FedoraForum.org - Fedora Support Forums and Community
Page 2 of 3 FirstFirst 123 LastLast
Results 16 to 30 of 31
  1. #16
    jdillick Guest
    Quote Originally Posted by kjellhansen
    e2fsck -f -b 32768 /dev/VolGroup00/LogVol00

    -----------------------------------------------------------------
    e2fsck 1.39 (29-May-2006)
    Pass 1: checking inodes, blocks, and sizes
    Pass 2: Checking directory structure
    i_file_acel for inode 2 (/) is 36910, should be zero.
    Clear <y>?
    -----------------------------------------------------------------

    I am confused about "Clear." By typing 'y' will I be accepting or rejecting the value 36910?
    You can press enter, and it will take the default answer, 'y'. This, I believe, will set the acl value for that inode to 0. I don't know what that means, but you are kind of trusting the tool to do the right thing.

    I just did this procedure on my external usb drive, and answered all these questions by holding down the enter key until they all went away. I was still able to mount the drive fine, and everything looked good.

  2. #17
    kjellhansen Guest
    After holding down enter forever (or resting a 9V battery on the key rather), I reached the end of the item list, hoping all had been restored to its right place.

    I typed in mkdir temp and then mount/dev/VolGroup00/LogVol00 temp. Neither command received any reply from the system (can I assume these commands were performed and that the computer would have complained had it not recognized them?). Then I exited the shell and tried booting from the hard drive. The kernel has apparently been restored to its former self, as there was no panic, but there were clearly multiple things that FAILED among the list of PASSes that streams out following the kernel's successful activation. After this list had been run down, a new screen came up to inform me:

    "The X-server (your graphical interface) cannot be started. It is likely that it is not set up correctly. You will need to log in on a console and rerun the X configuration application. then restart GDM."

    Clicked on OK, then a non-graphical login screen appeared. Upon entering my username and password, there would be a momentary flash of another text block, and then the login prompt would reappear. After entering my info many times (trying to glimpse the ephemeral text), the system got tired of my persistence and took a 5 minute break:

    ----------------------------------------------------------------------------
    Last login Thu Aug 30 19:17:23 on tty1
    audit (1188526650.254:17): avc: denied {entrypoint} for pid=2575 comm= "login
    " name="bash" dev=dm-0 ino=107315213 scontext= user_u: system_r: hotplug_t: s0 tcont
    ext= system_u: object_r: file_t: s0 tclass= file
    login: no shell: Permission denied.
    INIT: Id "1" respawning too fast: disabled for 5 minutes
    ----------------------------------------------------------------------------

    I wasn't sure if this screen's info would provide any useful information, but I decided to include it just in case. Will I be able to do any setup reconfiguration booting from the hard drive now, or should I boot back into rescue mode from the installation disc?

  3. #18
    jdillick Guest
    Quote Originally Posted by kjellhansen
    After holding down enter forever (or resting a 9V battery on the key rather), I reached the end of the item list, hoping all had been restored to its right place.

    I typed in mkdir temp and then mount/dev/VolGroup00/LogVol00 temp. Neither command received any reply from the system (can I assume these commands were performed and that the computer would have complained had it not recognized them?). Then I exited the shell and tried booting from the hard drive. The kernel has apparently been restored to its former self, as there was no panic, but there were clearly multiple things that FAILED among the list of PASSes that streams out following the kernel's successful activation. After this list had been run down, a new screen came up to inform me:

    "The X-server (your graphical interface) cannot be started. It is likely that it is not set up correctly. You will need to log in on a console and rerun the X configuration application. then restart GDM."

    Clicked on OK, then a non-graphical login screen appeared. Upon entering my username and password, there would be a momentary flash of another text block, and then the login prompt would reappear. After entering my info many times (trying to glimpse the ephemeral text), the system got tired of my persistence and took a 5 minute break:

    ----------------------------------------------------------------------------
    Last login Thu Aug 30 19:17:23 on tty1
    audit (1188526650.254:17): avc: denied {entrypoint} for pid=2575 comm= "login
    " name="bash" dev=dm-0 ino=107315213 scontext= user_u: system_r: hotplug_t: s0 tcont
    ext= system_u: object_r: file_t: s0 tclass= file
    login: no shell: Permission denied.
    INIT: Id "1" respawning too fast: disabled for 5 minutes
    ----------------------------------------------------------------------------

    I wasn't sure if this screen's info would provide any useful information, but I decided to include it just in case. Will I be able to do any setup reconfiguration booting from the hard drive now, or should I boot back into rescue mode from the installation disc?
    Yeah, you are part of the way back. Obviously, some system stuff got messed up when your filesystem took a dump. Try booting the system in runlevel 3 and reconfiguring X.

