of processes? Do you have a /proc, you could do
something like 'echo /proc/[0-9]* | wc -l' to see whether you get near the
process limit.
There's an include file in /usr/src/linux with a definition of NR_TASKS,
you might need to increase it.
I don't think 128MB of ram is that much, if it's a busy web site, it could
simply be thrashing.
That said, my impression was that a 2.2 kernel is better than 2.0,
presumably upgrading is a problem or you would have done it long ago.
If you have a lot of static images or pages, you might consider serving
them in a less resource intensive manner than apache. Some use the squid
proxy as an HTTP accelerator, it uses less processes, with a threaded
design to serve requests. This will not help, if all your content is
dynamic, though you might be able to hive off images.
>> This server handling a very popular web site. During the
>> beginnnig of the peak hours everyday (5pm-9pm), the loading
>> will raise from around 1 to 5. At 9pm-1am is the busy time.
>> The loading always around 7.
>>
>> Sometimes, the loading suddenly raise from 8 to 150 in only
>> 2 minutes. The highest loading I saw was 250. The server
>> was hanged up and extremely slow. The login prompt even couldn't
>> show up. If I was logged in already, it said "Out of memory!"
>> after I typed any command 3-5 minutes ago.
>
> What do you mean with loading? load average?
He means what linux does, the number of processes that are runnable in the
scheduler runqueue, averaged out over some time sample interval.
>> The loading will back to normal after the peak hours at around
>> 2am.
>>
>> I was using "top" to monitor the server when this problem
>> happened every night. The memory and the swap spaces were
>> ate up by something.
>
> Something? If you run top, you should be able to see which process
> take the RAM.
>
>> The number of "httpd" processes raise
>> from 15 processes to more than 50. "mysqld" also increased
>> from 3 processes to more than 15 processes.
>
> That's normal and no problem, perhaps raising the min. spare server
> apache
> should start, would make it more responsible.
>
>> The file-max is 4096, inode-max is 12288 and freepages are
>> 300 400 500. Is that a problem?
>>
>> Does anyone got the idea what is going on? This is a commercial
>> server so it's unreasonable to have no service during these
>> hours. Actually this problem occured almost every night for
>> 2 months, but I have no idea what's the problem.
>
> Check /etc/crontab and users tabs for something that runs at the
> same time and shouldn't.
>
>> This is the data from "top", hope this helps.
>>
>> "top":
>>
>> 6:39pm up 28 min, 1 user, load average: 2.71, 2.30, 1.58
>> 69 processes: 61 sleeping, 7 running, 1 zombie, 0 stopped
>> CPU states: 21.2% user, 13.0% system, 2.2% nice, 66.8% idle
>> Mem: 119636K av, 105260K used, 14376K free, 91076K shrd, 28656K buff
>> Swap: 130748K av, 4K used, 130744K free 32252K
>> cached
>>
>> PID USER PRI NI SIZE RSS SHARE STAT LIB %CPU %MEM TIME
>> COMMAND
>> 958 nobody 9 0 3288 3288 2164 R 0 5.5 2.7 0:10 httpd
>> 362 nobody 4 0 2944 2944 2160 R 0 2.7 2.4 0:14 httpd
>> 370 nobody 2 0 2956 2956 2164 S 0 2.7 2.4 0:08 httpd
>> 385 nobody 4 0 0 0 0 Z 0 2.7 0.0 0:14 httpd
>> <zombi
>> 315 nobody 2 0 3068 3068 2176 R 0 2.5 2.5 0:11 httpd
>> 316 nobody 4 0 2912 2912 2172 S 0 2.5 2.4 0:13 httpd
>> 360 nobody 3 0 2968 2968 2176 R 0 2.5 2.4 0:16 httpd
>> 465 nobody 2 0 3024 3024 2180 S 0 2.3 2.5 0:14 httpd
>> 369 nobody 6 0 3364 3364 2180 R 0 2.3 2.8 0:23 httpd
>> 936 nobody 3 0 2972 2972 2156 S 0 2.3 2.4 0:09 httpd
>> 2274 root 16 5 2124 2124 1332 R N 0 2.1 1.7 0:00 mysqld
>> 358 nobody 2 0 3108 3108 2160 S 0 1.5 2.5 0:10 httpd
>> 2163 root 3 0 748 748 544 R 0 1.3 0.6 0:02 top
>> 352 root 5 5 2124 2124 1332 S N 0 0.1 1.7 0:00 mysqld
>> 317 nobody 0 0 2944 2944 2172 S 0 0.1 2.4 0:15 httpd
>> 340 root 0 0 228 228 188 S 0 0.1 0.1 0:00 update
>> 1 root 0 0 376 372 316 S 0 0.0 0.3 0:03 init
I think this output is useless, as you need it from when the system is
busy. Better to use 'ps aux' to a log file every 60 seconds or so, so you
can see what's going on when the system seems to get busier.
