Linux-Misc Digest #933, Volume #27 Thu, 24 May 01 03:13:01 EDT
Contents:
Re: linux libraries (Richard)
Re: Network Configuration mystery ("Sony Antony")
Re: fdisk help please (David Efflandt)
Re: rh 7.1 driving me nuts (Dave Uhring)
Re: linux libraries (Dave Uhring)
autoloader ("Patrick Chan")
Re: rh 7.1 driving me nuts (Jacob Kristensen)
Re: cdrecord fails only with 2.4.x kernel ("Rob")
xmms on kde (mndoci)
Re: ziplinux ("Xoran99")
LILO limit still there? ("Vladimir Florinski")
Re: Swapping on RAID1 ("Support")
Re: Dumb newbie question about setting the system time. (Martin Brown)
Re: Squid 2.4 and Redhat 7.1 (Neil Goldberg)
Strange Question.. ("Kevin from Hamilton")
Re: Test don't reply ("Xoran99")
----------------------------------------------------------------------------
From: Richard <[EMAIL PROTECTED]>
Subject: Re: linux libraries
Date: Thu, 24 May 2001 03:14:58 GMT
Richard wrote:
>
> I am running RH 7.1 at work and here at home. I have had a problem with
> a couple differnt programs trying to install or compile from source or
> from rpm. It is asking for libcrypto.so.0 and libssl.so.0 .
>
> Now i have both of these files in /usr/lib. And also have the latest (i
>
> think) of these files in there, libcrypto.so.0.9.6 and
> libssl.so.0.9.6. I have tried adding /usr/lib to my path also. I am
> still quite the newbie with linux but i can usually figure these things
> out and this one is stumping me good. I cannot figure out how to make
> it read these files and know that they are in there. Any suggestions
> would be great
>
> rich
Also this is the exact error i am getting if it will help any.
error: failed dependencies:
libcrypto.so.0 is needed by licq-base-1.0.3-1
libssl.so.0 is needed by licq-base-1.0.3-1
------------------------------
From: "Sony Antony" <[EMAIL PROTECTED]>
Subject: Re: Network Configuration mystery
Date: Wed, 23 May 2001 23:43:01 -0400
Hi Peter:
thanks a real lot for the detailed explanation.
Yes the problem was that I had not setup the route to the default gateway
When I did the following ( put them in local.rc ), it got fixed.
route add -host 63.96.44.87 dev eth0
route add default gw 63.96.44.87
Essentially I created a static route to the IP address through eth0 and
*then* specified that IP as the def gateway.
In My redhat distribution using netcfg..
1. I had to add a new route to the network 63.96.44.87 with netmask
255.255.255.255 inorder to create the route to the host. In other words I
had to use -net ( I thought I was doing a hack due to poor interface )
instead of -host. Does anybody know if it s possible to give the equivalent
of route add -host IP eth0, using netcfg.
2. I could not find a way to specify this IP as the default gateway at all.
There was an option to give the default gateway and it s device, which did
not work when I gave the above IP.
Obviously I m missing something. If somebody could pl help.
More questions below...
> >
> > But I could not access the internet. Neither could I ping the any
> > addresses outside our network.
> > Then when I checked one of the office windows machine, I saw that they
all
> > have an IP address **outside our network** as the default gateway.
>
> outside of network ? if they have a public IP the whole world is the
> network ..
I said outside the network by looking at our subnet mask. The default
gateway IP was very different.
Why do you say that the whole world is the network.
As far as I understood..
1. They have the def gateway in the same ethernet network This def gw is the
gate to one or more higher level ( broader - bigger subnet mask ) networks
( which in turn will have more gateways and routers )
2. Our office network is actually a part ( I m guessing ) of a bigger
ethernet network which they own. They have many small networks like our
office ( with small subnet masks ) running in the same physical ethernet
network. ( In this case I should be able to snoop at other company s
traffic ). The default gateway they have given us is a machine sitting in
the same physical ethernet network.
Could my theory be correct. Can you see any flaw in this.
>
> usual (and deadly) setup offered by some ISP:
>
>
> internet,gateway----HUB------ client1
> ------ client2
> ------ clientx
If client1, client2 means small subnets for each customer, this is the same
as what I was describing.
>
> needless to say that this is an security nightmare. Note that all clients
> sit in the same networksegment and the gateway does not.
Why everything connected to the HUB forms the network segment ( meaning
physical ethernet network ) right.
Also the only security problem I see is that every company can see every
other company s ( served by the same ISP ) data.