    When grub comes up on boot, hit a key to stop it, and press 'e'. Select the kernel line and hit 'e', and put a space and '3', and press enter. Press 'b' to boot.
    See if it will let you log in to the CLI at that point and run:
    system-config-display --reconfig
    If you can't log in: Bootup the same way except boot into single user mode with a '1', instead of the '3'.
    You should get a CLI prompt without the password, and can then reset your password. Let me know if I misunderstood the problem you were having on login.

  4. #19
    kjellhansen Guest
    I tried booting on runlevel 3 from my hd, but the only difference was that I no longer get the "The X server (your graphical interface) cannot be started. It is likely that it is not set up correctly. You will need to log in on a console and rerun the X configuration application, then restart GDM." message screen after my first login attempt.

    Despite the absence of this message, the system's behavior is the same. It does not appear likely that the system is failing to recognize my login (as indeed if I enter a false username/password it tells me it's incorrect), rather, it is unable to load a graphical interface beyond that point, or something that's getting it stuck.

    In any case, I can't seem to get past that login loop to a shell on my hd, so working from my install disc's rescue mode still seems like the way to go. In this environment, when I type "system-config-display --reconfig" my system spits out:

    -------------------------------------------
    Traceback (most recent call last):
    File "/usr/share/system-config-display/xconf.py", line 32,
    in <module> import rhpxl.monitor
    Import error: No module named rhpxl.monitor
    --------------------------------------------

    I tried the suggestion of another linux forum member to change my Xorg configuration by "nano /etc/X11/xorg.conf" and thereupon edit my nvidia driver "nv" to "vesa" (as a stopgap to get a GUI from which to work, then reconfigure for nv later), then run startx.

    But I get the following response to the simple command "startx", whether I have my driver saved to vesa or nv:

    -----------------------------------------------------------------
    /user/bin.startx: line 132: xauth: command not found
    /user/bin.startx: line 142: xauth: command not found
    /user/bin.startx: line 144: xauth: command not found
    /user/bin.startx: line 142: xauth: command not found
    /user/bin.startx: line 144: xauth: command not found

    xinit: No such file or directory (errno 2): no server "X" in PATH

    Use the -- option, or make sure that /usr/bin is in your path and that "X" is a program or a link to the right type of server for your display. Possible server names include:
    Xorg X.Org displays

    giving up.
    xinit: Network is unreachable (errno 101): unable to connect to X server
    xinit: No such process (errno 3): Server error.
    /usr/bin/startx: line 167: xauth: command not found
    -----------------------------------------------------------------

    To see if I could find out more about why the "system-config-display --reconfig" you suggested was getting hung up, I used nano to open up /usr/share/system-config-display/xconf.py and looked down to line 32, which did indeed list rhpxl.monitor, but isn't this what it's supposed to be? I left this file alone.

  5. #20
    kjellhansen Guest
    I tried booting the hd on runlevel 1, allowing me to bypass login, but I got no system response from "system-config-display --reconfig", in this state.

  6. #21
    kjellhansen Guest
    I mentioned before that some PASS and FAIL entries stream by after the kernel boots. I rebooted a whole bunch of times trying to copy down the ones I could read quickly enough and this is what I got with the hope it might possibly be useful:

    "line 3 not found" , try iptables-restore

    hidd: HIDP protocol:
    autofs (4 failed entries)

    hpiod

    xfs: (?)

    libxfont.so.1

  7. #22
    Join Date
    Aug 2007
    Posts
    455
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    LVM some times create problems reinstall FC7 selecting custom lay out

  8. #23
    jdillick Guest
    Quote Originally Posted by kjellhansen
    I tried booting on runlevel 3 from my hd, but the only difference was that I no longer get the "The X server (your graphical interface) cannot be started. It is likely that it is not set up correctly. You will need to log in on a console and rerun the X configuration application, then restart GDM." message screen after my first login attempt.

