[High cpu loads] with CentOS + 2.6.9-22/2.6.9-34 linux kernel

[eva2000]

Update: Oct 19, 2006: - Seems CentOS related 2.6.9-42.0.3 causes the same load issue as before so stick with 2.6.9-42.0.2 for now http://www.vbulletin.com/forum/showthread.php?p=1233538#post1233538

Fix: Looks like CentOS 2.6.9-39 kernel is out which possibly fixes this issue, read http://www.vbulletin.com/forum/showpost.php?p=1151870&postcount=64 and http://bugs.centos.org/view.php?id=1174 and http://www.vbulletin.com/forum/showpost.php?p=1151875&postcount=65

Update: seems to occur sometimes with 2.6.9-34 kernel as well as 2.6.9-22 kernel http://www.vbulletin.com/forum/showpost.php?p=1095380&postcount=49

Lately, there has been alot more incidents of high cpu loads on servers with CentOS linux OS using 2.6.9-22 linux kernel

i.e.
- http://www.vbulletin.com/forum/showthread.php?p=1051263#post1051263
- http://www.vbulletin.com/forum/showpost.php?p=1054734&postcount=16


vB customers who had this issue reported it being fixed by downgrading their linux kernel to 2.6.9-11 instead. However, this may open you up to security exploits and bugs. So you may want to try upgrading to 2.6.15 linux kernel if your web host can oblige.

It's best to get your web host to do the kernel upgrade/downgrades since if you mess up you could render your server unbootable heh.

How to find want kernel you're using:

2 ways to find out

1. in ssh telnet type

uname -a

2. load phpinfo.php url and look at the top line under System
Linux yourhostname 2.6.13.1 #3 SMP Sun Oct 16 20:08:38 EDT 2005 i686


Please, post a reply if you experienced the exact same issue using CentOS and 2.6.9-22 linux kernel where downgrading or upgrading fixed the high cpu load issues etc.



Update:

A possible fix for the CentOS kernel issue in post #7 at http://www.centos.org/modules/newbb/viewtopic.php?topic_id=2287

[mentalrz]

2.4.21-37 i use and i get random high loads regardless of activity , but since i ordered a new opteron, who cares =) dont use xeons..


PS, thanks for the info.

[turkforum]

Linux web1.turkforum.net 2.6.9-22.ELsmp #1 SMP Sat Oct 8 21:32:36 BST 2005 x86_64 x86_64 x86_64 GNU/Linux

same here

But dont know why i have load right now it can be or possiblly mis configuration

[eva2000]

You using centOS 64bit ... 64bit has issues and centos with that kernel has issues... i'd request your web host to change you to CentOS 32bit OS with more recent kernel than 2.6.9-22 smp i.e. 2.6.15

[Abe Babe]

I recently had this issue and had posted in this forum with a server optimization request. We were experiencing very slow load times [often 5 minutes per page and sometimes more]. We tried many different things including suggestions from eva2000 and other things, but nothing worked.

On his suggestion of getting the kernel downgraded, since then the forum has been running at expected speed, if not better than the previous server we were on [current server is Pentium 4, 3.8 Ghz, 1GB ram], and loads have decreased dramatically.

Although just a further question, we got our kernel downgraded to 2.6.9. Are you aware of any security issues with this version? I don't really want to go back and ask for an upgrade again so soon after asking for a downgrade, but on reading the comment here I want to make sure the version we are running is okay. Plus now it is fixed, I'm hesitant to get more changes done.

Abe Babe...

[eva2000]

There might be security issues hence why kernel's were updated... maybe try 2.6.15 and see ?

[ubuntu-geek]

Centos 4.2, with kernel 2.6.9-22 causes serve high loads. Down graded in the meantime and loads are normal. Just wanted to confirm..

[pulsorock]

Recently an update for the kernel was released:
Linux localhost 2.6.9-22.0.2.ELsmp #1 SMP Tue Jan 17 07:10:04 CST 2006 i686 i686 i386 GNU/Linux Anyone knows if this one performs better?

I just upgraded to it... and it seems to run normal. It has its peak loads like always, but I guess it has to do with heavy traffic at specific moment. I yet have to see it on a very heavy load time (usually at night). See the current uptime results:

[root@localhost tmp]# uptime
14:20:05 up 40 min, 1 user, load average: 0.15, 0.22, 0.18
[root@localhost tmp]# uptime
14:23:36 up 43 min, 1 user, load average: 0.01, 0.12, 0.15
[root@localhost tmp]# uptime
14:28:31 up 48 min, 1 user, load average: 0.17, 0.14, 0.15
[root@localhost tmp]# uptime
14:29:50 up 50 min, 1 user, load average: 0.41, 0.21, 0.17
[root@localhost tmp]# uptime
14:29:56 up 50 min, 1 user, load average: 0.46, 0.22, 0.17
[root@localhost tmp]# uptime
14:33:12 up 53 min, 1 user, load average: 0.31, 0.22, 0.18
[root@localhost tmp]# uptime
14:45:42 up 1:05, 1 user, load average: 0.05, 0.19, 0.18
[root@localhost tmp]# uptime
14:55:02 up 1:15, 1 user, load average: 0.11, 0.11, 0.13
[root@localhost tmp]# uptime
14:55:25 up 1:15, 1 user, load average: 0.20, 0.13, 0.14
[root@localhost tmp]# uptime
14:58:53 up 1:19, 1 user, load average: 0.09, 0.09, 0.11
Has anyone else running CentOS 4.2 upgarded to this latest kernel?

[scottct1]

I am running Linux uplink.satelliteguys.us 2.6.9-22.0.1.ELsmp #1 SMP Thu Oct 27 14:49:37 CDT 2005 x86_64 x86_64 x86_64 GNU/Linux

Is there anything I need to do or am I ok?

Here are my loads with 1100 people online.

top - 08:04:30 up 10:56, 1 user, load average: 0.29, 0.36, 0.35
Tasks: 181 total, 4 running, 177 sleeping, 0 stopped, 0 zombie
Cpu(s): 12.1% us, 1.2% sy, 0.0% ni, 82.3% id, 4.1% wa, 0.0% hi, 0.2% si
Mem: 8112016k total, 2635592k used, 5476424k free, 172784k buffers
Swap: 2040244k total, 160k used, 2040084k free, 1136556k cached

[speedway]

You using centOS 64bit ... 64bit has issues and centos with that kernel has issues... i'd request your web host to change you to CentOS 32bit OS with more recent kernel than 2.6.9-22 smp i.e. 2.6.15
George

We are going to CentOS 64bit on dual Opterons, are you saying that moving to the 64bit operating system is not a good idea?

Cheers
Bruce

[eva2000]

I am running Linux uplink.satelliteguys.us 2.6.9-22.0.1.ELsmp #1 SMP Thu Oct 27 14:49:37 CDT 2005 x86_64 x86_64 x86_64 GNU/Linux

Is there anything I need to do or am I ok?

Here are my loads with 1100 people online.

top - 08:04:30 up 10:56, 1 user, load average: 0.29, 0.36, 0.35
Tasks: 181 total, 4 running, 177 sleeping, 0 stopped, 0 zombie
Cpu(s): 12.1% us, 1.2% sy, 0.0% ni, 82.3% id, 4.1% wa, 0.0% hi, 0.2% si
Mem: 8112016k total, 2635592k used, 5476424k free, 172784k buffers
Swap: 2040244k total, 160k used, 2040084k free, 1136556k cached
it's only a problem if yo have load issues

George

We are going to CentOS 64bit on dual Opterons, are you saying that moving to the 64bit operating system is not a good idea?

Cheers
Bruce

for now it's better if you use 32bit centOS ... although i hear Suse Enterprise 64bit is probably best for dual opteron rigs.. the real pain is 64bit hardware linux drivers

[speedway]

it's only a problem if yo have load issues



for now it's better if you use 32bit centOS ... although i hear Suse Enterprise 64bit is probably best for dual opteron rigs.. the real pain is 64bit hardware linux drivers Thanks George

[BoYagoob]

eva2000,
I have a question..
do you know anybody who upgraded to 2.6.15 linux kernel so it worked fine and fast with his forum?

Regards

[eva2000]

eva2000,
I have a question..
do you know anybody who upgraded to 2.6.15 linux kernel so it worked fine and fast with his forum?

Regards
Not personally, i've seen a few with 2.6.14 with no probs so assume same for 2.6.15

[roscoe36]

Linux rip.xxxxxxxx.com 2.6.9-22.0.1.EL #1 Thu Oct 27 12:26:11 CDT 2005 i686

I've experienced some loading spiralling out of control until a reboot was forced.

Is my version outdated? I'm no expert at this stuff. This OS was just loaded less than 2 weeks ago.

[eva2000]

you using CentOS for linux ?

[roscoe36]

Yep, that's what this thread is about right :confused:

[eva2000]

then either downgrade to 2.6.9-11 or better yet upgrade to 2.6.15 kernel

[eva2000]

A possible fix for the CentOS kernel issue in post #7 at http://www.centos.org/modules/newbb/viewtopic.php?topic_id=2287

[pulsorock]

A possible fix for the CentOS kernel issue in post #7 at http://www.centos.org/modules/newbb/viewtopic.php?topic_id=2287 I just changed from

127.0.0.1 mydomain.com localhost
12.174.245.206 mydomain.com

to


127.0.0.1 localhost
12.174.245.206 mydomain.com


Currently running:
Linux mydomain.com 2.6.9-22.0.2.ELsmp #1 SMP Tue Jan 17 07:10:04 CST 2006 i686 i686 i386 GNU/Linux

Lets see if that helps.... Thanks! :)

[neuromancer11]

Linux host.******.com 2.6.9-22.0.1.ELsmp #1 SMP Thu Oct 27 13:14:25 CDT 2005 i686 i686 i386 GNU/Linux


I have a load problem with that kernel. How can I downgrade my kernel ?

[Hackhell]

hosts.conf doesn't fix the issue for me i have a scsi card that does not boot with vanilla kernel and all of the 2.6.9.11 rpm kernels are deleted from repos is there any one that can send or upload the 2.6.9.11 rpm kernel so we can use it

[Hackhell]

i found the kernel keep in mind it is a vulnerable kernel and if you dont have any chance downgrade myscsi card dont like the vanilla kernel

just do
wget ftp://ftp.nluug.nl/pub/ibiblio/distributions/e-smith/devel/repo/RPMS/kernel-2.6.9-11.EL.i686.rpm

rpm --force -ivh kernel-2.6.9.11.EL.i686.rpm

then check /etc/grub.conf anor /etc/lilo.conf dont make this kernel as a default kernel boot once with that kernel

[patriotcow]

Yup I have this kernal, looks like I am not the only one who randomly gets high load and the system gets totally slow till you can't login.