>>
>> Ken
>
> Are you running lots of cgi/dynamic pages? I seldom saw a single apache
> process with more than 0.1% CPU.
>
> Top shows only very little swap usage, that's kind of curious, as you
> wrote
> above the machine would start swapping? And the CPU is 66.8% idle?
>
> Case it's swapping, put some more RAM in, Linux should never ever swap.
A little paging never hurt anyone, linux never does 'swap' the VM system
does not have the ability to swap out all the pages of a process. It
merely steals pages when it tries to free some memory, which may or may not
involve writing the dirty pages first to disk, in the swap partition.
What happens if his bandwidth is swamped? Folk then it hit reload then,
making the problem worse.
Rob
------------------------------
From: Jesper Petersen <jesperp**Delete**@nork.auc.dk>
Subject: Make user settings similar to root setting
Date: Fri, 15 Jun 2001 12:50:24 +0200
Hi!
I just changed from Mandrake 8.0 to Red Hat 7.1 reusing the kernel. When I
start Emacs as root I get a green background with yellowcharacters (which I
like), but when I start Emacs as a normal user I just get black characters on
a white background.
In order to get the green Emacs I have to start X as root, it's not enough to
"su -" when I'm logged in as user.
A similar thing happens when I start an Xterm, as root I get a black
background and as a normal user I get a white one.
Does anyone have an idea how I make the root settings default for normal
users?
Doesn't know if it has anything to say, but to make X work for normal users I
did the following:
chown root /usr/X11R6/bin/X
chmod -s /usr/X11R6/bin/X
chown root /etc/X11/X
chmod -s /etc/X11/X
Jesper
------------------------------
From: "Kristijan Schmidt" <[EMAIL PROTECTED]>
Subject: prob with install LinuxRedHat
Date: Fri, 15 Jun 2001 13:03:50 +0200
the partion prog says boot -> 1024
on c: and on d: is w2kpro
I wanted to install RedHat on e: which is 5GB big.
I made a swap partition. I cant make a boot and / or a root part. it says
boot -> 1024
obviously its the win boot!
Can you help me and/or understand me?
Greetings from Germany
------------------------------
From: glyn jones <[EMAIL PROTECTED]>
Subject: Re: Make modules fails with RH 7.1
Date: Fri, 15 Jun 2001 12:34:07 +0100
John Grana wrote:
> I am trying to re-build my 2.4.X Red Hat 7.1 kernel. I needed to patch it
> with the Win4Lin kernel patch...
> I can compile and boot the kernel just fine. The problem starts at trying
> to make modules. The first module it hits (dummy.c for example) that
> includes /usr/src/linux/include/linux/module.h fails with a stream of
> strange errors. I don't have the exact text in front of me, but things
> like line 173 parse error, not a number and not hexadecimal, function
> declared that returns a function, things like that.
>
> I can vi the module.h file and the lines look fine. No strange looking
> characters or numbers....
>
> BTW, the reason for re-building the kernel is to make the NVdriver.o file
> (that also gets the errors since it is a loadable module as well!).
>
> Any pointers?
>
> Thanks
> John Grana
> [EMAIL PROTECTED]
>
>
>
>try typing mrproper before you do anything else
worked for me
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: ide-tape.o errors
Crossposted-To: comp.os.linux.hardware,comp.os.linux.questions
Date: Fri, 15 Jun 2001 12:16:46 +0100
Leonard Evens wrote:
> Denis Leroy wrote:
>>
>> Ga Mu <[EMAIL PROTECTED]> wrote in message
>> news:<[EMAIL PROTECTED]>...