    Despite the absence of this message, the system's behavior is the same. It does not appear likely that the system is failing to recognize my login (as indeed if I enter a false username/password it tells me it's incorrect), rather, it is unable to load a graphical interface beyond that point, or something that's getting it stuck.

    In any case, I can't seem to get past that login loop to a shell on my hd, so working from my install disc's rescue mode still seems like the way to go. In this environment, when I type "system-config-display --reconfig" my system spits out:

    -------------------------------------------
    Traceback (most recent call last):
    File "/usr/share/system-config-display/xconf.py", line 32,
    in <module> import rhpxl.monitor
    Import error: No module named rhpxl.monitor
    --------------------------------------------

    I tried the suggestion of another linux forum member to change my Xorg configuration by "nano /etc/X11/xorg.conf" and thereupon edit my nvidia driver "nv" to "vesa" (as a stopgap to get a GUI from which to work, then reconfigure for nv later), then run startx.

    But I get the following response to the simple command "startx", whether I have my driver saved to vesa or nv:

    -----------------------------------------------------------------
    /user/bin.startx: line 132: xauth: command not found
    /user/bin.startx: line 142: xauth: command not found
    /user/bin.startx: line 144: xauth: command not found
    /user/bin.startx: line 142: xauth: command not found
    /user/bin.startx: line 144: xauth: command not found

    xinit: No such file or directory (errno 2): no server "X" in PATH

    Use the -- option, or make sure that /usr/bin is in your path and that "X" is a program or a link to the right type of server for your display. Possible server names include:
    Xorg X.Org displays

    giving up.
    xinit: Network is unreachable (errno 101): unable to connect to X server
    xinit: No such process (errno 3): Server error.
    /usr/bin/startx: line 167: xauth: command not found
    -----------------------------------------------------------------

    To see if I could find out more about why the "system-config-display --reconfig" you suggested was getting hung up, I used nano to open up /usr/share/system-config-display/xconf.py and looked down to line 32, which did indeed list rhpxl.monitor, but isn't this what it's supposed to be? I left this file alone.
    I think the approach I would take if I wanted to save this install, would be to try to reinstall some of the binary packages that are causing the problem. If you can mount the F7 DVD from rescue mode, you may be able to. Or perhaps use rescue + network + yum...

    You can't login for some reason, with this mysterious "login: no shell: Permission denied."
    Some things to check here (this could be a nasty one):
    Boot up into rescue and check on /etc/passwd, make sure you have for root:
    root:x:0:0:root:/root:/bin/bash
    Also, make sure /bin/bash exists and can check its permissions:
    #after booting into rescue mode
    chroot /mnt/sysimage
    #You can check all the permissions for the whole package
    rpm -qV bash
    ls -l /bin/bash

    # For your reference
    # S file Size differs
    # M Mode differs (includes permissions and file type)
    # 5 MD5 sum differs
    # D Device major/minor number mismatch
    # L readLink(2) path mismatch
    # U User ownership differs
    # G Group ownership differs
    # T mTime differs
    Here was an article about someone having a similar problem, and it turned out to be a permission problem with some of the shared-object libraries on the system; link

    You can do a system-wide verify as well, to see what kind of things vary from the original files placed during install:
    rpm -qaV | less
    That will run for a long time, and may have a lot of output. You may want to redirect that to a file and get it off the system, instead of writing it down. But it could be some useful info.

    In regard to the rhpxl problem:
    The rhpxl package seems to be a problem for you right now, otherwise it is python or the X server itself. system-config-display and firstboot require rhpxl. You shouldn't need firstboot at all, so you can remove it:

    From the rescue environment chroot to /mnt/sysimage:

    rpm -e firstboot #You don't need this
    Now, remove and reinstall system-config-display and rhpxl together.

    rpm -e system-config-display rhpxl
    ##############
    #Reinstall
    ##############
    #If doing from F7 media
    rpm -ivh /path-to-f7-rpms/system-config-display*.rpm /path-to-f7-rpms/rhpxl*.rpm

    #If doing from yum
    yum install system-config-display rhpxl

  9. #24
    kjellhansen Guest
    Root from /etc/passwd was normal and the check on /bin/bash read out:
    -rwxr-xr-x 1 root root 797312 2007-02-12 07:18 /bin/bash

    Do the reference fields you posted relate to each block within that line (e.g. D: 1, G: 797312, etc)? I've only read a little through the article you linked to so far.