[DefenceTalk]

Same issue here:


Linux host.xxxxxx.com 2.6.9-22.0.2.EL #1 Tue Jan 17 06:51:40 CST 2006 i68
6 athlon i386 GNU/Linux

[minzhong]

then either downgrade to 2.6.9-11 or better yet upgrade to 2.6.15 kernel

how do i upgrade? i'm on 2.6.9-22 and my server keeps crashing when the users online hit ~200.

I'm on P4 2.8GHz + HT :confused:

[minzhong]

Do i just comment (#) out the first part of the title ?

# grub.conf generated by anaconda
#
# Note that you do not have to rerun grub after making changes to this file
# NOTICE: You have a /boot partition. This means that
# all kernel and initrd paths are relative to /boot/, eg.
# root (hd0,0)
# kernel /vmlinuz-version ro root=/dev/hda3
# initrd /initrd-version.img
#boot=/dev/hda
default=0
timeout=5
splashimage=(hd0,0)/grub/splash.xpm.gz
hiddenmenu
title CentOS (2.6.9-22.ELsmp)
root (hd0,0)
kernel /vmlinuz-2.6.9-22.ELsmp ro root=LABEL=/
initrd /initrd-2.6.9-22.ELsmp.img
title CentOS (2.6.9-22.EL)
root (hd0,0)
kernel /vmlinuz-2.6.9-22.EL ro root=LABEL=/
initrd /initrd-2.6.9-22.EL.img
title CentOS-4 i386 (2.6.9-11.ELsmp)
root (hd0,0)
kernel /vmlinuz-2.6.9-11.ELsmp ro root=LABEL=/
initrd /initrd-2.6.9-11.ELsmp.img
title CentOS-4 i386-up (2.6.9-11.EL)
root (hd0,0)
kernel /vmlinuz-2.6.9-11.EL ro root=LABEL=/
initrd /initrd-2.6.9-11.EL.img

[eva2000]

To be safe get your web host to do the upgrade... i always do :)

[minzhong]

Hello. I'm on an unmanaged server. How do I upgrade .. I ran 'yum update' what's next?

[minzhong]

root@server [~]# yum update
Repository base is listed more than once in the configuration
Repository addons is listed more than once in the configuration
Repository extras is listed more than once in the configuration
Repository extras is listed more than once in the configuration
Repository updates is listed more than once in the configuration
Repository base is listed more than once in the configuration
Repository addons is listed more than once in the configuration
Setting up Update Process
Setting up repositories
Reading repository metadata in from local files
Excluding Packages in global exclude list
Finished
No Packages marked for Update/Obsoletion

[minzhong]

finally figured out everything. thanks for the guide.

[NickFritz]

I'm getting that problem too.

Linux naf-server01.com 2.6.9-22.0.2.ELsmp #1 SMP Tue Jan 17 07:10:04 CST 2006 i686 athlon i386 GNU/Linux

I have edited my hosts file and will see if it still happens as it happens daily at peak time.

I would also like to know how to upgrade to the newer kernel as all i can find in yum is the kernel version i am using.

Is there a tutorial on upgrading kernel in yum for centos?

[minzhong]

I downgraded mine to 2.6.9-11 .. server load issues are now gone!! You might want to check whether 2.6.9-11 is available in your server

[NickFritz]

No its not available as its a brand new box.

[Gamingforce]

I have successfully upgraded to CentOS (2.6.15.4) using this easy to use guide!

http://www.webhostingtalk.com/showpost.php?p=3691073&postcount=6

Check it out.

[eva2000]

I have successfully upgraded to CentOS (2.6.15.4) using this easy to use guide!

http://www.webhostingtalk.com/showpost.php?p=3691073&postcount=6

Check it out.
excellent thanks alot of folks will be greatful :)

has it fixed high load/iowait issue for you ?

edit: note that wht thread has a correction at http://www.webhostingtalk.com/showpost.php?p=3691175&postcount=7

sorry,

cp /boot/config-2.6.9-22.0.2.EL /usr/src/linux-2.6.15.4/.config

should instead be:

cp /boot/config-2.6.9-22.0.2.EL .config

[dizzi]

I have the same problem with high load in Centos 4x and the 2.6.9-22 kernel ,my server got high load everyday for over a month since i have it, After checked this thread i decided upgrade and ask my host upgrade to 2.6.15.4 kernel two day ago, and the server run smooth even at peak hour with no high load or slow down ,it definitely fixed the highload headache . Thank Eva and everyone else have show the tip in here.

[eva2000]

Excellent news dizzi .. glad i could help :)

[grecostimpy]

Having lots of issues lately with server loads and forced reboots...

Linux xxxxxxxxxxx 2.4.20-13.9ensim-3.5.0-14 #1 Wed Nov 24 05:08:31 PST 2004 i686

[eva2000]

Having lots of issues lately with server loads and forced reboots...

Linux xxxxxxxxxxx 2.4.20-13.9ensim-3.5.0-14 #1 Wed Nov 24 05:08:31 PST 2004 i686
if this is a separate forum and issue from original poster please start your own thread with info asked at http://www.vbulletin.com/forum/showthread.php?t=70117

[Joe Gronlund]

I am running 2.6.9-5.0.3.EL with no problems

[Stevez]

Hi,

Im using 2.6.11.9-grsec which i have downloaded from kernel.org and complied myself with grsecurity..

I also get massive loads at times (80+).

Anyone know if this kernel would be effected?

[Mahilu]

...
Please, post a reply if you experienced the exact same issue using CentOS and 2.6.9-22 linux kernel where downgrading or upgrading fixed the high cpu load issues etc.
...


I was experiencing similar issues on a very busy forum and upgrading the Centos Kernel 2.6.15 seems to have helped considerably. Load average dropped from around 4 to around 1. I did the upgrade manually, following the guide posted here, and everything worked fine

http://www.webhostingtalk.com/showpost.php?p=3691073&postcount=6

[Gavin.]

Is this still a problem in the 2.6.14 kernal? I'm having a lot of troubles with the server reaching the point of not even accepting an SSH connection, and it's a dual xeon!

[eva2000]

Is this still a problem in the 2.6.14 kernal? I'm having a lot of troubles with the server reaching the point of not even accepting an SSH connection, and it's a dual xeon!
not sure... depends on each server i guess

[Marco van Herwaarden]

Not sure what it is worth, but the following was posted at the CentOS forum:

DO NOT go higher than Linux r83pc66.rentokil.com 2.6.9-22.0.1.EL because the kernel has been compiled with a version of the gcc compiler that has not been released yet and you will not be able to compile or link in any third party kernel modules such as those required by vmware player server or workstation
http://www.centos.org/modules/newbb/viewtopic.php?topic_id=3161&forum=31 (post #3)

[Marco van Herwaarden]

I seem to be running a 2.6.9-34.ELsmp kernel on my new CentOS4.2 box. Anyone has experiences with this kernel version?

[Marco van Herwaarden]

Well using the 2.6.9-34.ELsmp kernel on my new Dual Xeon 2.4 server with CentOS 4.2/ 1Gb RAM give me the same symptoms. High loads occasionally, but also the box is swapping more then it should. Not sure if this is caused by the kernel problem or other config problems. (used the same apache/php & mysql config as i used on the previous P4 box with 1Gb).

Will try to upgrade to the .15 kernel this weekend to see if it makes a difference.

[Marco van Herwaarden]

I upgraded yesterday from 2.6.9-34 to the 2.6.15.6 kernel, and loads seem to have dropped dramatic, no more spikes and swap is almost 0. So sounds promissing.

[pulsorock]

I have my questions... doesn’t manually compiling a newer kernel affects CentOS update system? (ie: yum update)

What are the pros and cons about this?

[pulsorock]

CentOS 4.3 has been released
http://lists.centos.org/pipermail/centos-announce/2006-March/001261.html

I just updated mine via "yum update"

[eva2000]

thanks for the heads up... can't find what kernel version 4.3 comes with out of the box ?

[pulsorock]

thanks for the heads up... can't find what kernel version 4.3 comes with out of the box ?
This is what I got:

Linux domain.com 2.6.9-34.ELsmp #1 SMP Wed Mar 8 00:27:03 CST 2006 i686 i686 i386 GNU/Linux

[eva2000]

thanks for that :)

[rmd708]

Anyone have the same problems with this kernel?

Linux domain.com 2.6.9-34.ELsmp #1 SMP

My new server comes with this as default, so I am thinking of upgrading to

2.6.15.4 #1 SMP

[Marco van Herwaarden]

Anyone have the same problems with this kernel?

Linux domain.com 2.6.9-34.ELsmp #1 SMP

My new server comes with this as default, so I am thinking of upgrading to

2.6.15.4 #1 SMPSee my posts on this page.

[rmd708]

Does anyone know the difference between kernel 2.6.15.6 and kernel 2.6.15.4? I installed .4 (as written in most of this thread) without realizing there is a newer version.

[eva2000]

kernel.org has change logs for versions but the changelogs are normally pretty log (lots of bug fixes, tweaks and changes but nothing major to warrant major version jumps hence)

[RoutineX]

Yeah, was wondering the same thing. I was running 2.6.9-22 with constant high load problems, and went up to 2.6.9-34, but don't notice any difference.
Do we need a lower version perhaps?

[eva2000]

Yeah, was wondering the same thing. I was running 2.6.9-22 with constant high load problems, and went up to 2.6.9-34, but don't notice any difference.
Do we need a lower version perhaps?
see post #49 http://www.vbulletin.com/forum/showpost.php?p=1095380&postcount=49

[pulsorock]

I upgraded yesterday from 2.6.9-34 to the 2.6.15.6 kernel, and loads seem to have dropped dramatic, no more spikes and swap is almost 0. So sounds promissing.
How did you upgraded?
I see it's been three months since you did it... How has it work during all that time? Have you had any problems?

I have an 2.8GHz Single Xeon/ CentOS 4.3 and 2GB RAM (Dell PowerEdge 1800)

I do all my updates via "yum update". Would doing this upgrade damage any of the yum update functionality? If for any reason (ie. I got some errors) can I go back and how?

Thanks...

[Marco van Herwaarden]

Are you running a control panel like cPanel or Plesk? If yes, i suggest you also check inon there support forums.

To upgrade using yum, see http://www.webhostingtalk.com/showthread.php?p=3913165
^^^ didn't try this, so i can't say if it is correct.

I upgraded by rpm. Most of the steps outlined here: http://www.webhostingtalk.com/showthread.php?t=489019 (read starting at post #6)
The thread is a bit older, and you probably want to look what the current kernel version is and use that instead of the kernel they suggest there.

