I've recently set up a Linux box in order to expand my knowledge on the
subject, as well as run a few dedicated services that i see being useful
for my home network.
i choose fedora core 2, as it was the next evolution of red hat 9 (many
of the web servers i host on are red hat 9).
i was at first impressed that it seemed to include packages for all of
the things i was looking at accomplishing:
Thinking these packages would be ready to go out of the box, has me in a
bit of a quandary, as many seem to take much more to get going.
Today i was successful in getting apache and subversion running. Apache
ended up being installed, i just need to get it to autostart as a
service on boot-up.
Subversion took a bit of work, but was able to piece it together using
the RPMs linked to on their server.
However, I have still yet to get php working with apache. but i will
save that for another day.
I am interested in getting LDAP set up. I would like to consolidate my
address book, for both my windows and linux clients into an LDAP
directory. However, I am a bit lost at getting it installed / configured
/ and running. I have found openldap directories scatterd about my
system... and when reviewing the documentation there was a mention of
adding config options to a sldap.conf file. only l file similiar was
found and that was ldap.conf
Any help on getting openldap working on my fedora core 2 system would be
This is what has me in a bit of a quandary... so far the things that the
o/s install said got installed with it are either not fully installed or
configured. I understand that it is up to me to get it configured, and
all of that, however, when doing a which or find for the actual program,
nothing is returned. even broader searches of the system find no mention
of the files even existing on my system (case in point... when looking
for the subversion install... searches for subversion, svn, or svnadmin
pulled back nothing... which had me install it from the RPMs linked to
from the subversion site).
I am new to Linux, and am willing to learn / teach myself. However, i'm
left a bit confused by having it tell me that these things have been
installed for me, yet not actually finding evidence of such.
Well, my samba book just arrived from amazon, so i'm going to pour
myself into that tonight and see what I come up with. My other use for
LDAP was to combine it with samba, so i'd like to get ldap up and
thanks in advance for your input, answers, and links to resources. i
- eric marden
the mahine hangsup when
ENABLING SWAP SPACE is encountered during bootup....??
Do you Yahoo!?
New and Improved Yahoo! Mail - Send 10MB messages!
I know this has prob been asked a thousand times before but....
I have recently downloaded the source for kernel 2.6.7 and recompiled with
8K stacks instead of 4. Nvidia Driver still freezes machine on boot. I am
thinking I missed changing something in the kernel config. I ran make
oldconfig to generate a .config file (based on kernel 2.6.6-1.435) in my
linux-2.6.7 dir, than ran make xconfig changed my CPU to Athlon, made sure I
was using 8K stacks and left the rest defaulted to my .config
file......Booted new kernel, Installed the Nvidia Drivers, made changes to
xorg.conf. System freezes on boot. Did I miss changing something in the
kernel config?? Any suggestions??
STOP MORE SPAM with the MSN Premium and get 2 months FREE*
I've been having problems with the contents of files simply disappearing.
I am running fedora core 1, AMD64 version. I have a Western Digital
250GB USB 2.0 drive attached to the system onto to which I often
download video files.
Just this morning I had finished a particularly long download to this
drive. I moved the files over to another directory on the same drive.
I've been watching these files all morning until now, when I tried to
watch the next in the series, the contents of all these files was
suddenly 0 (except the one I was just watching).
This has happened before, but I took it for a fluke. I've had trouble
with this system before, but this seems very unusual.
If anyone has any advice, I'd love to hear it.
Up to Fedora 1, I could mount encrypted files but I
can't do it on Fedora 2.
I used to do:
# mount -o loop,encryption=xor encrypted_file /mnt/dir
Password: <my password here>
Now I get:
# mount -o loop,encryption=xor encrypted_file /mnt/dir
Password: <my password here>
ioctl: LOOP_SET_STATUS: Invalid argument
I've seen similiar complaints on the net, but people
complain about des or some other stronger encryption,
while they say xor works. I can't even get xor
I've tried loading modules such as cryptoloop, xor,
des, etc. They all load, but I can't never mount the
file. What's the matter with losetup? Please help.
Yahoo! Messenger - Fale com seus amigos online. Instale agora!
Until last night, I was running Fedora Core 1 with FreeSWAN. I was
using FreeSWAN to connect to my place of work, and two different
customer sites via IKE IPsec tunnels. It was working great.
Since upgrading to Fedora Core 2, I cannot get FreeSWAN to work. So I
decided to try the ipsec implementation included with the 2.6 kernel. I
am having no luck. Is there any documentation for this anywhere? I'm
referring specifically to the RedHat/Fedora implementation. I'm trying
to make use of it (including the GUI set tool under System Settings -->
Network --> IPsec).
Sometimes racoon doesn't even respond to pings from my Windows XP PC
behind my Linux firewall. When it does, it fails to connect. Racoon
logs a "hash mismatch". The Sidewinder G2 firewall on the work end logs
a similar message about the configuration not matching. I've noticed
that racoon keeps trying aggressive mode. I don't know why. I've
manually edited racoon.conf (and the individual .conf files for the
other end) to eliminate aggressive mode (just leaving main mode).
Also, when setting up an IPsec tunnel using the GUI Network tool, it
asks for "Local network address". According to the documentation I
found for RedHat Enterprise Linux, this should be the internal interface
for the Linux firewall. However, I don't think this is right. I think
this should actually be the subnet address, correct? In the Remote
Network screen, it asks for both "Remote IP address" and "Remote network
gateway". Aren't these the same thing? There's a separate field for
"Remote network address", so it would seem that "Remote IP address" and
"Remote network gateway" should both be the external interface IP
address of the remote firewall.