    I ran "rpm -qaV | less > rpm-output2.txt 2>&1" (I'd done rpm -aV --quiet > rpm-output.txt 2>&1 a day ago) and the list was indeed a massive thing to sift through. I used nano's crtl+W to search for terms that flashed by as FAIL after hd kernel startup and searched a few other items that seem to be causing issues:

    missing /usr/bin/Xorg
    (many entries for) /usr/lib64/xorg/modules/...
    " /usr/lib64/python2.5/site-packages/rhpxl/...
    SM5....T c /etc/sysconfig/iptables-config
    missing c /etc/X11/fs/config
    S.?..... /usr/bin/xfs
    missing /usr/sbin/hpiod
    missing c /etc/pam.d/system-config-users
    missing /usr/lib64/libXfont.so.1(.4.1)

    I was eager to try out your advice RE the rhpxl problem. At first I tried "rpm -e firstboot" from runlevel 1 of my hd, but the system response was "error: Failed dependencies: firstboot is needed by (installed) system-config-keyboard-1.2.11-1.fc7.noarch". Then I tried "rpm -e firstboot system-config-keyboard", which responded "error: %postun (system[...].noarch) sciptlet failed, exit status 255. error: %preun (firstboot[...]noarch) sciptlet failed, exit status 255."

    But when I booted into the rescue disc environment "rpm -e firstboot system-config-keyboard" worked and so did "rpm -e system-config-display rhpxl".

    I haven't been able to reinstall though:
    rpm -ivh /path-to-f7-rpms/system-config-display*.rpm /path-to-f7-rpms/rhpxl*.rpm

    ------------------------------------------------------------------------
    error: File not found by glob: /path-to-f7-rpms/system-config-display*.rpm
    error: File not found by glob: /path-to-f7-rpms/rhpxl*.rpm
    ------------------------------------------------------------------------

    Attempts to reinstall these in runlevel 1 on my hd with "yum install system-config-display rhpxl" is not working. But, whether this is due to a problem with yum, my network, or the syntax of the command, I am not sure.

  10. #25
    jdillick Guest
    Quote Originally Posted by kjellhansen
    Root from /etc/passwd was normal and the check on /bin/bash read out:
    -rwxr-xr-x 1 root root 797312 2007-02-12 07:18 /bin/bash

    Do the reference fields you posted relate to each block within that line (e.g. D: 1, G: 797312, etc)? I've only read a little through the article you linked to so far.

    I ran "rpm -qaV | less > rpm-output2.txt 2>&1" (I'd done rpm -aV --quiet > rpm-output.txt 2>&1 a day ago) and the list was indeed a massive thing to sift through. I used nano's crtl+W to search for terms that flashed by as FAIL after hd kernel startup and searched a few other items that seem to be causing issues:

    missing /usr/bin/Xorg
    (many entries for) /usr/lib64/xorg/modules/...
    " /usr/lib64/python2.5/site-packages/rhpxl/...
    SM5....T c /etc/sysconfig/iptables-config
    missing c /etc/X11/fs/config
    S.?..... /usr/bin/xfs
    missing /usr/sbin/hpiod
    missing c /etc/pam.d/system-config-users
    missing /usr/lib64/libXfont.so.1(.4.1)

    I was eager to try out your advice RE the rhpxl problem. At first I tried "rpm -e firstboot" from runlevel 1 of my hd, but the system response was "error: Failed dependencies: firstboot is needed by (installed) system-config-keyboard-1.2.11-1.fc7.noarch". Then I tried "rpm -e firstboot system-config-keyboard", which responded "error: %postun (system[...].noarch) sciptlet failed, exit status 255. error: %preun (firstboot[...]noarch) sciptlet failed, exit status 255."

    But when I booted into the rescue disc environment "rpm -e firstboot system-config-keyboard" worked and so did "rpm -e system-config-display rhpxl".

    I haven't been able to reinstall though:
    rpm -ivh /path-to-f7-rpms/system-config-display*.rpm /path-to-f7-rpms/rhpxl*.rpm

    ------------------------------------------------------------------------
    error: File not found by glob: /path-to-f7-rpms/system-config-display*.rpm
    error: File not found by glob: /path-to-f7-rpms/rhpxl*.rpm
    ------------------------------------------------------------------------

    Attempts to reinstall these in runlevel 1 on my hd with "yum install system-config-display rhpxl" is not working. But, whether this is due to a problem with yum, my network, or the syntax of the command, I am not sure.
    The /path-to-f7-rpms path I have in the command is just to prompt you to find the *actual* path to your F7 media. If you have the dvd in the drive, it would probably be mounted under /media/Fedora or something like that, I'm not sure.