I can't garantee that a kernel upgrade will not damage something. You should know what you are doing and make good backups. With most install methods, you will end up with a new kernel sitting next to your old one, and if 1 don't work you should be able to easy switch back to the old one.

[kmike]

I've successfully used 2.6.16 FC4 kernels in RHEL4 which is what CentOS 4.3 based off AFAIK, so you can try to download and install them using rpm.

[mlx]

It's amazing how crappy 2.6.9-34.0.1 (released 2006/05/24) still is.

We have finally moved a quite busy vBulletin yesterday to a new CentOS 4.3 server which was running that 2.6.9-34.0.1 kernel and everything looked fine for half an hour or something. After that the load went through the sky.

Luckily I found that thread here :)

As suggested at http://bugs.centos.org/view.php?id=1174 we have upgraded to 2.6.9-39 now and it looks like it solved those problems pretty much completely. It's available since 2006/06/15 at http://dev.centos.org/centos/4/testing/i386/RPMS/.

Everyone who downgraded should consider upgrading using those RPMs I guess!

[mlx]

Use at your own risk (worked fine for me)

First of all get the correct RPMs for your server.

If you have a dual processor server or a P4 CPU that has hyperthreading technology:
# wget http://mirror.centos.org/centos-4/4.5/os/i386/CentOS/RPMS/kernel-smp-2.6.9-55.EL.i686.rpm
# wget http://mirror.centos.org/centos-4/4.5/os/i386/CentOS/RPMS/kernel-smp-devel-2.6.9-55.EL.i686.rpm

If you have a single processor server:
# wget http://mirror.centos.org/centos-4/4.5/os/i386/CentOS/RPMS/kernel-2.6.9-55.EL.i686.rpm
# wget http://mirror.centos.org/centos-4/4.5/os/i386/CentOS/RPMS/kernel-devel-2.6.9-55.EL.i686.rpm

Now install them ... but do not use 'rpm -U'! 'rpm -i' will keep your old kernel(s), installing the new one instead of upgrading. This will make sure you can go back in case of any errors.

If you have a dual processor server or a P4 CPU that has hyperthreading technology:
# rpm -ivh kernel-smp-devel-2.6.9-55.EL.i686.rpm kernel-smp-2.6.9-55.EL.i686.rpm

If you have a single processor server:
# rpm -ivh kernel-devel-2.6.9-55.EL.i686.rpm kernel-2.6.9-55.EL.i686.rpm

The server needs to be rebooted to load the new kernel now, so first of all let's edit the grub bootloader options now.

(applies to both single and dual processor servers)
# nano /boot/grub/grub.conf

Installing the kernel should put the new one on top, having the old kernel(s) below. Make sure that the default says "default=1". Grub starts at 0. So the second kernel in the list would be your default (1). You can change it back to 0 later if the reboot is working fine. For now let's configure grub to boot the new kernel only once making sure that if it fails and the datacenter has to do a reboot it will bring back your old default kernel.

(applies to both single and dual processor servers)
# grub
> savedefault --default=0 --once
> quit
# shutdown -r now

That should be it.

[eva2000]

wow thanks for the heads up.. updating first post!

[NickFritz]

OMg, thanks the kernel upgrade works WONDERFULLY, and i had been looking for a tutorial for upgrading my kernel.

Remember for those who want to upgrade their kernel

SMP = multiple cpu

i accidently installed the single cpu one and it nearly killed my server from the load spike :P

[mlx]

Please note that I have updated my tutorial post above as kernel-2.6.9-39.0.2 has been released on the dev repository a few days ago.

This kernel-2.6.9-39 is working like a charm by the way. Our server has been very stable since we upgraded!

P.S. eva2000: one of the links in your first post isn't working correctly, http://www.vbulletin.com/forum/showpost.php?p=1151870&postcount=65 should be http://www.vbulletin.com/forum/showpost.php?p=1151875&postcount=65

[eva2000]

thanks fixed it :)

[Deimos]

I also have the same issues with my server, I ran through the yum process of upgrading the kernel, it now reports.

2.6.9-34.0.2.EL

Which is the upgraded version correct? (I completely forgot which version I had before, but I think it was 0.1.EL?)

[Deimos]

Oops, completely missed the mini guide above, thanks to the poster, I upgraded AOK!

[Estad]

Linux vvv.vvv.com 2.6.11.7.dn2.64 #3 SMP Mon Apr 11 11:47:02 EDT 2005 i686 i686 i386 GNU/Linux

with
CentOS 3.7 i686

and I have tired with every day hihj load and slowly in the server .
what I have to do ??:(

[eva2000]

request your web host to upgrade your linux kernel.. but not sure if centOS 3.7 is affected by this it's mainly for centOS 4.2/4.3.. worth a try though

[kerplunk]

I find it odd that you guys are having all of these problems with CentOS. I have ran all kinds of differen versions of CentOS and never experienced any of these problems. In fact, I actually had some issues with Fedora so I had to move to CentOS on some servers. For enterprise linux, CentOS is the only OS I would recommend.

[eva2000]

I find it odd that you guys are having all of these problems with CentOS. I have ran all kinds of differen versions of CentOS and never experienced any of these problems. In fact, I actually had some issues with Fedora so I had to move to CentOS on some servers. For enterprise linux, CentOS is the only OS I would recommend.
Yeah it's strange.. but as you can see from this thread, it's a real issue for some folks... i haven't had any probs with CentOS myself, but that's because my web host custom compiles thier own linux kernerls 2.6.13/14 and 2.6.17-6 or higher kernels bypassing any of these above mentioned issues.

[NIS-Francisco]

Just upgraded to kernel-smp-2.6.9-39.EL.i686.rpm

Let's see how it goes.

[FreDyz]

I've upgraded kernel following mlx instructions and now everything seems to be OK, but if I execute "cat /proc/cpuinfo" the system tell me that I have only a processor, and I have two... How I can solve this?

Thanks

[Tomek]

Install a SMP-Kernel.

[FreDyz]

Yes... is that... thanks!

[pspcrazy]

What is your using a dual core operton server with only one dual core processor then what do you use?

Oh yea is there any older rpms that are safe to use as well? That would b awesxome to know because its better to downgrade then to upgrade.

[eva2000]

smp means more than one cpu core. so for dual core single cpu = smp kernel still

[NIS-Francisco]

DO NOT USE THIS KERNEL!

IT IS VULNERABLE TO r00t exploits.

One of our servers got hacked, customers unhappy. Don't make the mistake I did.

http://forums.navigatoris.net/support-and-questions/282-my-sites-were-hacked-cant-log-cpanel.html

[eva2000]

DO NOT USE THIS KERNEL!

IT IS VULNERABLE TO r00t exploits.

One of our servers got hacked, customers unhappy. Don't make the mistake I did.

http://forums.navigatoris.net/support-and-questions/282-my-sites-were-hacked-cant-log-cpanel.html
you mean 2.6.39 right ? please keep us informed how it goes

[jsharper]

DO NOT USE THIS KERNEL!

IT IS VULNERABLE TO r00t exploits.

One of our servers got hacked, customers unhappy. Don't make the mistake I did.

http://forums.navigatoris.net/support-and-questions/282-my-sites-were-hacked-cant-log-cpanel.html

Can you elaborate? I see nothing at all in that referenced thread that indicates that there was any kernel vulnerability or even anything to suggest that the defacer got root. Perhaps I missed it...?

[iRain]

I have two server .

both of the server is : CentOS 4.X

first one :
Linux myHostName 2.6.17.8 #1 SMP Mon Aug 14 21:22:53 AST 2006 i686 i686 i386 GNU/Linux
working like hill , load keep more than 25 for all time ..


second one :
Linux myHostName 2.6.9-22.0.1.ELsmp #1 SMP Thu Oct 27 13:14:25 CDT 2005
working fine with any problem .. !



any help ..!

[iRain]

Multipost

[eva2000]

Seems 2.6.9-40 and 2.6.9-42 kernels are out

http://forums.cpanel.net/showpost.php?p=265314&postcount=13

and full thread http://forums.cpanel.net/showthread.php?t=54708

[clubplanet]

I had these ridiculous high CPU utilizations too and just upgraded from 2.6.9-34.0.2.ELsmp to 2.6.9-42.ELsmp. Will let you know how it goes.

[pspcrazy]

what is Devel for? Im still wondering.

[Tomek]

The devel-Package includes the kernel headers, which are needed for compiling foreign kernel modules i.e.

[pulsorock]

CentOS just released version 4.4

I'm downloading the updates via yum.

I think it has kernel 2.6.9-42.0.2.EL

[fury]

I have 2.6.16-1 and have been having some stability problems. Should I look for an upgrade to this, or is this a branch unaffected by the problems in the 2.6.9 line?

[bajabob]

For me, upgrading to .39 did help to reduce the load on my servers a bit, but .42 did a much better job.

What about centos 4.4? does it worth the upgrade? would that help reduce the load on my servers?

[mlx]

I have upgraded to CentOS 4.4 a few days ago. Besides that the postfix update messed up my MailScanner setup and I had to repair that, everything worked like a charm. Staying up2date is always worth it I guess ;) Not sure though if that will lower your server load.

But the server is very stable with the 2.6.9-42.0.2 kernel. No need to manually use something else anymore. Just use the latest kernel from the standard CentOS repository.

[pulsorock]

I have upgraded to CentOS 4.4 a few days ago. Besides that the postfix update messed up my MailScanner setup and I had to repair that, everything worked like a charm.

I think I'm having the same problem... my mail is not longer working after the update... How did you managed to fix it?

[mlx]

I think I'm having the same problem... my mail is not longer working after the update... How did you managed to fix it?

My lazy solution was to just download the latest stable MailScanner package and re-install it by running that install.sh. Afterwards I just restartet MailScanner and everything was alright again.

[activa]

i have now centos 4.4 , but when i type in ssh comman uname -a it give me

Linux server.*****.com 2.6.9 #1 SMP Wed Jul 19 16:24:18 MSD 2006 i686 athlon i386 GNU/Linux

is still the karnel 2.6.9 ?
in whm panel it show centos 4.4 ....

is this normal ?
other things , the acp cache now it not working , i have to disable the line in php.ini to php work , i have re-compiled but no way , how i can make acp cache wok with this version ? note i have php 4.4.3 now installed

[mlx]

is still the karnel 2.6.9 ?
in whm panel it show centos 4.4 ....

is this normal ?
other things , the acp cache now it not working , i have to disable the line in php.ini to php work , i have re-compiled but no way , how i can make acp cache wok with this version ? note i have php 4.4.3 now installed

WHM = cPanel, is it? I don't have any experience with cPanel sorry. But as mentioned above CentOS 4.4 comes along with kernel 2.6.9-42.0.2.EL.