Are you concerned that anybody from the world can intride the network
because there is no firewall.
> My ISP offers such solutions and when I set my nic to promiscous mode I
see
> a lot of stuff floating by that does not belong there ... :)
>
>
> You can also have the following:
>
> internet, isp-gate------router-------client1
> -------client2
> -------client3
>
> where the router can also be a firewall or whatever. In that case the
> clients have the router as gateway or (if bad configuration) the
> isp-gateway.
>
> In that case the router must allow the clients connections to the internet
> and maybe this was your problem.
>
> > Even
> > this IP I could not ping from my Linux box. Needless to say that setting
> > this as default gateway in my Linux box did not solve the problem.
>
> Not needless :) You are sure, all routes were set proper ?
I wasn t when I set it as default gateway, I did not have a route to the
default gateway. Even my setting as default gateway I dont think was proper.
>
> > My understanding about default gateway was
> > 1. This is the IP address the request should be sent if the host does
not
> > know of any route to a destination.
>
> no, this is the machines where packages are sent to next. In other words:
> every route from inside to outside goes via gateway.
Suddenly I remembered one confusion I had from the school days.
1. Because some machine is set as def gateway, if my m/c is asked to contact
an unknown ( no routs ) destination, will it do something like a NAT and
form a packet whose dest address will be that of def gateway. Then def
gateway will forward the packet after changing the dest to the actual dest
( if it can deliver directly - or to that of the next gateway in case of
next hop ). Similarly doing the reverse when the reply comes. If this is
true, there is extra expense in sending a packet through a def gateway. ( I
m almost sure I m wrong here )
OR
2. Will it do a broadcast for the unknown dest address, to which the m/c
designated as the def gateway will respond. Then the originator will
send/receive packet as if the def gateway is the actual dest. Had this been
true there is no packing/unpacking expense for the originator. But there is
no need for it to know who is the def gateway either, since it broadcasted
it in the ethernet network ( for which the def gateway responded )
Maybe there is a 3rd and correct way I dont know. Does anybody know.
>
> > 2. The defaul;t gateway should always be a member of the network.
> >
>
> iit should but it need not. you can set routes per interface or per
> gateway. If set per gateway the specified gateway needs not to be
reachable
> via one hop.
This is the route add -net way of doing it right. Essentially you will set
the first hop gateway first ( has to be in the same ethernet network ). Then
the net hop gateway ( which will be passing through the first one ) so on
and so forth right.
>
> > Also can anybody see how the windows machines were able to work
> > mysteriously. The only other configuration I could see was for DNS (
which
> > was also pointing to the same machine as the one set as default
gateway -
> > an outside the network host )
> >
>
> I think you have one of the above configurations but of course there are
> some other mystical things that can take place here.
> I would recommend reading the network-howto.
Thanks you once again for the patience.
--sony
------------------------------
From: [EMAIL PROTECTED] (David Efflandt)
Crossposted-To: alt.os.linux.mandrake,comp.os.linux
Subject: Re: fdisk help please
Date: Thu, 24 May 2001 04:03:44 +0000 (UTC)
Reply-To: [EMAIL PROTECTED]
On Wed, 23 May 2001 20:31:00 -0400, Dale Winters <[EMAIL PROTECTED]> wrote:
> Am trying to change the partition Id on one of my linux part. I tried
> using sfdisk with...... /sbin/sfdisk -c /dev/hda2 85 [83] --force....
> 85 is the Id it now has and I want to change it to 83. I tried lots of
> other combo's but this came the closest to working.I get an error msg
> something like "no partition 85".
> Can someone please help with the correct syntax ?
You are telling sfdisk to change partition number 85 on hard drive hda2
(not a drive) to type "[83]", which with brackets is not a valid type.
When you see brackets in an option, it means the parameter is optional, do
not include the actual brackets.
However, you are attempting to change an extended partition to a primary.
Does it contain any logical partitions that need to be removed first?
Are you aware that Linux has man pages that explain commands?
'man sfdisk' shows you should use: sfdisk --change-id /dev/hda 2 83
Is there some reason you don't use 'fdisk' (per your subject), which might
be easier (are you doing this from a shell script or something)?
> Does it matter if this part. is set active or not ?
Probably not, but I doubt if it is active if it is currently extended.