I have noticed two error messages that seem particularly troubling to
me. First, when racoon first starts, it complains with
ERROR: isakmp.c:1378:isakmp_open(): failed to bind to address [address
of NIC] (no such device)
for both NICs.
Also, if I try to activate a configured tunnel with "ifup [tunnel
name]", it replies with "RTNETLINK answers: Network is unreachable". What?
If there is some good documentation for this ipsec system, please point
me to it! Any other help would be greatly appreciated--I need this
connection up so that I can work from home!
I ran into a problem setting up a Linux system for a co-worker. He has
an external firewire disk.
The problem is: when it is connected, the system detects a new device:
# cat /proc/bus/ieee1394/devices
Vendor ID: `Linux OHCI-1394' [0x004063]
IRMC(1) CMC(1) ISC(1) BMC(0) PMC(0) GEN(0)
LSPD(2) MAX_REC(2048) CYC_CLK_ACC(0)
Host Node Status:
Host Driver : ohci1394
Nodes connected : 2
Nodes active : 2
SelfIDs received: 2
Irm ID : [0-01:1023]
BusMgr ID : [0-63:1023]
In Bus Reset : no
Root : yes
Cycle Master : yes
IRM : yes
Bus Manager : no
Vendor ID: `Granite Digital' [0x0004da]
IRMC(0) CMC(0) ISC(0) BMC(0) PMC(0) GEN(0)
LSPD(0) MAX_REC(64) CYC_CLK_ACC(255)
Unit Directory 0:
Vendor/Model ID: Granite Digital [0004da] / FireVue 1394-IDE Bridge LUN0 
Software Specifier ID: 00609e
Software Version: 010483
Driver: SBP2 Driver
Length (in quads): 8
But what it detects is just the electronics in the enclosure, which is
essentially an IDE-controller to which the actual disk (Maxtor 300 GB)
is connected. So an sbp2 driver gets loaded, and nothing more.
I figured out that it works to do:
# echo "scsi add-single-device 1 0 0" > /proc/scsi/scsi
after which /dev/sda1 exists. So basically my only remaining question
is: where to add this in the scripts of hotplug or devlabel or whichever
component can do such a task.
Hi, I was wondering what was the right way to start and mount a software
RAID array at boot time in FC2. I have a computer at home with a RAID-1
and a RAID-5 array and just adding them both in the fstab is not enough
because I need to start the arrays first. Where do I set them up? I
could just add mdadm -As /dev/md0 and mount /dev/md0 in rc.local and do
the same for /dev/md1 but I'd like to do it "The right way".
Jean-Rene Cormier <jean-rene.cormier(a)cipanb.ca>
Long story, short question: Should I fall back to FC1 instead of using
FC2? I am somewhat very concerned about the release quality of FC2.
Here's what happened:
I just upgraded a laptop (Fujitsu-Siemens E2010) from RH9 to FC2. After
that, I couldn't log on to Gnome, after username/password, X just hung
without any error message, neither in the logs. Never mind, I did a
fresh install instead & immediately after updated all the packages. Then
the surprises really started:
- Service PCMCIA doesn't load the yento-sockets module. This is known @
bugzilla, one has to patch the pcmcia script by hand. Through this bug,
the PCMCIA subsystem is non-functional!
- Kernel 2.6.6-1.435 gives a "failed to bind key 255 to value..." every
time you log in at a console, whereas kernel -358 doesn't. This appeared
at bugzilla but was closed as "magically disappeared"
- Hotplug calls the script ide, which in turn calls /sbin/ide_info.
/sbin/ide_info simply doesn't exist! Had to copy it in from a FC1
install. Through this bug, you can't mount CF-Cards in a PCMCIA adapter.
Just found out that bugzilla also knows about this -- since FC2 test3!
- USB updfstab doesn't work. Devices get detected in the kernel, hotplug
@ updfstab gets executed, but /mnt & fstab entries are not created.
Trying to debug, I realized that the issue is somehow timing-related.
When I do some sleep at the "right" points in the updfstab scripts,
things work sometimes. Can't resolve the issue though and have to live
without using my Flash-Stick & Camera reliably.
- Kernel -435 sometimes segfaults when plugging USB devices in/out,
whereas -358 seems reliable here.
- My console-beep (emitted through the sound card) disappeared ever
since FC2. I don't seem to be able to find a pointer on how to trace
This is the result after a few hours of FC2 experience and I really
start wondering about what level of testing this release experienced
before being released into the public -- even IMHO obvious things like
the PCMCIA subsystem doesn't work! Seems to me that FC1 was by far the
more reliable system.
My question now is whether I should give up on FC2 (I'm really tempted)
and go back to FC1 or hang in there.
- Will I experience more of these surprises with FC2?
- How fast will issues be attented to through updates? The PCMCIA issue
is known, so is the ide_info issue (since FC2 test3), but still no fix
- Is this what we have to expect from a "community distro" in the long
run? As with RH9 the last consumer-distro is discontinued, my logical
conclusion would then be to leave RH behind and move on to another
packager, say Suse?
A bit disappointed,
Just want to be sure. The instructions from nVidia are not real precise
on the point of exiting X and/or booting to something other than
Is the a key stroke/combination to exit X in FC2? And, to set my default
runlevel do I only need to edit my /etc/inittab lines:
id:5:initdefault: (changing 5 to 3)
x:5:respawn:/etc/X11/prefdm -nodaemon (changing 5 to 3)
As I recall I can leave the runlevel at 3 and use 'startx' to Start X
until which time I am satisfied with the driver installation?
If there is something I am missing I would like to get it before I start
on this installation. I remember when I used IBM OS/2 Warp and upgrading
to faster video driver incorrectly could really do a number on a system.
Thanks for your support,