With APC in your php.ini - does "php -v" display any error messages?

I'm running PHP 5.1.6 with APC 3.0.12p2 without any problems.

[activa]

i have upgraded to php 4.4.4
when i type php -v no error display ....

but the service httpd get stoped when i use the apc extension ... even if i restart it , it get stoped automaticly

[alshref]

hi

i have 2.4.20-31.9
work all

but if come:
grub> savedefault --default=0 --once

message

Error 15: File not found

where problem ?

[neuromancer11]

Which kernel version is latest for Centos 4.2. Can you give me the kernel link ? (For Intel Xeon Processors)

[vinzphua]

I'm running 'Linux vps.bennyng.net 2.6.9-022stab078.14-enterprise #1 SMP Wed Jul 19 14:35:02 MSD 2006 i686'

Should I be worried about this? My host tech support tells me it's different, but I am getting problems with my resource usage creeping up.

[-=Moosa=-]

What About 2.6.9-34.ELsmp on RHEL 4?
should i upgrade my Kernel?

[beansbaxter]

I am running the 2.6.9-34 kernel but my problems seem to be memory leaks and what appears to be too many open sessions. I dont think I'm having a problem with CPU load.

[Delphy]

Just want to add that I upgraded from -22 to -42 and the load on my fileserver went from 170+ (no kidding!) to less than 4. :)

[egyptsons]

Use at your own risk (worked fine for me)

First of all get the correct RPMs for your system. I have used:

# wget http://dev.centos.org/centos/4/testing/i386/RPMS/kernel-devel-2.6.9-39.0.2.EL.i686.rpm
# wget http://dev.centos.org/centos/4/testing/i386/RPMS/kernel-2.6.9-39.0.2.EL.i686.rpm

Now install them ... but do not use 'rpm -U'! 'rpm -i' will keep your old kernel(s), installing the new one instead of upgrading. This will make sure you can go back in case of any errors and it won't damage yum of course.

# rpm -ivh kernel-devel-2.6.9-39.0.2.EL.i686.rpm kernel-2.6.9-39.0.2.EL.i686.rpm

# nano /boot/grub/grub.conf

Installing the kernel should put the new one on top, having the old kernel(s) below. Make sure that the default says "default=1". Grub starts at 0. So the second kernel in the list would be your default (1). You can change it back to 0 later if the reboot is working fine. For now let's configure grub to boot the new kernel only once making sure that if it fails and the datacenter has to do a reboot it will bring back your old default kernel.

# grub
> savedefault --default=0 --once
> quit
# shutdown -r now

That should be it.
kernel-2.6.9-39.0.2.EL is very good
when I update to kernel-2.6.9-42.0.3.EL
Load come again and memory is gone :(
anyone tell me the way to downgrade ?

[pulsorock]

kernel-2.6.9-39.0.2.EL is very good
when I update to kernel-2.6.9-42.0.3.EL
Load come again and memory is gone :(
anyone tell me the way to downgrade ?
I just noticed that also... I upgraded the kernel via yum a few days ago from 2.6.9-42.0.2.ELsmp to 2.6.9-42.0.3.ELsmp and now is using swap memory...

[tactic]

Has anyone experienced any similar issues with this kernel version?
2.4.21-4.0.1.ELsmp #1 SMP Thu Oct 23 01:27:36 EDT 2003 i686 i686 i386 GNU/Linux

Is upgrading kernels relatively easy?? I am running an Ensim control panel as well..

[Ig@r]

Can anyone please confirm that kernel-2.6.9-39.0.2.EL is the best choise for now?

[hescominsoon]

I just noticed that also... I upgraded the kernel via yum a few days ago from 2.6.9-42.0.2.ELsmp to 2.6.9-42.0.3.ELsmp and now is using swap memory...
interesting. I have three different servers all running this kernel without issues. One thing to keep in mind is vm.swappiness by default in centos is set to 60. change that to zero and reboot. That should nearly or totaly eliminate your swapping unless you are having some memory pressure or application leakage problems.

[Ig@r]

So, what is the best version of centos kernal available?

[hescominsoon]

2.6.9-42.0.3.ELsmp

if you have swap issues set vm.swappiness to zero.

here's how you do it:

Right way:
On /etc/sysctl.conf, add "vm.swappiness = 0"

Or any other value you want. The default is 60.

Don't forget to run "sysctl -p" to load this new value.

[egyptsons]

2.6.9-42.0.3.ELsmp

if you ahve swap issues set vm.swappiness to zero.

here's how you do it:

Right way:
On /etc/sysctl.conf, add "vm.swappiness = 0"

Or any other value you want. The default is 60.

Don't forget to run "sysctl -p" to load this new value.
Thanx I do this but I still have Load and memory is gone

top - 09:30:39 up 8 min, 1 user, load average: 3.48, 2.70, 1.37
Tasks: 198 total, 1 running, 197 sleeping, 0 stopped, 0 zombie
Cpu(s): 23.6% us, 2.7% sy, 0.0% ni, 24.0% id, 49.6% wa, 0.2% hi, 0.0% si
Mem: 1034604k total, 1019412k used, 15192k free, 9120k buffers
Swap: 2096472k total, 0k used, 2096472k free, 315620k cached

[jsharper]

Thanx I do this but I still have Load and memory is gone

top - 09:30:39 up 8 min, 1 user, load average: 3.48, 2.70, 1.37
Tasks: 198 total, 1 running, 197 sleeping, 0 stopped, 0 zombie
Cpu(s): 23.6% us, 2.7% sy, 0.0% ni, 24.0% id, 49.6% wa, 0.2% hi, 0.0% si
Mem: 1034604k total, 1019412k used, 15192k free, 9120k buffers
Swap: 2096472k total, 0k used, 2096472k free, 315620k cached

The mem isn't gone. You've got 1 gig total with about 331meg still available. Most of that 331meg is currently being taken advantage of as disk cache (and should become usable ram if a process needs it). But you're a ways off from being out of ram and you aren't touching swap at all. Sounds pretty decent to me.

What is making you think the 3.48/2.70/1.37 load numbers are not representative of the work that the box is doing?


</grain of salt>

[Ig@r]

So, what is the best version of centos kernal available?2.6.9-42.0.3.ELsmp

Sorry for nube question. Where to obtain sources for this kernel?
I need to install it manually, when I tried to use "yim upadate" server ended up with "kernel panic"

[hescominsoon]

Sorry for nube question. Where to obtain sources for this kernel?
I need to install it manually, when I tried to use "yim upadate" server ended up with "kernel panic"

if you can't yum it then installing from sources may work. i would figure out the cause of the panic. If you shoehorn the kernel in you will likely have an unworkable system.

[Ig@r]

if you can't yum it then installing from sources may work.
OK, that is what I'm trying to do.
Can you point me to the right direction where can I download the latest source?

i would figure out the cause of the panic. It stopped with a
kernel panic after trying to mount the file system

[hescominsoon]

OK, that is what I'm trying to do.
Can you point me to the right direction where can I download the latest source?
It stopped with a
kernel panic after trying to mount the file system


I would take this issue to the centos mailing list. They'll help you quite quickly and thoroughly. YOu can find the mailing list signup in the support section at centos.org

[Kristijan[cro]]

Linux polux 2.6.16.13-4-default #1 Wed May 3 04:53:23 UTC 2006 i686 i686 i386 GNU/Linux i have this version and i have very big problem

i just have transfered to new server, excellent server p4

and server load is 95% ???? 60 users online????

in the last server are 100 users and server is being slow, this is new and works fine, but i have problem with quick answer and server load

top - 23:31:56 up 2:22, 4 users, load average: 13.41, 11.12, 7.97
Tasks: 117 total, 10 running, 107 sleeping, 0 stopped, 0 zombie
Cpu(s): 95.0% us, 4.7% sy, 0.0% ni, 0.0% id, 0.0% wa, 0.0% hi, 0.3% si
Mem: 775732k total, 730084k used, 45648k free, 91424k buffers
Swap: 1028120k total, 136k used, 1027984k free, 139728k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
17073 wwwrun 15 0 66460 19m 3556 S 20.5 2.5 0:05.73 httpd2-prefork
18085 wwwrun 15 0 66912 19m 3544 S 20.2 2.6 0:00.73 httpd2-prefork
16705 wwwrun 15 0 65836 18m 3552 S 7.1 2.4 0:07.49 httpd2-prefork
17956 wwwrun 15 0 65656 18m 3572 S 7.1 2.4 0:00.82 httpd2-prefork
16694 wwwrun 16 0 66820 19m 3572 R 5.5 2.6 0:06.84 httpd2-prefork
16943 wwwrun 15 0 66264 18m 3564 S 5.5 2.5 0:08.20 httpd2-prefork
4717 mysql 15 0 118m 27m 4500 S 5.1 3.6 1:54.37 mysqld
16400 wwwrun 15 0 66004 18m 3572 S 4.8 2.5 0:05.63 httpd2-prefork
14403 wwwrun 16 0 65356 18m 3612 R 4.2 2.4 0:03.24 httpd2-prefork
14996 wwwrun 16 0 66148 18m 3616 R 3.9 2.5 0:23.72 httpd2-prefork
17367 wwwrun 16 0 66444 19m 3584 S 3.2 2.5 0:01.85 httpd2-prefork
15341 wwwrun 16 0 66512 19m 3596 R 1.3 2.5 0:08.11 httpd2-prefork
17083 wwwrun 15 0 65668 18m 3564 S 0.6 2.4 0:03.74 httpd2-prefork
2850 root 16 0 5416 1744 1412 S 0.3 0.2 0:00.08 master
9389 postfix 15 0 5376 1708 1384 D 0.3 0.2 0:00.01 pickup
17954 wwwrun 16 0 60272 12m 3368 S 0.3 1.7 0:00.33 httpd2-prefork
17983 kdrvar 16 0 2188 1036 756 R 0.3 0.1 0:00.06 top
1 root 16 0 720 284 244 S 0.0 0.0 0:01.03 init
2 root 34 19 0 0 0 R 0.0 0.0 0:00.00 ksoftirqd/0
3 root 10 -5 0 0 0 S 0.0 0.0 0:00.06 events/0
4 root 10 -5 0 0 0 S 0.0 0.0 0:00.00 khelper
5 root 10 -5 0 0 0 S 0.0 0.0 0:00.00 kthread
7 root 10 -5 0 0 0 S 0.0 0.0 0:00.10 kblockd/0
8 root 20 -5 0 0 0 S 0.0 0.0 0:00.00 kacpid
119 root 16 -5 0 0 0 S 0.0 0.0 0:00.00 aio/0
118 root 15 0 0 0 0 S 0.0 0.0 0:03.85 kswapd0
325 root 11 -5 0 0 0 S 0.0 0.0 0:00.00 cqueue/0
326 root 10 -5 0 0 0 S 0.0 0.0 0:00.00 kseriod
366 root 11 -5 0 0 0 S 0.0 0.0 0:00.00 kpsmoused
766 root 10 -5 0 0 0 S 0.0 0.0 0:00.74 reiserfs/0
823 root 13 -4 1832 640 344 S 0.0 0.1 0:00.57 udevd
1371 root 10 -5 0 0 0 S 0.0 0.0 0:00.00 khubd
1479 root 20 0 0 0 0 S 0.0 0.0 0:00.00 shpchpd_event
1984 messageb 17 0 3420 960 728 S 0.0 0.1 0:00.33 dbus-daemon
1998 root 16 0 1520 504 428 S 0.0 0.1 0:00.00 acpid
1999 root 15 0 1800 656 528 S 0.0 0.1 0:00.00 resmgrd
2020 root 16 0 4304 2924 1468 S 0.0 0.4 0:00.94 hald
2093 root 18 0 1820 608 524 S 0.0 0.1 0:00.00 hald-addon-acpi
2307 root 16 0 1816 740 648 S 0.0 0.1 0:00.11 hald-addon-stor
2330 root 16 0 1656 520 324 S 0.0 0.1 0:00.00 klogd
2335 root 15 0 1896 780 576 S 0.0 0.1 0:00.07 syslog-ng
2344 nobody 19 0 1556 404 324 S 0.0 0.1 0:00.00 portmap
2414 root 15 0 103m 1052 776 S 0.0 0.1 0:00.12 nscd
2602 root 16 0 4260 1724 1384 S 0.0 0.2 0:00.02 powersaved
2747 root 16 0 4956 1216 876 S 0.0 0.2 0:00.13 sshd
2871 root 16 0 1820 540 440 S 0.0 0.1 0:00.00 cron