--
David Efflandt (Reply-To is valid) http://www.de-srv.com/
http://www.autox.chicago.il.us/ http://www.berniesfloral.net/
http://cgi-help.virtualave.net/ http://hammer.prohosting.com/~cgi-wiz/
------------------------------
From: Dave Uhring <[EMAIL PROTECTED]>
Subject: Re: rh 7.1 driving me nuts
Date: Wed, 23 May 2001 23:06:18 -0500
alik blochin wrote:
> how exactly do i do it?
> "Dave Uhring" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
>> alik blochin wrote:
>>
>> > i am going just crazy... guys :
>> >
>> > i am trying to install rh 7.1 for god knows what time
>> > and still without success :
>> >
>> > the installation gets stacked at the point of :
>> > "transferring install image to harddrive"
>> >
>> > and it gives the error
>> > "it was an error, maybe you haven't got enough free space"
>> > or something like that but that's not true:
>> >
>> > i have a 14 GB allocated just for linux
>> > 512MB swap
>> > / 7GB
>> > /usr all the rest
>> >
>> >
>> > i tried all the option available :
>> > disk druid
>> > fdisk
>> > with all options with those tools
>> >
>> > nothing just helps
>> > the system hangs on the point of
>> > "transferring install image to harddrive"
>> >
>> > it's driving me nuts....
>> > what the hell i am suppose to do to get it work !!!!?????????
>> >
>> > I know that if i turn to Red Hat
>> > god knows when i'll get an answer....
>> >
>> > if somebody knows something please help..
>> > i need to install it...
>> >
>> > and i don't want to mess with mandrake
>> > mandrake just freezes on my machine...
>> >
>> > Configuration:
>> >
>> > 900mHZ AMD
>> > Chaintech motherboard 7AJA
>> > 20GB WD +
>> > 30GB Quantum : 14GB for linux
>>
>> Drop your memory bus speed from 133MHz to 100MHz for the install.
Look, if you can't even get your mail address right so that I could respond
to your personal e-mail, then it's a waste of time to try to tell you a
second time. Your mail server does not recognize your name.
------------------------------
From: Dave Uhring <[EMAIL PROTECTED]>
Subject: Re: linux libraries
Date: Wed, 23 May 2001 23:09:06 -0500
Richard wrote:
> Dave Uhring wrote:
>>
>> Lack Mr G M wrote:
>>
>> > In article <[EMAIL PROTECTED]>, Richard <[EMAIL PROTECTED]>
>> > writes:
>> > |>
>> > |> I am running RH 7.1 at work and here at home. I have had a problem
>> > |> with a couple differnt programs trying to install or compile from
>> > |> source or
>> > |> from rpm. It is asking for libcrypto.so.0 and libssl.so.0 .
>> > |>
>> > |> Now i have both of these files in /usr/lib. And also have the
>> > |> latest (i
>> > |> think) of these files in there, libcrypto.so.0.9.6 and
>> > |> libssl.so.0.9.6.
>> >
>> > How did they get there? These are from OpenSSL.
>> >
>> > |> I have tried adding /usr/lib to my path
>> >
>> > That will have no effect. LD_LIBRARY_PATH is what affects dynamic
>> > library finding, and /usr/lib is going to be searched anyway. However,
>> > if you haven't rebooted the system since adding them you might neeed to
>> > run:
>> >
>> > /sbin/ldconfig -v
>> >
>> > as root.
>> >
>>
>> Red Hat 7.1 - XFS-SGI system, almost everything installed:
>>
>> /usr/lib/libcrypto.a
>> /usr/lib/libcrypto.so
>> /usr/lib/libcrypto.so.0.9.6
>> /usr/lib/libcrypto.so.1
>> /usr/lib/libcrypto.so.0.9.5a
>> /usr/lib/libcrypto.so.0
>>
>> /usr/lib/libssl.a
>> /usr/lib/libssl.so
>> /usr/lib/libssl.so.0.9.6
>> /usr/lib/libssl.so.1
>> /usr/lib/apache/libssl.so
>> /usr/lib/libssl.so.0.9.5a
>> /usr/lib/libssl.so.0
>
>
> Exactly. I have all of those files in those locations. Today i got it to
> work for me at work by making symbolic links in the /lib directory.
> Still cant get it to do it at home. Still get a dependency problem when
> i try to run the rpm. Any other ideas?
>
> rich
>
/etc/ld.so.conf
/usr/X11R6/lib
/usr/lib
/usr/kerberos/lib
/usr/i486-linux-libc5/lib
/usr/lib/gconv
This one is from my RH6 install where I am now. Then execute "ldconfig".