    If you are in runlevel 1 (single user mode), this implies there is no network connectivity. You can sometimes selectively bring network services up, with:
    /sbin/service network start
    If you can get network access through fedora rescue, then I would continue working that way. When you execute the chroot /mnt/sysimage command, you will be executing commands (even network commands like yum) using the executables from your hard drive. You should be able to yum install there after chroot'ing.

    Sorry if I confused you.

    Oh, one other thing. When you are booting into single user mode, and get the sh# prompt, try to login to a local account (replace youraccount below with your local non-root user), and see if it errors out:
    su - youraccount
    Especially if this fails, tail the /var/log/message log file to see if it gives us any info on the failure:
    tail /var/log/message

  11. #26
    kjellhansen Guest
    su - myaccount worked fine within single user mode, thus no message to tail.

    Lol, I thought the "path-to-f7-rpms" part of that command was a bizarre directory name. My mistake. I tried some various combinations/guesses, but ultimately couldn't figure out what the name of the path to the disc's rpms was.

    Fortunately however, after properly configuring my network before entering the rescue disc's shell (another embarassing admission--I'd just been skipping over the network configuration options each time), I did a successful yum reinstall of rhpxl, system-config-display and system-config-keyboard (firstboot also got automatically reinstalled as a dependency).

    With these fresh packages in place, I tried "system-config-display --reconfig", resulting in a new error message:

    --------------------------------------------------------------------------------
    Traceback (most recent call last):
    File "/usr/share/system-config-display/xconf.py", line 111, in start_x_server
    (server, temp-config) = fork_off_server (xconfig)
    File "/usr/share/system-config-display/xconf.py", line 104, in fork_off_server
    os.execv (args [0], args)
    OSError: [Errno 2] No such file or directory

    Error, failed to start X server.
    Couldn't start x server on card 1
    Error, failed to start X server.
    Couldn't start X server on card 0
    -------------------------------------------------------------------------

    Later, back on runlevel 1 I tried "/sbin/service network start" (as root) and got:

    -------------------------------------------------------------------------
    FATAL: Could not load /lib/modules/2.6.22.4-65.fc7/modules.dep: No such file or directory
    Bringing up loopback interface: ./network-functions: line 166: cd: /var/run/net
    report: No such file or directory.
    [OK] ## <-- green font
    --------------------------------------------------------------------------

    As a miscellaneous note, I was able to read one FAILed item of the start up list pretty clearly:

    hpiod: /bin/bash: ./hpiod: No such file or directory.

    Regarding the resulting contents of "rpm -qaV | less > rpm-output2.txt 2>&1", I am curious if I'm able to (and might it be a good idea to) select some of these items for removal using the "rpm -e xxx" command syntax and perform fresh reinstalls via yum (within the network configured rescue disc environment)? For example, if it's possible to do "rpm -e Xorg" followed by "yum install Xorg", would this be good to try? Please let me know if this syntax holds up and whether some of the damaged or missing items listed by rpm -qaV | less are unable to be refreshed in such a manner.

  12. #27
    jdillick Guest
    Quote Originally Posted by kjellhansen
    su - myaccount worked fine within single user mode, thus no message to tail.

    --snip--

    With these fresh packages in place, I tried "system-config-display --reconfig", resulting in a new error message:

    --------------------------------------------------------------------------------
    Traceback (most recent call last):
    File "/usr/share/system-config-display/xconf.py", line 111, in start_x_server
    (server, temp-config) = fork_off_server (xconfig)
    File "/usr/share/system-config-display/xconf.py", line 104, in fork_off_server
    os.execv (args [0], args)
    OSError: [Errno 2] No such file or directory

    Error, failed to start X server.
    Couldn't start x server on card 1
    Error, failed to start X server.
    Couldn't start X server on card 0
    -------------------------------------------------------------------------

    Later, back on runlevel 1 I tried "/sbin/service network start" (as root) and got:

    -------------------------------------------------------------------------
    FATAL: Could not load /lib/modules/2.6.22.4-65.fc7/modules.dep: No such file or directory
    Bringing up loopback interface: ./network-functions: line 166: cd: /var/run/net
    report: No such file or directory.
    [OK] ## <-- green font
    --------------------------------------------------------------------------

    As a miscellaneous note, I was able to read one FAILed item of the start up list pretty clearly:

    hpiod: /bin/bash: ./hpiod: No such file or directory.