Nov 07 23:56:33 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:33 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:34 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:34 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:34 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:34 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:34 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:34 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:34 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:35 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:35 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:35 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:35 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:35 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:35 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:35 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:36 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:36 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:36 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:36 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:36 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:36 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:36 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:36 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:36 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:37 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:37 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:37 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:37 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:37 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:37 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:37 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:37 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:37 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:38 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:38 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:38 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:38 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:38 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:38 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:39 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'
[Tue Nov 07 23:56:39 2006] [error] Failed to change_hat to 'HANDLING_UNTRUSTED_INPUT'

[hescominsoon]

apache is killing you there. You aren't swapping anything. I have been digging around trying to find the cause and a fix..:) Right now it's something with either apache or an apache module or a script(like php) that's killing apache. If this is centos have you tried centos.org?

[Kristijan[cro]]

hej hescominsoon, what is centos, can u explain a litle bit, thanks ;)

[hescominsoon]

;1247364']hej hescominsoon, what is centos, can u explain a litle bit, thanks ;)

http://www.centos.org


Red Hat releases their Enterprise Linux in source code format. Centos take those source rpm's..removed REd Hat trademarks and calls it CentOS. Wile RHEL costs money for support..CentOS is a self-supported version. RHEL has 5 year life while the fedora series typically have an 18 month lifecycle.

[creativepart]

My server is CentOS 4.2 and is running kernel 2.6.9-34.107 and I've been get high load readings and the server stalling out and pausing. We've been looking for a good reason and this post seems to apply to my server.

I've read the whole sticky but I'm not positive about what kernel is best to install.

My webhost is all set to upgrade the kernel to 2.6.9-42. Is there consensus that this is the correct kernel to solve this problem?

Should I go ahead with this upgrade?

[hescominsoon]

yes you should. If your host is willing set vm.swappiness=0

[egyptsons]

The mem isn't gone. You've got 1 gig total with about 331meg still available. Most of that 331meg is currently being taken advantage of as disk cache (and should become usable ram if a process needs it). But you're a ways off from being out of ram and you aren't touching swap at all. Sounds pretty decent to me.

What is making you think the 3.48/2.70/1.37 load numbers are not representative of the work that the box is doing?


</grain of salt>
thats status is after 8 minuet of restarting the server :( I don't know tell now how can I fix this upgrade

[jsharper]

I'm still confused about why you think that is abnormal even 30 seconds after starting the server, and why you think something needs to be fixed. Are you experiencing a problem with slow performance?

[activa]

i have a vps ....

Linux server.********.*** 2.6.9 #2 SMP Fri Sep 29 02:47:35 EDT 2006 i686 athlon i386 GNU/Linux

i want to make a upgrade kernel when i type the

yum upgrade kernel

the output is



Setting up Upgrade Process
Setting up repositories
Reading repository metadata in from local files
Excluding Packages in global exclude list
Finished
Could not find update match for kernel-smp
No Packages marked for Update/Obsoletion



when i type grup no command found ....

how i can upgrade to the latest kernel .....
thanks in advance

[Marco van Herwaarden]

Pleaqse contact your host for instructions specific for their setup.

[egyptsons]

I'm still confused about why you think that is abnormal even 30 seconds after starting the server, and why you think something needs to be fixed. Are you experiencing a problem with slow performance?
coz this load appear again after upgrade the Kernel

[lazyindian]

Well this is my problem my present host gave me 2.6.9-23stab version kernel, present server is clovertown 1.66. We have some 100,000+ posts, 768 mb ram with vbshout, arcade, vbplaza,vbimagehosting installed.

60-70 concurrent users online at anytime and we are having a hell lot of problem with the loads what could be the reason ? :(


And i am planning of a new vps from liquidweb which offfers
512 MB ram , Dual Opteron Processor
The kernel installed would be 2.4.20-021stab028.19.777-smp

Will i be having any issues with vbulletin 3.6.3 on the 2.4.20 kernel? ?

[Ig@r]

60-70 concurrent users online at anytime and we are having a hell lot of problem with the loads what could be the reason ? :(Just upgrade kernel to the newest version.

[lazyindian]

they said they wont be able to upgrade the kernel ...as i said its a VPS :(

[TPTP]

With 2.6.9-42.0.3 installed we were crashing every 3-4 days like clockwork. It's off-season for us, and our dual irwindale / scsi box was crapping out with no warning with only 150-250 online. Typically 2500 online is an expected norm in our "on-season". Load averages were climbing out of control, memory usage was soaring, and swap usage buried it until it bought the farm.

Dropping back to 2.6.9-42.0.2 solved that issue soundly. We still have some other issues, but the difference was like coming home. We're back in business. We went from lashing the rudder to the mast, putting on raingear, and huddling below deck, to blue skies and only a hand lazily resting on the rudder. What an amazing difference.

[dsotmoon]

My webhost is all set to upgrade the kernel to 2.6.9-42. Is there consensus that this is the correct kernel to solve this problem?

Should I go ahead with this upgrade?

yes you should. If your host is willing set vm.swappiness=0

what does changing "vm.swappiness=0" do exactly? is this what eliminates the centos 4/kernel problem?

I am getting ready to move a very busy forum to a centos4.4 server running 2.6.9-42.0.3.ELsmp #1 , i want to make sure I wont have any problems or try to correct them before I make the transfer

thanks

[Mike Warner]

I am getting the problems described here, but I am not running CentOS, but RedHat EL.

Linux mydomain.com 2.6.9-22.ELsmp #1 SMP Mon Sep 19 18:32:14 EDT 2005 i686

Even though I'm not running Centos, could this still be the cause of my problems? Afterall, CentOS is based off Redhat EL.

Many thanks.

[Mike Warner]

I am getting the problems described here, but I am not running CentOS, but RedHat EL.

Linux mydomain.com 2.6.9-22.ELsmp #1 SMP Mon Sep 19 18:32:14 EDT 2005 i686

Even though I'm not running Centos, could this still be the cause of my problems? Afterall, CentOS is based off Redhat EL.

Many thanks.Anyone?

[eva2000]

could be same problem.. only way to know is updating your kernel as well

[Luca_000]

[root@server mt]# uname -a
Linux server.hostingtalk.it 2.6.9-42.0.3.ELsmp #1 SMP Fri Oct 6 06:28:26 CDT 2006 x86_64 x86_64 x86_64 GNU/Linux

[mlx]

CentOS Errata and Security Advisory 2007:0014

https://rhn.redhat.com/errata/RHSA-2007-0014.html

The following updated files have been uploaded and are currently
syncing to the mirrors:

i386:
kernel-2.6.9-42.0.8.EL.i586.rpm
kernel-2.6.9-42.0.8.EL.i686.rpm
kernel-devel-2.6.9-42.0.8.EL.i586.rpm
kernel-devel-2.6.9-42.0.8.EL.i686.rpm
kernel-doc-2.6.9-42.0.8.EL.noarch.rpm
kernel-hugemem-2.6.9-42.0.8.EL.i686.rpm
kernel-hugemem-devel-2.6.9-42.0.8.EL.i686.rpm
kernel-smp-2.6.9-42.0.8.EL.i586.rpm
kernel-smp-2.6.9-42.0.8.EL.i686.rpm
kernel-smp-devel-2.6.9-42.0.8.EL.i586.rpm
kernel-smp-devel-2.6.9-42.0.8.EL.i686.rpm

src:
kernel-2.6.9-42.0.8.EL.src.rpm
We've just upgraded to 2.6.9-42.0.8 ... running the new kernel for only about 30 minutes yet, but looks good so far :)

[eva2000]

We've just upgraded to 2.6.9-42.0.8 ... running the new kernel for only about 30 minutes yet, but looks good so far :)
Let us know how it goes in a few days :)

[mlx]

Let us know how it goes in a few days :)

The server has been up for about 6 days now. Everything's alright :)

[eva2000]

Good to hear so 2.6.9-42.0.8 get's the tick of approval i guess :)

[als3ada]

Hi,
I hve this kernel 2.6.9-42.0.3.ELsmp and i was installed kernel-2.6.9-44.EL
now when i try to install 2.6.9-42.0.8.EL kernel it say to me
root@linux [~]# rpm -ivh kernel-*
Preparing... ########################################### [100%]
package kernel-2.6.9-44.EL (which is newer than kernel-2.6.9-42.0.8.EL) is already installed

how can i install 2.6.9-42.0.8.EL ?

[Bob Isaac]

I've never messed with kernel upgrades, I leave that to the network guys. There is a --force (http://www.redhat.com/docs/books/max-rpm/max-rpm-html/s1-rpm-install-additional-options.html#S2-RPM-INSTALL-FORCE-OPTION) option that may work, but check it out before you use it.

Bob

[hescominsoon]

Hi,
I hve this kernel 2.6.9-42.0.3.ELsmp and i was installed kernel-2.6.9-44.EL
now when i try to install 2.6.9-42.0.8.EL kernel it say to me


how can i install 2.6.9-42.0.8.EL ?
the best way to upgrade your kernel in centos is yum update. it will grab the latest kernel and install it..a reboot and you are good to go. If you already have the latest kernel then you are good to go.

[als3ada]

the best way to upgrade your kernel in centos is yum update. it will grab the latest kernel and install it..a reboot and you are good to go. If you already have the latest kernel then you are good to go.
but i think the best is kernel 2.6.9-42.0.8.EL

i upgrade to kernel-2.6.9-44.EL manually not by yum, and now when i update by yum, i think it is faild coz the kernel-2.6.9-44.EL installed

how can i uninstall kernel-2.6.9-44.EL to can downgrade to 2.6.9-42.0.8.EL ?

[hescominsoon]

you dn't want to downgrade as the latest kernel has the latest security patches installed. If you downgrade you are opening up security issues.

[hescominsoon]

Hi,
I hve this kernel 2.6.9-42.0.3.ELsmp and i was installed kernel-2.6.9-44.EL
now when i try to install 2.6.9-42.0.8.EL kernel it say to me


how can i install 2.6.9-42.0.8.EL ?

you have to reboot to use the newly installed kernel. Then you can run yum update again and it will install the latest kernel available. You have to reboot for kernel installs to take effect.

[als3ada]

you have to reboot to use the newly installed kernel. Then you can run yum update again and it will install the latest kernel available. You have to reboot for kernel installs to take effect.
Hi,

when i make yum search kernel , i have this

Install the kernel-smp package if your machine uses two or more CPUs.
System Environment/Kernel

kernel-smp.i686 2.6.9-42.0.8.EL updates
Matched from:
kernel-smp
The Linux kernel compiled for SMP machines.
This package includes a SMP version of the Linux kernel. It is
required only on machines with two or more CPUs as well as machines with
hyperthreading technology.
Install the kernel-smp package if your machine uses two or more CPUs.
System Environment/Kernel

kernel-smp-devel.i686 2.6.9-42.0.8.EL updates
Matched from:
kernel-smp-devel
Development package for building kernel modules to match the SMP kernel.
This package provides kernel headers and makefiles sufficient to build modules
against the SMP kernel package.
System Environment/Kernel

kernel-doc.noarch 2.6.9-42.0.8.EL updates
Matched from:
kernel-doc
Various documentation bits found in the kernel source.
This package contains documentation files from the kernel
source. Various bits of information about the Linux kernel and the
device drivers shipped with it are documented in these files.
You'll want to install this package if you need a reference to the
options that can be passed to Linux kernel modules at load time.

keyutils.i386 1.0-2 installed
Matched from:
Utilities to control the kernel key management facility and to provide
a mechanism by which the kernel call back to userspace to get a key
instantiated.

selinux-policy-targeted.noarch 1.17.30-2.140 installed
Matched from:
Security-enhanced Linux is a patch of the Linuxآ� kernel and a number
of utilities with enhanced security functionality designed to add
mandatory access controls to Linux. The Security-enhanced Linux
kernel contains new architectural components originally developed to
improve the security of the Flask operating system. These
architectural components provide general support for the enforcement
of many kinds of mandatory access control policies, including those
based on the concepts of Type Enforcementآ�, Role-based Access
Control, and Multi-level Security.
This package contains the SELinux example policy configuration along
with the Flask configuration information and the application
configuration files.

hotplug.i386 3:2004_04_01-7.7 installed
Matched from:
The term "hotplugging" refers to the dynamic reconfiguration performed
after a device has been attached to a running system. This package
contains the application which is called by the kernel when a USB
device is added; hotplug then loads the required modules for that
device.
System Environment/Kernel

kernel.i686 2.6.9-42.EL installed
Matched from:
kernel
the linux kernel (the core of the linux operating system)
The kernel package contains the Linux kernel (vmlinuz), the core of any
Linux operating system. The kernel handles the basic functions
of the operating system: memory allocation, process allocation, device
input and output, etc.
System Environment/Kernel

crash.i386 4.0-2.30 installed
Matched from:
The core analysis suite is a self-contained tool that can be used to
investigate either live systems, kernel core dumps created from the
netdump and diskdump packages from Red Hat Linux, the mcore kernel patch
offered by Mission Critical Linux, or the LKCD kernel patch.

iptables.i386 1.2.11-3.1.RHEL4 installed
Matched from:
Tools for managing Linux kernel packet filtering capabilities.
The iptables utility controls the network packet filtering code in the
Linux kernel. If you need to set up firewalls and/or IP masquerading,
you should install this package.

nfs-utils.i386 1.0.6-70.EL4 installed
Matched from:
NFS utlilities and supporting daemons for the kernel NFS server.
The nfs-utils package provides a daemon for the kernel NFS server and
related tools, which provides a much higher level of performance than the
traditional Linux NFS server used by most users.
This package also contains the showmount program. Showmount queries the
mount daemon on a remote host for information about the NFS (Network File
System) server on the remote host. For example, showmount can display the
clients which are mounted on that host.

pam-devel.i386 0.77-66.17 installed
Matched from:
http://www.us.kernel.org/pub/linux/libs/pam/index.html

iproute.i386 2.6.9-3.EL4.3.centos4 installed
Matched from:
The iproute package contains networking utilities (ip and rtmon, for
example) which are designed to use the advanced networking
capabilities of the Linux 2.4.x and 2.6.x kernel.

kernel-devel.i686 2.6.9-42.0.3.EL installed
Matched from:
kernel-devel
Development package for building kernel modules to match the kernel.
This package provides kernel headers and makefiles sufficient to build modules
against the kernel package.
System Environment/Kernel

kernel-hugemem-devel.i686 2.6.9-42.0.8.EL installed
Matched from:
kernel-hugemem-devel
Development package for building kernel modules to match the hugemem kernel.
This package provides kernel headers and makefiles sufficient to build modules
against the hugemem kernel package.
System Environment/Kernel

ethtool.i386 1.8-4 installed
Matched from:
http://sourceforge.net/projects/gkernel/

libselinux.i386 1.19.1-7.2 installed
Matched from:
Security-enhanced Linux is a patch of the Linuxآ� kernel and a number
of utilities with enhanced security functionality designed to add
mandatory access controls to Linux. The Security-enhanced Linux
kernel contains new architectural components originally developed to
improve the security of the Flask operating system. These
architectural components provide general support for the enforcement
of many kinds of mandatory access control policies, including those
based on the concepts of Type Enforcementآ�, Role-based Access
Control, and Multi-level Security.
libselinux provides an API for SELinux applications to get and set
process and file security contexts and to obtain security policy
decisions. Required for any applications that use the SELinux API.

checkpolicy.i386 1.17.5-1 installed
Matched from:
Security-enhanced Linux is a patch of the Linuxآ� kernel and a number
of utilities with enhanced security functionality designed to add
mandatory access controls to Linux. The Security-enhanced Linux
kernel contains new architectural components originally developed to
improve the security of the Flask operating system. These
architectural components provide general support for the enforcement
of many kinds of mandatory access control policies, including those
based on the concepts of Type Enforcementآ�, Role-based Access
Control, and Multi-level Security.
This package contains checkpolicy, the SELinux policy compiler.
Only required for building policies.

audit.i386 1.0.14-1.EL4 installed
Matched from:
User space tools for 2.6 kernel auditing.
The audit package contains the user space utilities for
storing and searching the audit records generate by
the audit subsystem in the Linux 2.6 kernel.

MAKEDEV.i386 3.15.2-3 installed
Matched from:
This package contains the MAKEDEV program, which makes it easier to create
and maintain the files in the /dev directory. /dev directory files
correspond to a particular device supported by Linux (serial or printer
ports, scanners, sound cards, tape drives, CD-ROM drives, hard drives,
etc.) and interface with the drivers in the kernel.
You should install the MAKEDEV package because the MAKEDEV utility makes
it easy to manage the /dev directory device files.

sysklogd.i386 1.4.1-26_EL installed
Matched from:
System logging and kernel message trapping daemons.

pam.i386 0.77-66.17 installed
Matched from:
http://www.us.kernel.org/pub/linux/libs/pam/index.html

policycoreutils.i386 1.18.1-4.9 installed
Matched from:
Security-enhanced Linux is a patch of the Linuxآ� kernel and a number
of utilities with enhanced security functionality designed to add
mandatory access controls to Linux. The Security-enhanced Linux
kernel contains new architectural components originally developed to
improve the security of the Flask operating system. These
architectural components provide general support for the enforcement
of many kinds of mandatory access control policies, including those
based on the concepts of Type Enforcementآ�, Role-based Access
Control, and Multi-level Security.
policycoreutils contains the policy core utilities that are required
for basic operation of a SELinux system. These utilities include
load_policy to load policies, setfiles to label filesystems, newrole
to switch roles, and run_init to run /etc/init.d scripts in the proper
context.

setools.i386 1.5.1-5 installed
Matched from:
Security-enhanced Linux is a patch of the Linux kernel and a number of
utilities with enhanced security functionality designed to add mandatory access
controls to Linux. The Security-enhanced Linux kernel contains new
architectural components originally developed to improve the security of the Flask
operating system. These architectural components provide general support for the
enforcement of many kinds of mandatory access control policies, including those
based on the concepts of Type Enforcement, Role-based Access Control, and
Multi-level Security.
The tools and libraries in this release include:
1. seuser: A GUI and command line user manager tool for SELinux. This
is a tool that actually manages a portion of a running policy (i.e.,
user accounts).
2. seuser scripts: A set of shell scripts: seuseradd, seusermod, and
seuserdel. These scripts combine the functions of the associated s*
commands with seuser to provide a single interface to manage users in
SE Linux.
3. libapol: The main policy.conf analysis library, which is the core
library for all our tools.
See the help files for apol, sepcut, and seuser for help on using the
tools.

udev.i386 039-10.15.EL4 installed
Matched from:
udev is a implementation of devfs in userspace using sysfs and
/sbin/hotplug. It requires a 2.6 kernel to run properly.
http://kernel.org/pub/linux/utils/kernel/hotplug/

mkinitrd.i386 4.2.1.8-1 installed
Matched from:
Mkinitrd creates filesystem images for use as initial ramdisk (initrd)
images. These ramdisk images are often used to preload the block
device modules (SCSI or RAID) needed to access the root filesystem.
In other words, generic kernels can be built without drivers for any
SCSI adapters which load the SCSI driver as a module. Since the
kernel needs to read those modules, but in this case it isn't able to
address the SCSI adapter, an initial ramdisk is used. The initial
ramdisk is loaded by the operating system loader (normally LILO) and
is available to the kernel as soon as the ramdisk is loaded. The
ramdisk image loads the proper SCSI adapter and allows the kernel to
mount the root filesystem. The mkinitrd program creates such a
ramdisk using information found in the /etc/modules.conf file.

kernel-smp.i686 2.6.9-42.EL installed
Matched from:
kernel-smp
The Linux kernel compiled for SMP machines.
This package includes a SMP version of the Linux kernel. It is
required only on machines with two or more CPUs as well as machines with
hyperthreading technology.
Install the kernel-smp package if your machine uses two or more CPUs.
System Environment/Kernel

kernel-ib.i386 1.0-1 installed
Matched from:
kernel-ib
Infiniband updates for udev and boot time kernel module loading
Infiniband updates for udev and boot time kernel module loading

setarch.i386 1.6-1 installed
Matched from:
This utility tells the kernel to report a different architecture than the
current one, then runs a program in that environment. It can also set various
personality flags.
System Environment/Kernel

dapl.i386 1.2.0-1 installed
Matched from:
Along with the OpenIB kernel drivers, libdat and libdapl provides a userspace
RDMA API that supports DAT 1.2 specification

irda-utils.i386 0.9.16-3 installed
Matched from:
IrDA(TM) (Infrared Data Association) is an industry standard for
wireless, infrared communication between devices. IrDA speeds range
from 9600 bps to 4 Mbps, and IrDA can be used by many modern devices
including laptops, LAN adapters, PDAs, printers, and mobile phones.
The Linux-IrDA project is a GPL'd implementation, written from
scratch, of the IrDA protocols. Supported IrDA protocols include
IrLAP, IrLMP, IrIAP, IrTTP, IrLPT, IrLAN, IrCOMM and IrOBEX.
The irda-utils package contains a collection of programs that enable
the use of IrDA protocols. Most IrDA features are implemented in the
kernel, so IrDA support must be enabled in the kernel before any IrDA
tools or programs can be used. Some configuration outside the kernel
is required, however, and some IrDA features, like IrOBEX, are
actually implemented outside the kernel.

diskdumputils.i386 1.3.15-1 installed
Matched from:
The diskdump utilities set up the kernel to save memory image to
the specified partition.

nscd.i386 2.3.4-2.25 installed
Matched from:
Nscd caches name service lookups and can dramatically improve
performance with NIS+, and may help with DNS as well. Note that you
can't use nscd with 2.0 kernels because of bugs in the kernel-side
thread support. Unfortunately, nscd happens to hit these bugs
particularly hard.

syslinux.i386 2.11-1 installed
Matched from:
Simple kernel loader which boots from a FAT filesystem
Syslinux is a simple kernel loader. It normally loads the kernel (and an
optional initrd image) from a FAT filesystem. It can also be used as a
PXE bootloader during network boots.

ipsec-tools.i386 0.3.3-6.rhel4.1 installed
Matched from:
This is the IPsec-Tools package. You need this package in order to
really use the IPsec functionality in the linux-2.5+ kernels. This
package builds:
- setkey, a program to directly manipulate policies and SAs
- racoon, an IKEv1 keying daemon

kernel-utils.i386 1:2.4-13.1.83 installed
Matched from:
kernel-utils
Kernel and Hardware related utilities
kernel-utils contains several utilities that can be used to control
the kernel or your machines hardware. Included are
* cpuspeed - dynamically change the speed of CPUs (if CPU is capable)
* dmidecode - gives information about the bios and motherboard revisions
* irqbalance - Evenly distribute interrupt load across CPUs.
* microcode_ctl - updates the microcode on Intel cpus
* rng-tools - Hardware random number generation tools.
* smartctl - monitor the health of your disks

netdump.i386 0.7.16-2 installed
Matched from:
Client setup for network kernel message logging and crash dumps
The netdump client sets up the kernel to send crash dumps and/or
console messages as syslog packets to a remote system. Some manual
setup is required as documented in the netdump (8) man page.

pciutils.i386 2.1.99.test8-3.2 installed
Matched from:
The pciutils package contains various utilities for inspecting and
setting devices connected to the PCI bus. The utilities provided
require kernel version 2.1.82 or newer (which support the
/proc/bus/pci interface).

rp-pppoe.i386 3.5-22 installed
Matched from:
PPPoE (Point-to-Point Protocol over Ethernet) is a protocol used by
many ADSL Internet Service Providers. This package contains the
Roaring Penguin PPPoE client, a user-mode program that does not
require any kernel modifications. It is fully compliant with RFC 2516,
the official PPPoE specification.

kernel-smp.i686 2.6.9-42.0.3.EL installed
Matched from:
kernel-smp
The Linux kernel compiled for SMP machines.
This package includes a SMP version of the Linux kernel. It is
required only on machines with two or more CPUs as well as machines with
hyperthreading technology.
Install the kernel-smp package if your machine uses two or more CPUs.
System Environment/Kernel

kernel.i686 2.6.9-42.0.3.EL installed
Matched from:
kernel
the linux kernel (the core of the linux operating system)
The kernel package contains the Linux kernel (vmlinuz), the core of any
Linux operating system. The kernel handles the basic functions
of the operating system: memory allocation, process allocation, device
input and output, etc.
System Environment/Kernel

dev86.i386 0.16.16-1 installed
Matched from:
The dev86 package provides an assembler and linker for real mode 80x86
instructions. You'll need to have this package installed in order to
build programs that run in real mode, including LILO and the kernel's
bootstrapping code, from their sources.
You should install dev86 if you intend to build programs that run in real
mode from their source code.

kernel.i686 2.6.9-44.EL installed
Matched from:
kernel
the linux kernel (the core of the linux operating system)
The kernel package contains the Linux kernel (vmlinuz), the core of any
Linux operating system. The kernel handles the basic functions
of the operating system: memory allocation, process allocation, device
input and output, etc.
System Environment/Kernel

kernel-smp-devel.i686 2.6.9-42.0.8.EL installed
Matched from:
kernel-smp-devel
Development package for building kernel modules to match the SMP kernel.
This package provides kernel headers and makefiles sufficient to build modules
against the SMP kernel package.
System Environment/Kernel

kernel-devel.i686 2.6.9-42.0.8.EL installed
Matched from:
kernel-devel
Development package for building kernel modules to match the kernel.
This package provides kernel headers and makefiles sufficient to build modules
against the kernel package.
System Environment/Kernel

libsepol.i386 1.1.1-2 installed
Matched from:
Security-enhanced Linux is a patch of the Linuxآ� kernel and a number
of utilities with enhanced security functionality designed to add
mandatory access controls to Linux. The Security-enhanced Linux
kernel contains new architectural components originally developed to
improve the security of the Flask operating system. These
architectural components provide general support for the enforcement
of many kinds of mandatory access control policies, including those
based on the concepts of Type Enforcementآ�, Role-based Access
Control, and Multi-level Security.
libsepol provides an API for the manipulation of SELinux binary policies.
It is used by checkpolicy (the policy compiler) and similar tools, as well
as by programs like load_policy that need to perform specific transformations
on binary policies such as customizing policy boolean settings.

module-init-tools.i386 3.1-0.pre5.3.2 installed
Matched from:
Kernel module management utilities.
The modutils package includes various programs needed for automatic
loading and unloading of modules under 2.2 and later kernels, as well
as other module management programs. Device drivers and filesystems
are two examples of loaded and unloaded modules.
System Environment/Kernel

SysVinit.i386 2.85-34.3 installed
Matched from:
The SysVinit package contains a group of processes that control
the very basic functions of your system. SysVinit includes the init
program, the first program started by the Linux kernel when the
system boots. Init then controls the startup, running, and shutdown
of all other programs.

ibmasm.i386 3.0-7 installed
Matched from:
System Environment/Kernel

lksctp-tools.i386 1.0.2-6.4E.1 installed
Matched from:
User-space access to Linux Kernel SCTP
This is the lksctp-tools package for Linux Kernel SCTP Reference
Implementation.
This package is intended to supplement the Linux Kernel SCTP Reference
Implementation now available in the Linux kernel source tree in
versions 2.5.36 and following. For more information on LKSCTP see the
package documentation README file, section titled "LKSCTP - Linux
Kernel SCTP."
This package contains the base run-time library & command-line tools.

kernel-devel.i686 2.6.9-42.EL installed
Matched from:
kernel-devel
Development package for building kernel modules to match the kernel.
This package provides kernel headers and makefiles sufficient to build modules
against the kernel package.
System Environment/Kernel

glibc-kernheaders.i386 2.4-9.1.98.EL installed
Matched from:
Header files for the Linux kernel for use by glibc
Kernel-headers includes the C header files that specify the interface
between the Linux kernel and userspace libraries and programs. The
header files define structures and constants that are needed for
building most standard programs and are also needed for rebuilding the
glibc package
root@linux [~]#


abd i make yum install kernel i have this

root@linux [~]# yum install kernel
Repository base is listed more than once in the configuration
Repository addons is listed more than once in the configuration
Repository extras is listed more than once in the configuration
Repository extras is listed more than once in the configuration
Repository updates is listed more than once in the configuration
Repository base is listed more than once in the configuration
Repository addons is listed more than once in the configuration
Setting up Install Process
Setting up repositories
update 100% |=========================| 951 B 00:00
extras 100% |=========================| 1.1 kB 00:00
updates 100% |=========================| 951 B 00:00
base 100% |=========================| 1.1 kB 00:00
addons 100% |=========================| 951 B 00:00
Reading repository metadata in from local files
Excluding Packages in global exclude list
Finished
Parsing package install arguments
Nothing to do


and after reboot the kernel ver. not change


(root@linux)root@linux [~]# uname -r
2.6.9-42.0.3.EL

(root@linux)

i want to install 2.6.9-42.0.8.EL

please help me.

[als3ada]

anyone here ?

[mlx]

you dn't want to downgrade as the latest kernel has the latest security patches installed. If you downgrade you are opening up security issues.

Installing 2.6.9-42.0.8.EL shouldn't be a security issue at all as it's the latest stable CentOS/RHEL4 kernel available!

anyone here ?

As Bob Isaac already said, use rpm --force -ivh to install the kernel anyway ...

[kerplunk]

Edit your /etc/grub.conf file to change which kernel you boot up with.

[Scott.Mc]

I am getting the problems described here, but I am not running CentOS, but RedHat EL.

Linux mydomain.com 2.6.9-22.ELsmp #1 SMP Mon Sep 19 18:32:14 EDT 2005 i686

Even though I'm not running Centos, could this still be the cause of my problems? Afterall, CentOS is based off Redhat EL.

Many thanks.

Just to note on this, you REALLY should upgrade your kernel as you are running a vulnerable kernel.

#####

As for the kernel upgrades do not use --force on the kernel. Stick to using your package manager if it's centos that will be yum or if it's RHEL it will be up2date.

For yum edit your /etc/yum.conf and remove kernel* from the skip packages.

For up2date run up2date --configure, go into the package skip list and clear that.

Then you can update the kernel and verify that it was installed via your boot loader(generally grub or lilo).

-Scott

[Scream And Fly]

You using centOS 64bit ... 64bit has issues and centos with that kernel has issues... i'd request your web host to change you to CentOS 32bit OS with more recent kernel than 2.6.9-22 smp i.e. 2.6.15


Eva, is this situation still true? My server is experiencing massive load problems that bring it down.

[hescominsoon]

unless you are running either more than 4 gigabytes of physical ram or processes larger than 4 gigabytes..64 bit is more trouble than it's worth..go with 32 bit.

[Scream And Fly]

Thank you so much for your help. Right now, we're ready to go back to a 32-bit OS since we cannot solve this sudden server load issue that's bringing things down. What 32-bit OS distro do you recommend?

My original post is here: http://www.vbulletin.com/forum/showthread.php?p=1313061#post1313061

[kerplunk]

Are these 64-bit problems just with high CPU loads?

Our main database server runs Fedora (64-bit) and our CPU loads are usually not more than 50% but our load averages can get up to 20.

[eva2000]

Eva, is this situation still true? My server is experiencing massive load problems that bring it down.
not 100% sure but it's possible - really it's the job of a good web host to properly have stressed test a server including OS/hardware/kernel test so as such problems don't rise.

[hescominsoon]

There's a new kernel out for Centos

yum update should get it for you.

[egyptsons]

I update to 2.6.9-42.0.10.EL :)
I will tell u if the load disappeared ;)

[dbembibre]

Running Transaction
Installing: kernel-smp ######################### [1/1]
Installed: kernel-smp.i686 0:2.6.9-42.0.10.EL
Complete!
[root@srv01 daniel]#


Testing now :D

[mlx]

We are running 2.6.9-42.0.10.EL for 6 days now. No problems here.

[dbembibre]

No load issues with the new 2.6.9-42.0.10.ELsmp

[DeeperImage]

I've been running into this problem with CentOS from time to time. I've updated my Kernel twice, and the load issues seem to go away for a few months, but reappear shortly after that time frame.

I just upgraded to latest one and hope it will fix me for good.

It's very annoying to have speed issues with a max number of concurrent members of 100 or less on a pretty descent machine.

[ant0nio]

No load problems with 2.6.9-42.0.10.ELsmp. Everything seems to be stable.

[dbembibre]

Running Transaction
Installing: kernel-smp ######################### [1/1]
Installed: kernel-smp.i686 0:2.6.9-55.EL
Complete!
[root@srv01 daniel]#


Testing now :D

[Daky]

Ive been using 55 for 3 days now, no problems at all.

[mlx]

We've upgraded several servers as well. No problems at all.

As my kernel upgrade how-to is still linked in the sticky first post I've also upgraded that post for the latest 2.6.9-55 kernel.

Having said that, it might be time to unstick this thread as CentOS 4.5 comes along with kernel-2.6.9-55.EL.i686 and at least I didn't notice any problems with it on our servers.

[eva2000]

We've upgraded several servers as well. No problems at all.

As my kernel upgrade how-to is still linked in the sticky first post I've also upgraded that post for the latest 2.6.9-55 kernel.

Having said that, it might be time to unstick this thread as CentOS 4.5 comes along with kernel-2.6.9-55.EL.i686 and at least I didn't notice any problems with it on our servers.
sounds very good.. i'll keep it stickied otherwise how will existing CentOS 4.4 users know about the issue :)

[oldford]

I'm running the old 2.6.9-42.0.3.ELsmp #1 kernel and am having load issues. So I asked my server management company to upgrade the kernel. They show the latest version to be 2.6.18.4, which is different from the 9 series that everyone is talking about here.

Should I try this or is there a reason to stay with the 9 series? Thanks!

[eva2000]

try which ever is supported by your web host... especially if they're doing the upgrade :)

[oldford]

Well I just upgraded the kernel to 2.6.18.4 #1 SMP and the machine rebooted cleanly, so that's a good start!:)

I'll give it a few days to see how things are going.

[eva2000]

K.. keep us updated :)

[oldford]

Update: Looks like 2.6.18.4 kernel isn't working for me. I can't reboot. My datacenter says the kernel panics and they had to reset me back to my old kernel. It's strange because the first reboot using the new kernel went fine. The server then ran most of the day fine, but I kept getting emails about iptable errors. So my management company fixed that, but then I couldn't boot.

So I went with the 2.6.9-55 kernel. Booted fine this morning and has been running without errors so far. Too early to tell about server load yet.

[kerplunk]

Unfortunately, I believe this thread is just a place for people to blame their high loads on CentOS. Maybe at one time a problem existed with a specific kernel that caused this, but as these problems are obviously in the past, I think this thread should be unsticky'd to prevent future confusion amongst people who might say, "Hey, I run CentOS... this is probably my problem!" (When in reality their problems are caused by the user.)

[eva2000]

well it's stuck so users can be aware of the issue IF they read this thread poperly ;)

[Tourmeister]

So does upgrading to 2.6.9-55.EL fix the load problems? I am using 2.6.9-42.EL and am getting load averages close to 200!! Before I start trying to upgrade PHP, MySQL, VB, etc,... it was suggested that I try updating to the -55 kernel.

[mlx]

So does upgrading to 2.6.9-55.EL fix the load problems? I am using 2.6.9-42.EL and am getting load averages close to 200!! Before I start trying to upgrade PHP, MySQL, VB, etc,... it was suggested that I try updating to the -55 kernel.


I don't think that we had any problems with 2.6.9-42.EL. But it's more than a year old, so upgrading to the latest 2.6.9-55.0.6.EL kernel would definitely be a good start. Not sure if that will fix all your problems though.

[Tourmeister]

The upgrade to 2.6.9-55 was a bust. It actually made things even worse. Prior to the upgrade, my forum would go down, but I could still get into the command prompt on the server using Putty and restart MySQL. Now when the spikes hit, I cannot access the server at all. It does not even respond to pings. Also, the frequency of the spikes has gone up considerably, from a few times a day to a few times an hour! I'm having it rolled back to the previous kernel version.

[mlx]

That's why I said upgrading the kernel is a good start but probably not your only problem. Now that's one of the reasons why this thread should indeed be unstuck :p

I'd suggest that you have a closer look at which processes are actually causing the load. Or post all required information (http://www.vbulletin.com/forum/showthread.php?t=70117) for an optimization request.

All I can say is that we are running 2.6.9-55.0.6.EL for over a week now and have zero problems with it.

[Tourmeister]

That's why I said upgrading the kernel is a good start but probably not your only problem. Now that's one of the reasons why this thread should indeed be unstuck :p

I'd suggest that you have a closer look at which processes are actually causing the load. Or post all required information (http://www.vbulletin.com/forum/showthread.php?t=70117) for an optimization request.

All I can say is that we are running 2.6.9-55.0.6.EL for over a week now and have zero problems with it.

True, but the kernel upgrade was a quick and easy thing to check. I've been monitoring processes using htop on my box and seldom see anything that looks like it is really hammering the resources. However, with that new kernel, the box would just lock up. Htop would freeze. While frozen, I would look to see what the last highest resource using processes were and none of them were really doing much. After the box would unfreeze, the server loads did not reflect that anything had even happened. Danged peculiar...

George made some recommendations for me in an optimization request and I am in the process of going through that. However, I have this nagging feeling there is a hardware issue lurking somewhere in here... Hence the other thread about what components to use to build a custom server ;)

[mlx]

Hmm, that sounds weird indeed and pretty much like the problems we were having with those old kernels. Haven't seen any problems like that here since 2.6.9-42.EL though.

If this is a rented dedicated server I would contact the datacenter where you rent the hardware and ask them for advice, maybe they can do a hardware-check or something.

[Tourmeister]

If this is a rented dedicated server I would contact the datacenter where you rent the hardware and ask them for advice, maybe they can do a hardware-check or something.

Hehe... and therein lies a big part of my problem... In terms of money, I am a flea on an elephants butt to them... :p They have not been real helpful unless I am willing to spend money to pay them to trouble shoot their hardware :rolleyes: Soo... I am looking at options to either move to a new service, or to build my own machine if I can find a place to colocate it that does not cost a fortune.

[majidm]

Just wanted to add that we're hitting the same problems - running CentOS 4.4 32-bit w/2.6.9-42.0.10.ELsmp on an Athlon64 w/3GB RAM. I managed to get the load to stop hitting 240 by reducing MaxClients (now we're hovering between 10 and 50, depending if it's morning or evening), but this isn't really acceptable - it's set to 50 now, and we serve more hits than that simultaneously. :(

So it's (likely) a kernel problem after all?

[EDIT] strangely, it was working more or less fine since we installed the server about 6 months ago, only in the last month or two has it started becoming a major problem...
[EDIT] just updated our kernel to -55. we'll see how the load goes - hovering between 10-15 right now...

[majidm]

Hehe... and therein lies a big part of my problem... In terms of money, I am a flea on an elephants butt to them... :p They have not been real helpful unless I am willing to spend money to pay them to trouble shoot their hardware :rolleyes: Soo... I am looking at options to either move to a new service, or to build my own machine if I can find a place to colocate it that does not cost a fortune.

I'm not sure who you're with, but I've found ServerBeach to be _very_ good when it comes to customer service, even if you're a small fly. Our server kept locking up when we got our first server with them, and they swapped everything for us - motherboards, chassis, everything - until they realised that the only constant, besides HDs, were the drive cables. Soon as they swapped those, the server ceased crashing. So now when I call my rep, I tell him "I'm the guy because of whom you changed your datacenter policies", hehe. I'd recommend them highly if you were willing to switch - Youtube didn't start with them for nothing.

[Tourmeister]

Just an interesting update:

First, after tweaking everything possible to optimal performance and still having problems, we replaced the network cards and cables. When the load spikes were not occuring, the forum was smoking fast. So it would appear we had some hardware issues.

Second, even with the new hardware and the performance tweaks, we still were getting load numbers up around 275-300!! Sometimes, the server would just freeze and the last shown load numbers were around 1.0-2.0. When it would finally come back, the load numbers were still in that same range, showing no indication of ever having spiked. After much logging, we finally deduced that we were being subjected to SYN flood attacks (form of Denial of Service attack). Turns out that my current provider DOES NOT implement any kind of DoS filtering above the individual server level :( As of