Debian :: Kernel Keeps Reporting A Printk Bug In Syslog Recent Printk Recursion
Mar 5, 2010
I've got a VPS node running 2.6.26-2-openvz-amd64 and several guest machines. My kernel keeps reporting a printk bug in syslog:
[Code].....
This is somehow being replicated on the guests machines. I've read that printk is a kernel function that prints out a message, during bootup Has anyone experienced similar problem or have a more experience with this? Does it have negative impact to the host performance?
Kernel driver function was called via ioctl and returned success, but when I checked the kernel display buffer with dmesg the printk message was not there. Then when you do lsmod, the driver you were calling showed "used by zero". So it seems like the kernel driver was not accessed. In the kernel driver, I had many printk statements, but nothing printed in the buffer. if the driver get accessed and what could cause this?
A recent kernel update seems to have misplaced the Kernel Headers. VMWare needs these headers and cannot find them. Attempting to run VMWARE gets the message: Kernel headers for version 2.6.31.12-0.2-desktop were not found.
I have a hardware not so popular, it's a VIA Nehemiah 1,5 Ghz i686 (a low power mobo). I've used it since many years ago.
This mobo seems to have kind of problems with new kernels. If I use the Linux 2.6.20-16-generic kernel or an older one everything is working well. But if I use a newer kernel it has a lot of problems. A lot of different problems and I tried to solve it many times but I couldn't.
For example, with the recent Ubuntu distro, the GUI freezes before starting the installation. If I install by terminal, the GUI freezes at the first start. If I install the server version of Ubuntu (or Debian) without X-server, it works for some minutes then it freezes. If I use a kernel just a bit newer (I don't remember which one... it was years ago) than 2.6.20-16-generic, it works for some hours, then it freezs. So, it is mandatory to use the kernel Linux 2.6.20-16-generic!
Years ago I installed Ubuntu (maybe 6.10) then I updated, more times, till Ubuntu 8.04.4 LTS maintaining the old kernels. So now I'm running Ubuntu 8.04.4 with the 2.6.20-16-generic kernel and everything is working perfectly. Ubuntu 8.04 is fortunatly a LTS, so it will be supported till 04/2011, but I have to think about what to do after this date.
I don't remember well, but I think I've already tried to update from 8.04 to a recent Ubuntu (without jumping any version), maintaining my old working kernel, but I had some problems... I'm not sure, maybe will try again.
Anyway the question is: how install a recent Ubuntu (let's say 10.04) with the 2.6.20-16-generic kernel?
Possible solution: 1) Install Ubuntu 10.04 with an option which specify to use this old kernel. But I really don't know how to do it.
2) Try to install Ubuntu 10.04 with just basic service, eg. without GUI and networking. Then install (or compile and install) the old kernel there. Or compile it in another computer and then install it there.
The problem is that I had a lot of problems compiling that kernel (I tried on a Ubuntu 9.10 in a virtual machine) and i couldn't complete the compilation.
Anyway, let's forget for one moment about compilation, can I extract my current kernel from my working system and then install it on the new Ubuntu 10.04 (with basic service)?
A few days ago yast did update my computer to a new kernel-desktop 2.6.31.12-0.1.1.
My mainboard is an ASUS P4P800-VM with Intel ICH5 chipset. This new kernel is not able, to poweroff my computer. Older kernels were able to poweroff the computer.
How can I roll back to the older kernel? Yast does only offer the recent kernel.
In /boot there are only files of the recent kernel.
ASUS P4P800 Intel Pentium 4 HT, 3000 MHz, Frontside 800 MHz/Cache 1 MB 3 GB Ram AGP nVidia 7600GS/512 MB, nVidia driver 190.53, installed from yast. PCI WLan TP-Link 951N (Atheros chipset, WLan N, driver ath9 was automatically installed)
Having ubuntu 8.04, when I updated my computer to the latest kernel and headers through Update Manager, an error popped up. The error stated that it wouldn't update at this time. When I tried to download again, the update was gone in Update Manager. I checked and I don't have the most recent kernel installed. I still have 2.6.24-19. Accrding to [URL]the most recent one is: 2.6.27.
where I can find a list of Fedora 9 kernel updates (i.e. the updates that come through yum update) along with the dates they were released? I'd like to know when the last kernel I have installed (2.6.25.14-108.fc9.x86_64) was publicly released.