>> > HELP!
>> >
>> The solution is to use the device in SCSI emulation mode, it'll work just
>> fine.
>>
>> modprobe ide-scsi
>> modprobe st
>>
>> drive is now available as /dev/st0 and /dev/nst0
>>
>> -denis
>
> I tried this, but attempts to access the device through mt or
> tar respond that there is no such device. Of course, there is
> such a device, but something can't find it. Any suggestions?
Yes, you need to pass some options on boot in your lilo append line.
To set a drive hdc to scsi emulation I have :
append = "reboot=warm hdb=ide-scsi hdc=ide-scsi"
This is needed to tell the IDE drivers that they don't 'own' this drive.
I had to put the modprobe ide-scsi in a file called 'boot.local' which is
run before the rc?.d stuff.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: New H/W, Disk Tuning and risks of corruption from tweaking with hdparm
Date: Fri, 15 Jun 2001 12:23:31 +0100
Michael Perry wrote:
> On Wed, 13 Jun 2001 10:49:11 +0100, Robert Davies
> <[EMAIL PROTECTED]> wrote:
>
> Also one could not go very wrong by looking at the VIa ide drivers
> contained
> in the 2.4.5 kernel. The driver file mentions the versions of the VIA
> chipset/southbridge that the driver now works with. I have been running
> this driver on my systems with the KT133 southbridge stuff for awhile now
> and have had quite good results.
Now that is interesting! I'm hoping the VM will be fixed in 2.4.6 now
they've added various patches, on that issue.
It has kept me on 2.2, though I run SMP and would benefit from 2.4.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: Cannot open master raw device '/dev/rawctl' (No such device)
Date: Fri, 15 Jun 2001 12:19:03 +0100
Sean Marshall wrote:
>> So I was right, now go and get a RH 2.2.19 kernel rather than a Linus
>> one. Unless sanargy needs patches applied that come with the software
>> (unlikely) then it should work.
>>
>> If you previously used a RH binary generic kernel, stick to it, you'll
>> not get much from compiling your own kernel, and it'll save you lots of
>> time. 5 minute upgrade to new version, with rpm, and running lilo.
>
> The precompiled kernel isn't very efficient nor does it see all my RAM
> either. I need a custom kernel I'm afraid.
My tests showed minimal differences, to compiling in drivers, I did use a
binary kernel of right CPU architecture however.
For RAM, you can set it in append lines, newer kernel versions are much
better at recognising correct amount correctly however.
Is it because you have really shed loads of RAM, and need the large memory
models, >1GB?
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: How do i reinstall Lilo?
Date: Fri, 15 Jun 2001 12:33:11 +0100
David wrote:
> mj topper wrote:
>>
>> I'm new to linux. Is it possible to reinstall the boot
>> selector. I've got win98 on C:\ and redhat 6.2 on D:\ so
>> far the only way to boot linux is from the "floppy".
>> Thanks
>> Mike
>
>
> As root run the command "lilo" if the /etc/lilo.conf file is configured
> properly.
Before messing with lilo, make sure you have a working boot floppy,
personally I think it's best to have lilo in MBR, and use it for booting.
The original questioner, can keep his M$ boot code in C: 's MBR, and boot
off the linux disk.
This lilo fragment shows how to boot Windows off another disk.
image = /boot/vmlinuz
label = linux
root = /dev/hda6
initrd = /boot/initrd
append = "noapic hdb=ide-scsi hdc=ide-scsi apm=poweroff reboot=warm"
other = /dev/hde2
label = windows
map-drive = 0x80
to = 0x81
map-drive = 0x81
to = 0x80
table = /dev/hde
To boot off D: keeping windows as C: you would need to alter the BIOS to
bood D:, and then have lines like :
# Keep things straight after booting from different disk with bios.
disk=/dev/hdb
bios=0x80
disk=/dev/hda
bios=0x81
I haven't tried booting off of another drive in BIOS with lilo eg) D: and
then booting the Windows on the first disk. It shouldn't be a problem
though.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: Kernel 2.2, IP masq pool, possible?