------------------------------
From: "Patrick Chan" <[EMAIL PROTECTED]>
Subject: autoloader
Date: Thu, 24 May 2001 11:22:56 +0800
I have Compaq ML530 and a Compaq SCSI 20/40 GB autoloader.
How can I use the autoloader?
Do I need to use the driver for the autoloader?
The CD accompanied with the autoloader does not contain driver for Linux.
I also can't find the driver over the Internet.
Or I can use it without a driver?
------------------------------
From: Jacob Kristensen <[EMAIL PROTECTED]>
Subject: Re: rh 7.1 driving me nuts
Date: Thu, 24 May 2001 06:15:32 +0200
alik blochin wrote:
>
> i am going just crazy... guys :
If ur BIOS got "virus-somethign" turned on, the installprogram migth not
be able to write to the MBR.
Jacob
--
There he goes... One of Gods own prototypes.
A high-powered mutant of some kind,
never even considered for mass production.
Too weird to live, and too rare to die.
------------------------------
From: "Rob" <[EMAIL PROTECTED]>
Subject: Re: cdrecord fails only with 2.4.x kernel
Crossposted-To: comp.os.linux.hardware
Date: Thu, 24 May 2001 06:23:15 +0200
In article <9ehghp$nt9$[EMAIL PROTECTED]>, "Chris Sherman"
<[EMAIL PROTECTED]> wrote:
> I have the following hardware:
>
> Dell OptiPlex GX1
> Pentium III running at 550 MHz
> Adaptec 2940 UW SCSI card
> Yamaha 8,4,24 CDR/CDRW writer
> External SCSI
> 128M RAM
>
> cdrecord versions:
> 1.9 (comes with RH 7.1)
> 1.11a01
>
> kernel versions:
> stock RH 7.1 kernel
> Linux version 2.4.4 ([EMAIL PROTECTED]) (gcc version 2.96
> 20000731 (Red Hat Linux 7.1 2.96-81)) #1 Wed May 9 11:09:12 EDT
> 2001
> (compiled myself--all drivers compiled into kernel)
>
>
This might be related to the new aic7xxx diver that is part of the 2.4.x
kernels. Suggest you recompile a kernel with support for the aic7xxx_old
driver. Solved the problems with my scsi cdr and cdrom
------------------------------
From: [EMAIL PROTECTED] (mndoci)
Subject: xmms on kde
Date: 23 May 2001 21:23:27 -0700
Hi Folks
I have been having trouble with XMMS when I use the KDE destop
environment (I run Mandrake 7.2). After every track, the cd or
whatever I am listening to stops, up pops an error message about the
device possibly being in use and I have to manually advance to the
next song. When I use gnome (enlightenment) as my desktop system, I
have no problems at all.
Please help!!!
Thanks
Mndoci
------------------------------
From: "Xoran99" <[EMAIL PROTECTED]>
Subject: Re: ziplinux
Date: Thu, 24 May 2001 00:03:54 -0500
If you're talking about Slackware's distro, I installed it and used it for a
while; I downloaded the 100MB version (30 MB download, I think), and learned
my way around Linux there. But because I got the smaller version, it didn't
come with the bells and whistles that the larger one had (it had X as well,
I think...), so it wasn't as robust as I wanted it to be. I eventually set
up a partition for Redhat 6.1, since I had access to a CD and all :P
Clinton
william <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Has anyone installed and had success with zip Linux?
>
> --
> Posted via CNET Help.com
> http://www.help.com/
------------------------------
From: "Vladimir Florinski" <[EMAIL PROTECTED]>
Subject: LILO limit still there?
Date: Wed, 23 May 2001 21:51:15 -0700
I was installing RH7.1 for someone who has some version of Windows (60GB
disk) and the installer would not go past the partitioning step, saying /
is above 1024 cylinders, etc., you know the routine. My understanding is that in
LILO this limit was lifted quite a while ago. Is Red Hat so far behind the
times? Or is there another reason?
--
Vladimir
------------------------------
From: "Support" <[EMAIL PROTECTED]>
Subject: Re: Swapping on RAID1
Date: Thu, 24 May 2001 15:28:42 +1000
I meant if the Drive goes west - it won't matter if I lost my
swap file, because it is not permenant data.
David
"Steve Wolfe" <[EMAIL PROTECTED]> wrote in message news:behhe9.q93.ln@helix...