What i'm trying to accomplish seems rather silly but is needed for my little project here.
I'm using OpenSuSE 11.2 as a media center PC and need it to login automatically(console not X). That one i accomplished without problems.
However, after login i can see all the info about the services that were started and that needs to "go away".
One could do a "clear" in the .bash_rc and or .bash_profile but it will still show the login prompt which i don't want to see either (i don't even want to see the blinking cursor as well).
Question: How do i accomplish that so that the login console shows only the "splash screen" without any output of the kernel,rc.status nor the "issue-file"... just a plain blank screen ?
The most recent kernel update, to version 2.6.38-11 has been a horror. It hangs on boot up. Sometimes I can get to the log in screen, then it hangs there as well. However, after 3-5 reboots (which is absolutely annoying to do when I just want to use my computer), it sometimes loads up finally. However, if I go to older Linux version in the Grub menu, then I can select 2.6.38-10 and it runs just fine.
So I am looking for a way to fix this problem, whether it be removing the most recent kernel update, or making 2.6.38-10 boot up by default. I noticed that the kernels are listed in the synaptic package manager, but I am afraid to remove the most recent one via synaptic. Is it as simple as removing the Linux version and headers I don't want, or are there other things I need to do in order to make 2.6.38-10 my default kernel at start-up?
I am running a headless Ubuntu 10.04 server with the 2.6.32-28-generic kernel. For what I can figure out no single direct cause I get a high load average and the following syslog output at random intervals. Generally the load average will drop back down to normal however the kernel errors will still continue What little I have been able to find has pointed to memory issues. I am not totaly convinced this is the cause as the server will be showing >50% free when the errors are happening.
I am looking for an open source syslog server which accumulate the each and every log of Windows, Solaris, Linux and network devices. Currently I am using Syslog-ng which is not fulfiling my requirement in Windows clients, as I need the logs of every action which user performed after logon.
I have installed fedora13 OS on it recently,earlier it has both windows and linux but now only fedora13. My laptop's harddisk has probably some bad sectors and a software package "Automatic Bug Reporting Tool" has reported kernel crash,gnome-panel crash. I reported them upto some extent but in totality all I can't report. How to deal with it, why is it happening, and will it be harmful for any part like CPU,RAM,HardDisk of my laptop?
I've tried the bug-reporting tool in Gnome to no avail so far. This isn't a huge problem, more a cosmetic one. I'm running Squeeze beta 2, which was a fresh install. I have a Maxtor USB portable HD, which I use to back up personal data. It auto-mounts fine, as soon as I connect it. I have full read-write access from my regular user account; but when I'm finished accessing it, when I right-click and select "Safely remove" I get an error every time. The error has to do with "unable to unmount . . . ." and I wish I had the exact text but I don't. Anyway, when I open a terminal and check the mount point, the drive *has* unmounted, so I'm not worried about data loss. But why the error? BTW, the drive is formatted as one partition with NTFS.
I'm not looking for an answer so much as I am just trying to see that someone involved in trying to get Squeeze ready to release hears about this. It's no big deal, but it looks bad.
installed squeeze, but top keeps reporting 0.00 most of the time, and def it's wrong cause we have another server same hardware but lenny, and its over 2.00 same load. what can i check or do to fix this? thx
Linux 2.6.32-5-amd64 #1 SMP Wed May 18 23:13:22 UTC 2011 x86_64 GNU/Linux
Quick background: the office in my shop is just a partitioned area so is subject to temperature and dust, just as the workshop area is. I've upgraded my desktop from an AMD Phenom II X4 955 which used to suffer overheating problems (despite regular cleaning) to an AMD FX-6300 on a Gigabyte 970A-D3P motherboard. For cooling I've fitted a Cooler Master Hyper TX3, plus various fans. Fresh install of Debian 8.3.
The cooler has certainly solved the overheating issues and the machine runs very quietly, rather than sounding like a 747 at take-off. But I'm now having some problems getting consistent reporting on temperatures.