Crossposted-To: comp.os.linux.networking,alt.os.linux.mandrake
Date: Fri, 15 Jun 2001 12:44:56 +0100
Sean McAvoy wrote:
> Hello.... just wondering if the following is possible under 2.2.... I need
> to have "pool"of IPs availible for masqing only when communicating with
> a specific subnet. If the packet is not going to the special subnet then
> it doesn't get one of those masq's, it get's the "default". But when
> someone opens a connection to the subnet, they need to "lock" the ip
> until they disconnect (Ideally users would get connection refused
> or the like). Any Ideas? I know this should be more than possible under
> 2.4, but I am working with embeded systems (x86 based), and 2.2 is the
> only kernel that's been ported to them and tested....
I don't think it is possible with 2.2 masquerading, for the simple reason,
that the NAT that masquerading does, makes the connections look like
they're coming from the masq-ing machine.
Your specs sound broken to me, whoever came up with that scheme?
I think the way you _could_ do it is with an application level gateway, and
use IP aliases, then you can do the locking issues and IP address admin in
user space.
Rob
------------------------------
From: Julian Bordas <[EMAIL PROTECTED]>
Subject: Re: Buying a CDR/RW
Date: Fri, 15 Jun 2001 22:06:05 +1000
Stanislaw Flatto wrote:
> So get SCSI 8x2x2 and you won't regret it.
> Just my opinion.
> NB writing from hard disk to CD writer on the same IDE interface creates
> traffic jams, caveat emptor!
>
> Stanislaw.
>
>
The HD is on a different interface to the CDR. Do you know of any
suppliers of SCSI CDR/RW ?
BTW Wheere is Ulla Dulla?
Julian
------------------------------
Date: 15 Jun 2001 7:47:38 -0500
From: "Gene Heskett" <[EMAIL PROTECTED]>
Subject: Re: 2.4.2 Kernel Compiling Problem
Crossposted-To: comp.os.linux.misc
Gene Heskett sends Greetings to Skylar Thompson;
ST> I am having some trouble compiling the 2.4.2 kernel. Whenever I try to
ST> compile it
ST> (make dep;make clean; etc.), it displays: .hdepend:1493 Unterminated
ST> variable reference. I have tried getting the source again from the CD, to
ST> no avail. Is there any way to fix this problem, or do I have to go back
ST> to the 2.2.18 kernel?
ST> -- Skylar Thompson ([EMAIL PROTECTED])
I stay up on kernels and ISTR I saw that in one of the 2.4.2 updates.
It wasn't something I fellt I could fix, so I just waited for the next
2.4.2-ac## patch, which did fix it.
I'd do two things.
1. goto rawhide and get the newest revision number of the whole gcc
family, this included libstdc++ and cpp. Its much more stable. Its
about 6 files to get it all.
2. Goto kernel.org and get the basic 2.4.5 sources, and maybe even the
-ac14 patch for it. The reasoning being that since 2.4.2, huge amounts
of the code have been subjected to the Stanford Code Verifier, catching
the things that USED to cause the compiler to output warnings by the
megabyte. The whole thing builds with amazingly few of those going by
today.
Cheers, Gene
--
Gene Heskett, CET, UHK |Amiga A2k Zeus040, 70MB ram, 31 gigs
| Linux @ 500mhz, 320MB ram, 50 gigs
email gene underscore heskett at iolinc dot net
#Amiga based X10 home automation program EZHome, see at:#
<http://www.iolinc.net/gene_heskett>
This messages reply content, but not any previously quoted material,
is © 2001 by Gene Heskett, all rights reserved.
--
------------------------------
From: "Tozzy" <[EMAIL PROTECTED]>
Subject: USB mouse
Date: Fri, 15 Jun 2001 14:00:37 +0200
Hello,
Is it posible to use a USB mouse under suse6.2?
I would like to use my USB Microsoft Itelliemouse Explorer.
thanx
------------------------------
** 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
******************************