    Regarding the resulting contents of "rpm -qaV | less > rpm-output2.txt 2>&1", I am curious if I'm able to (and might it be a good idea to) select some of these items for removal using the "rpm -e xxx" command syntax and perform fresh reinstalls via yum (within the network configured rescue disc environment)? For example, if it's possible to do "rpm -e Xorg" followed by "yum install Xorg", would this be good to try? Please let me know if this syntax holds up and whether some of the damaged or missing items listed by rpm -qaV | less are unable to be refreshed in such a manner.
    I'd say once we get your login problem fixed, you're golden... until then it will continue to be painful.

    As for hpiod: don't worry about it. It is part of cups (printer support) for HP printers, the service is hplip.

    The modules.dep message means that depmod needs to be run. Before starting network services, run:
    depmod -a
    Then check and make sure /lib/modules/2.6.22.4-65.fc7/modules.dep exists. You need this file so modprobe can add your kernel modules (such as your ethernet driver). This file is basically a list of kernel module dependencies and their paths for the current running kernel.

    Check your eth0 alias in /etc/modprobe.conf to see which driver is being loaded. You can manually add it after the depmod. For instance, my modprobe.conf has "alias eth0 tg3", so I can execute:
    modprobe tg3
    On my system, this would load the tg3 kernel module to support my ethernet nic.

    As for removing and reloading packages... This is very possible, however it will be a bit painful with dependencies. Basically you have to remove and reinstall the desired package, and any package that depends on it, or requires its CAPABILITIES so to speak (see man rpm if you are curious).

    rpm may not be the best tool to remove it. Yum might do a better job, as it will check dependencies and spell out all the packages that need to be removed.
    For example, if I run yum to remove xorg, I get this:
    # yum remove xorg-x11-server-Xorg-1.3.0.0-9.fc7
    Dependencies Resolved

    ================================================== ===========================
    Package Arch Version Repository Size
    ================================================== ===========================
    Removing:
    xorg-x11-server-Xorg i386 1.3.0.0-9.fc7 installed 7.1 M
    Removing for dependencies:
    compiz i386 0.3.6-8.fc7 installed 2.5 M
    firstboot noarch 1.4.35-1.fc7 installed 884 k
    linuxwacom i386 0.7.4_1-2.1 installed 402 k
    rhpxl i386 0.47-1.fc7 installed 396 k
    synaptics i386 0.14.4-8.fc6 installed 110 k
    system-config-display noarch 1.0.51-1.fc7 installed 724 k
    system-config-keyboard noarch 1.2.11-1.fc7 installed 193 k
    .......... -- snip-- ............
    xorg-x11-drv-void i386 1.1.0-4.fc7 installed 7.3 k
    xorg-x11-drv-voodoo i386 1.1.0-4.fc7 installed 28 k

    Transaction Summary
    ================================================== ===========================
    Install 0 Package(s)
    Update 0 Package(s)
    Remove 72 Package(s)
    This will not add 72 Packages back when I do the reverse however. I will only get the dependencies of xorg-x11-server-Xorg, which will likely still be installed, unless they have cross dependencies. In short, it is kind of a pain. Are you sure you want to do it this way (as opposed to reinstalling f7 without initializing your / partition)?

    I would say you definately want to reinstall most packages that have files show up as status "5" (md5sum differs) when you do your "rpm -qaV > rpmoutput.txt", unless the file in question is a config file (which will most certainly differ.) Also, some missing files can be bad, particularly any files ending in .so (libraries) or that are in some "bin" directory, like /usr/bin /bin /sbin /usr/sbin.

    If you are unsure about the contents of the file, run the "file" command on the full path, and if it is an ELF or other binary data format, you can probably assume something a md5sum discrepancy is bad. For example

    $ file /bin/bash
    /bin/bash: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.9, stripped
    I hope that if you don't get it fixed, at least you are learning a lot about your system. I know I am.

