help! can't log in system

[erossunny]

here is my boot.log find in /var/log/boot.log:
I find this with my ubuntu, I have two system.

[1;31mUnable to fix label of /sys/fs/cgroup/systemd: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/cpuset: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/ns: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/cpu: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/cpuacct: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/memory: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/devices: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/freezer: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/net_cls: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/blkio: No such file or directory[0m

Welcome to [0;34mFedora release 15 (Lovelock)[0m!

Starting Collect Read-Ahead Data...
Starting Replay Read-Ahead Data...
Starting Bootup hack...
Starting Syslog Kernel Log Buffer Bridge...
Started Syslog Kernel Log Buffer Bridge.
Starting Media Directory...
Started Lock Directory.
Started Runtime Directory.
Started Collect Read-Ahead Data.
Started Replay Read-Ahead Data.
Starting Load legacy module configuration...
Started File System Check on Root Device.
Starting Remount Root FS...
Starting Remount API VFS...
Starting Apply System Clock UTC Offset...
Starting Apply Kernel Variables...
Started Load Kernel Modules.
Starting Setup Virtual Console...
Started Set Up Additional Binary Formats.
Started Media Directory.
Starting Stdio Syslog Bridge...
Started Stdio Syslog Bridge.
Started Bootup hack.
Started Load legacy module configuration.
Started Remount Root FS.
Started Remount API VFS.
Started Apply System Clock UTC Offset.
Started Apply Kernel Variables.
Started Setup Virtual Console.
Starting udev Kernel Device Manager...
Starting Configure read-only root support...
Started Configure read-only root support.
Started udev Kernel Device Manager.
Starting udev Coldplug all Devices...
Started udev Coldplug all Devices.
Starting udev Wait for Complete Device Initialization...
Started udev Kernel Device Manager.
Started udev Wait for Complete Device Initialization.
Starting Wait for storage scan...
Started Show Plymouth Boot Screen.
Started Wait for storage scan.
Starting Initialize storage subsystems (RAID, LVM, etc.)...
Starting Huge Pages File System...
Started Huge Pages File System.
Starting POSIX Message Queue File System...
Started POSIX Message Queue File System.
Started Initialize storage subsystems (RAID, LVM, etc.).
Starting Initialize storage subsystems (RAID, LVM, etc.)...
Started Initialize storage subsystems (RAID, LVM, etc.).
Started Reconfigure the system on administrator request.
Started Relabel all filesystems, if necessary.
Started Mark the need to relabel after reboot.
Starting Enable all detected swap partitions...
Starting Recreate Volatile Files and Directories...
Starting Tell Plymouth To Write Out Runtime Data...
Starting Load Random Seed...
Started Enable all detected swap partitions.
Started Load Random Seed.
Started Tell Plymouth To Write Out Runtime Data.
Started Recreate Volatile Files and Directories.----------------------------------->It just stopped here

by the way, I delete the filesystem label of fedora volume in ubuntu with Disk Utility. Is that the reason? I can't remember the original label now.
I installed fedora use Fedora-15-Beta-x86_64-Live-Desktop.iso

[GoinEasy9]

This is due to the newest kernel kernel-PAE-2.6.38.4-20.fc15.i686, I just updated and got the same messages. Boot up into the previous kernel and the problem goes away. Or you can change to default=1 in /boot/grub/grub.conf until a new kernel is released. Or remove the new kernel "yum remove kernel-PAE-2.6.38.4-20.fc15.i686".
Ah just noticed you're on 64 bit, nice, so it isn't just 32 bit or 32 bit PAE kernels.

Edit: I must say it is nice seeing more info in boot.log now.

---------- Post added 30th April 2011 at 11:15 AM ---------- Previous post was 29th April 2011 at 11:55 PM ----------

I seemed to have been wrong, even the previous kernel is now getting those errors. I did a "yum history undo last" and the problems still persist. It removed kernel and perf, the only things that got updated last night in the first place. Before that everything was running fine. Anyone else know what's going on?

[jwmueller]

I just completed the 64-bit updates to kernel 2.6.38.4-20.fc15.x86_64 and am now experiencing long delays in the boot process. Found this in /var/log/boot.log amongst the normal startup messages:

Starting Cryptography Setup for luks-eaca5e23-6931-4c48-bb09-f6aa29a9564c aborted because a dependency failed.

I have LUKS encryption on my root ( / ) directory, which is what this UUID refers to.

I booted with live USB stick, performed cryptsetup luksOpen on this partition and ran fsck without incident.

Are any of you seeing this? If not, I can open a new thread.

JW.

---------- Post added at 02:07 PM ---------- Previous post was at 01:02 PM ----------

Update, I have another system (netbook) on F15 that still needed the latest updates and hasn't been showing startup problems. Including the kernel update, it has 60 updates waiting install. I'll update a product at a time to see if/when the boot slowdown gets introduced. Tho the netbook is 32-bit... kernel installed and system booted without delays. On to the other updates to see if I hit the problem or maybe it doesn't affect 32-bit.

jw.

[GoinEasy9]

I'm on 32 bit and it all started for me last night when only the new kernel and perf came in. Today's updates didn't help. The error claiming labeling error is where I'm focused. Especially since reverting to the old kernel didn't help.

[jwmueller]

If those were the only 2 updates, maybe it's "perf". That's one I'll get to in my list later. Based on this, I will plan to do chkconfig and perf as the last 2 individual updates. Will get there in about 30-45 minutes.

jw

[GoinEasy9]

I'm curious if one of your updates is for a new selinux policy, and if that causes the problem. BTW - all these files/directories exist, why can't they be relabeled?
[1;31mUnable to fix label of /sys/fs/cgroup/systemd: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/cpuset: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/ns: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/cpu: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/cpuacct: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/memory: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/devices: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/freezer: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/net_cls: No such file or directory[0m
[1;31mUnable to fix label of /sys/fs/cgroup/blkio: No such file or directory[0m

Another question, are you using KDE? My secure file is looking for /lib/security/pam_gnome_keyring.so and it's not there. This is a kdm call.

---------- Post added at 06:23 PM ---------- Previous post was at 06:14 PM ----------

You mentioned systemd from koji in the other thread, was this the answer?

[jwmueller]

On my 32-bit system, I ran into problems after updating policycoreutils* . Yum was upgrading to 2.0.86-7 for 3 items. Booted to init 3 and yum downgrade to 2.0.85-27 got me going again.

However, on my 64-bit system, initscripts (and initscripts-legacy) appeared to be causing my problem.

Now, on both, 32- and 64-bit systems, I'm current except for policycoreutils* and initscripts*. Both of those groups of products are 1 level back. No hangs on either system.

I don't have the "perf" package installed on my 32-bit system, so it wasn't involved for me.

jw

[GoinEasy9]

Never mind, updating to the most recent systemd, got rid of the errors. I knew I read somewhere that systemd used cgroups to monitor processes, I should have known the labeling wasn't an selinux issue.

Just saw your last post, I guess you have other issues, but, I thought that perf got installed by default. I didn't install it manually.

[jwmueller]

selinux didn't play a part for me... I'm running with selinux=0 on my grub kernel statement and /etc/sysconfig/selinux has SELINUX=disabled.

jw

[GoinEasy9]

From what I can see, the new kernel should have been accompanied by the new systemd. When I tried booting into the previous kernel, I still got the hang at the end of boot, but using the new kernel it doesn't happen. At least that's what I'm noticing without further testing. Right now I just want to go back to other things I'm trying, I'll let the devs figure out the core issues. I wonder if everyone else's boot and shutdown problems could be solved by getting the new systemd from koji?

[peteraltherr]

hi there, i have exactly the same problem as jwmueller in post #3. i have a beta installation of fedora 15 x86_64 on an external usb harddisk and the whole system is luks/lvm encrypted. yesterday everything was fine and while i wanted to start the system this morning it does not boot properly and gives me several error messages like:

Starting Cryptography Setup for luks-... aborted because a dependency failed.

i have several years of experience with ubuntu and luks/lvm encryption. also i was able to mount the fedora root file system using my ubuntu machine. the problem is, i have no clue how to fix this as in this early stage of the boot process there is still no terminal session available. at least according to my unterstanding and the failed tries i gave ctrl-alt-f2.

is there any chance to get the system updated so it reboots again properly.


thanks a lot

peter (a pretty rookie in terms o fedora)


ps: what made me thinking about the future use of fedora instead of my loved ubuntu can be expressed in a few words: 1. unity, 2. poor gnome3 support

[jwmueller]

Peter, what happens if you just wait a couple minutes? On my x86_64 system, after I get the crypt setup error, the system completes startup as normal. If you get that far, you can open a terminal and downgrade the initscripts packages:
su -
yum downgrade initscripts*

reboot and the delay should be gone.

If your system doesn't continue, there may be something else at play, too.

While I mention the policycoreutil* packages, they cause problems on my 32-bit system, but not the 64-bit one.

jw

[peteraltherr]

hi jw, thanks a lot for your suggestion. after your reply i gave it another try and started the machine. then i have waited for more than one hour but the system is still stuck on that error message. the only action i can do is hitting ctrl-alt-f2 and then i get another error message from the plymouth demon that there is something wrong and that's it.

sorry for not having better news. any more suggestions what to do? everything is appeciated :-)


peter

[jwmueller]

I'm sorry I can't help any further... can only share the experiences I've had.

jw

[peteraltherr]

hey jw, no problem. my next step would be booting into a live fedora 15, mounting the lvm/luks device and then trying to chroot it in order to be able to perform some yum updates. i hope this has been fixed meanwhile and a simple update gets the system back running. the only thing i have to do is learning how to chroot. i never did it by myself but i have read a lot about it.


greetings

peter

---------- Post added at 07:26 PM ---------- Previous post was at 03:52 PM ----------

at the moment i am updating the faulty system on my external usb hdd by using the fedora 15 beta live-cd on a usb thumbdrive. i have chrooted into the encrypted system after mounting successfully the lvm/luks partitions and right now yum is (hopefully) updating all packages (303mbytes). later on i will give you a report about the result.

peter

---------- Post added at 07:26 PM ---------- Previous post was at 07:26 PM ----------

at the moment i am updating the faulty system on my external usb hdd by using the fedora 15 beta live-cd on a usb thumbdrive. i have chrooted into the encrypted system after mounting successfully the lvm/luks partitions and right now yum is (hopefully) updating all packages (303mbytes). later on i will give you a report about the result.

peter