> > I am running a software raid on 2 x ide drives - one is bigger than the
> > other so I utilise the difference in space by filling that one with the
> > swap. Works fine. I think that as the swap only stores temporary data it
> > does't matter if it ever goes west.
>
> Uh.... if the data you've swapped out is lost, expect "real bad things"
> when the machine tries to swap them back in.
>
> steve
>
>
>
------------------------------
Subject: Re: Dumb newbie question about setting the system time.
From: [EMAIL PROTECTED] (Martin Brown)
Date: Thu, 24 May 2001 06:17:41 GMT
In article <9ef8vl$dci$[EMAIL PROTECTED]>,
Joel <[EMAIL PROTECTED]> wrote:
>Hi. I am reletively new to linux (I've been using it for about 2 months)
>and my clock is off. I have it set to local time, but I can't figure out
>how to set it to the correct time. I remember when I installed RedHat 7.1,
>a package was installed that is for setting the time, but I don't remember
>the name. How do I set the hardware time on my computer? I have Windows
>and Linux installed, BTW. Thanks in advance.
>
>
I don't run RH7.1 (RH6.1/2).
You can set the hardware clock from Linux using hwclock.
from notes taken from other posts on this matter:
If /etc/sysconfig/clock contains "UTC=false", then
the system assumes that the bios time is the correct local time and
doesn't update for daylight savings. (Actually, the rollover will
be handled correctly, but the next time you boot, you'll be an hour
off again.)
If this is a server (i.e., never boots anything but Linux), then
configure your hardware clock to use Universal time ("UTC=true" in
/etc/sysconfig/clock) and use
hwclock --utc --systohc
to set the clock. Then the system will correctly handle daylight time
under all circumstances.
The only reason to set your BIOS clock to local time is to support
Windows, which is too brain-dead to understand Universal time. Real
OS's always use UTC 8^).
=====
I use ntpdate to set the linux system time from a time server.
I have set the following aliases for 'root':
alias set_sysclock='/usr/sbin/ntpdate 132.239.254.49'
alias set_hwclock='/sbin/hwclock --utc --systohc'
alias set_time='/usr/sbin/ntpdate 132.239.254.49;/sbin/hwclock --utc
--systohc'
I run 'set_time' every day or so.
--
- Martin J. Brown, Jr. -
- [EMAIL PROTECTED] -
PGP Public Key ID: 0xCED9BD8A Key Server: http://www.keyserver.net/en/
------------------------------
From: Neil Goldberg <[EMAIL PROTECTED]>
Subject: Re: Squid 2.4 and Redhat 7.1
Date: Thu, 24 May 2001 06:26:55 GMT
cistron wrote:
> Dear friends,
>
> Is anyone using Squid 2.4 with Redhat 7.1. Are there any problems with
> it. Also has anyone used IWil ATA 100 RAID controller with 7.1. What are
> your experiences.
>
> Thanks and Regards.
I have tested Squid 2.4 on Redhat 6.2 and Solaris 2.6 and 2.8. I am sure it
works on 7.1 without a hitch. Try building it with --enable-async-io, with
Redhat 7.1, you will fully be able to realize the multithreaded goodness!
No comment on the RAID controller.
Neil
------------------------------
From: "Kevin from Hamilton" <ac.ocitapmys@kytrehod>
Crossposted-To: comp.os.linux.setup,linux.debian.user,muc.lists.debian.user
Subject: Strange Question..
Date: Thu, 24 May 2001 02:43:29 -0400
I was just wondering, if anyone can help me.
I have a PAS-16 Sound Card with the SCSI connector and I want to add a 9.1GB
SCSI Hard Drive.
My question is simple, can I do it? Is it possible or should I forget it
before I even make the purchase. And the other follow up question is does
anyone know if I can add a SCSI-2 or a SCSI-Fast drive? And finally how
could I find out what 50 Pin SCSI devices or MAX hard drive size or SCSI
types could I get for it. I don't have the book anymore and I can't find
anything that can help me on the net.
I turned to this newsgroup because I will be using it on a linux machine.
(Debian stable)
Well in any case... I hope someone can get back to me with some information.
Thanks in advance.
------------------------------
From: "Xoran99" <[EMAIL PROTECTED]>
Subject: Re: Test don't reply
Date: Thu, 24 May 2001 01:59:23 -0500
How unoriginal. Blech.
Clinton
Paul <[EMAIL PROTECTED]> wrote in message news:[EMAIL PROTECTED]...
> Test don't reply
>
>
------------------------------
** 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.misc.
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-Misc Digest
******************************