The BIOS reports temperatures from the CPU which seem to be fairly consistently in the mid to high 30s (C).
lm-sensors and hddtemp have been installed.
sensors-detect reports (just the last section) Code: Select allNow follows a summary of the probes I have just done. Just press ENTER to continue:
Driver `fam15h_power' (autoloaded): * Chip `AMD Family 15h power sensors' (confidence: 9)
[Code].....
psensor identifies AMD CPU and the NVidia GPU, apparently correctly, as it does the HDDs.
So, my problem is that the CPU temps reported by psensor and the panel app vary quite significantly with those that are reported in the BIOS. My thinking is that the BIOS is correct and that the software is either misreporting the temperatures or using the wrong sensors.
On the old machine I used to get a temperature from each core, just like in the BIOS, but now I'm only getting a single CPU reading.
I've installed Debian6 Testing KDE4 using a net-inst CD when it is recently released. Today , I did a aptitude clean && aptitude update && aptitude upgrade and got a weird output. aptitude clean && aptitude update ... ... Current status: 62 updates [+61], 1561 new [+270]. There are 7 newly obsolete packages: librasqal2, libwpd8c2a, libwpg-0.1-1, libwps-0.1-1, openoffice.org-base-core, openoffice.org-core, openoffice.org-report-builder-bin aptitude upgrade Resolving dependencies... open: 42543; closed: 32760; defer: 49; conflict: 8 The open/closed figures is still running (higher) after 10 minutes. Believe I am not the only one using Debian Testing with KDE4.
I am attempting to get Debian Lenny 5.0.8 to run on an OpenSolaris snv_134 lx-branded zone. Debian 6.0.0 is a total flop due to the included udev requiring a kernel >= 2.6.26 and the kernel provided by the lx26 brand being only 2.6.18.So I installed Debian 5.0.8 into a VirtualBox VM, configured it, then tarred up the install with the following command as root, from /:
tar -cjf /mnt/share/debian-5.0.8.tar.bz2 --exclude dev --exclude sys --exclude mnt --exclude proc *
I then installed that archive into a solaris zone, booted it, and had to run
perl -pi -e "s,1 1,1 0," /etc/fstab
in order to fix up the fstab. Now it boots (with a handful of warnings), but the problem I'm having is that I can't run apt. I get:
E: Couldn't determine free space in /var/cache/apt/archives/ - statvfs (75 Value too large for defined data type)
Today morning I logged on to one of our servers (through ssh -X) to perform a routine maintenance. When I tried to open a GUI application, it failed to connect to the X server of the workstation. Eventually I ended up issuing the command
Code:
df -h /
and shockingly it showed that / is 100% used. I checked / using
Code:
du --max-depth=1 -xh /
to check the sizes of the individual directories, which showed that only about 18% of the / is used and that confused me badly. There was no quick solutions when I googled around; but luckily I found the following link which nicely explained the issue I had:
[URL]
All I did was found out the services responsible for those unreleased files (using lsof +L1) and restarted them. That is it. Now df reports only 18% of the disk is used and all my X things started working again, thanks to Walker.
According to this article -- [URL] there are problems with FireFox security, as a veriety of Zeus is being used by crackers in Europe and China. Allowing for alarmist reporting, this still looks unpleasant. Does anyone have definitive information about how this attack affects the various operating systems?
wants to remove everything else that (presumably) has syslog as a dependency. how do I replace the dependency on syslog with a dependency on syslog-ng?
I have a server provided dns and sendmail service, dns keeping record for my public servers. I don't want my server to be open dns server, so I just add recursion and forwarders option to named.conf, like: allow-recursion {192.168.x.x/24; 123.123.x.x/28; 127.0.0.0/8;}; forwarders {publicDNSipA; publicDNSipB;}; After restart named, all sendmail user can't mail to other internet account (like xxxxx@gmail.com); but local account is ok
I noticed in my system that my root partition is getting full. I found a lot of old compacted syslogfiles. Had a look at etc/sysconfig editor eg cron but could not find a setting which allows to delete files older than a month. Where and how could I influence this ? I deleted manually all syslog files older than a month. Approx 6GB