  13. #28
    jdillick Guest
    Quote Originally Posted by kjellhansen
    Later, back on runlevel 1 I tried "/sbin/service network start" (as root) and got:

    -------------------------------------------------------------------------
    FATAL: Could not load /lib/modules/2.6.22.4-65.fc7/modules.dep: No such file or directory
    Bringing up loopback interface: ./network-functions: line 166: cd: /var/run/net
    report: No such file or directory.
    [OK] ## <-- green font
    --------------------------------------------------------------------------
    That /var/run/netreport is a directory that should be there. It is installed as part of the initscripts package. You can just add it back with "mkdir /var/run/netreport", make sure it is chown'd to root:root and is chmod'd to 775.

    Do a "rpm -qV initscripts" and make sure it isn't a bigger problem than just the one directory.

    $ rpm -qV initscripts
    ..5....T c /etc/inittab
    This is about normal.

  14. #29
    kjellhansen Guest
    Did depmod -a, then verified that /lib/modules/2.6.22.4-65.fc7/modules.dep exists.

    There actually was no ethernet alias in etc/modprobe.conf, so I added "alias eth1 tg3" (eth1, because I've got two ports on my mobo and the cable's on eth1 currently), then ran modprobe tg3.

    Though I did mkdir /var/run/netreport, I couldn't figure out how to "make sure it is chown'd to root:root and is chmod'd to 775" What command is used to verify the contents of a directory?

    rpm -qV initscripts gave me

    ---------------------------------------
    ..5....T c /etc/inittab
    missing /var/lib/stateless
    missing /var/lib/stateless/state
    missing /var/lib/stateless/writable
    .M...... /var/run/netreport
    ----------------------------------------

    So I did a mkdir on the three missing entries, then checked with rpm -qV initscripts again and it appears that the three middle items are now invisible.

    /sbin/service network start now runs without errors.

    Rather than miring myself in dependency purgatory with yum package reinstallation, I'll plan to take your advice of focusing on the login issue as the way to go. However, I appreciate your tips for selecting out the packages located in the various bins, ending in .so, etc. and I intend to search the list and choose some packages to try and reinstall later tonight.

    Learning a lot more about my system? Absolutely. If I can't recover this install, I'm at least becoming a lot more comfortable interacting with my system outside of a GUI. With all there is to know it's like treading water in the Atlantic, but one has to start somewhere.

  15. #30
    jdillick Guest
    Quote Originally Posted by kjellhansen
    There actually was no ethernet alias in etc/modprobe.conf, so I added "alias eth1 tg3" (eth1, because I've got two ports on my mobo and the cable's on eth1 currently), then ran modprobe tg3.
    Be aware that tg3 works on my system, because I have a Broadcom Corporation NetXtreme BCM5751 Gigabit Ethernet PCI Express

    You can get the details on a given driver (if depmod has been run) with the command:
    $ /sbin/modinfo tg3
    filename: /lib/modules/2.6.21-1.3228.fc7/kernel/drivers/net/tg3.ko
    version: 3.75.2
    license: GPL
    description: Broadcom Tigon3 ethernet driver
    ...snip...
    Your ethernet might very well use a different driver altogether. If you have no alias for eth0 in /etc/modprobe.conf, I'm not sure what it is loading. You can do /sbin/lspci to see the devices on you motherboard.

    Quote Originally Posted by kjellhansen
    So I did a mkdir on the three missing entries, then checked with rpm -qV initscripts again and it appears that the three middle items are now invisible.
    By the way, generally, most *nix - type operating systems have many commands that omit most/all output to indicate the command executed successfully.

    Quote Originally Posted by kjellhansen
    Learning a lot more about my system? Absolutely. If I can't recover this install, I'm at least becoming a lot more comfortable interacting with my system outside of a GUI. With all there is to know it's like treading water in the Atlantic, but one has to start somewhere.
    Just remember, every expert of any subject was once a novice. :-)

Page 2 of 3 FirstFirst 123 LastLast

Similar Threads

  1. booting Kernel stops with Kernel Panic due to missing library
    By ohommes in forum Installation, Upgrades and Live Media
    Replies: 0
    Last Post: 18th January 2009, 03:06 AM
  2. Updated Kernel Causing Kernel Panic - Image Provided
    By arpsmack in forum Using Fedora
    Replies: 7
    Last Post: 3rd September 2006, 02:31 AM

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •