Kernel update to 2.6.20.1

When posting - remember to include details of your server type.

Moderator: LiamM

Kernel update to 2.6.20.1

Postby jimbob on Sat Mar 29, 2008 6:56 pm

An updated kernel for the following hardware is now available:
RaQ3, RaQ4, RaQ550, RaQXTR, Qube

This update is both a security and features update.

The update does contain new netfilter modules for all those who have emailed requesting the new features that they offer.

To install this update, please log on to your server, become root and do:

Code: Select all
wget http://www.osoffice.co.uk/linux/strongbolt/update_kernel.sh -O /tmp/update_kernel && sh /tmp/update_kernel


(The above line is all one command).

Please post any questions regarding the new kernel in this thread.

Jimbob
Strongbolt - new life for old raqs
jimbob
Site Admin
 
Posts: 347
Joined: Sat Apr 08, 2006 9:09 am
Location: Devon, UK

Postby macwunder on Sat Mar 29, 2008 8:16 pm

Sweet! We've got NAT... :lol:
Jeff -- Yellow RaQ XTR, 1x (so far) 1GHz P-III, 2GB RAM, 4 x 80GB RAID 1+0, SB1
PowerMac G3 500 MHz, 512 MB RAM, 40 GB, Ubuntu
HP DC7100 3 GHz P4 HT, 2.5 GB RAM, 4 x 320 GB RAID 1+0, BlueOnyx
HP has 12,000 BOGOMIPS Baby!!!
User avatar
macwunder
Legendary Forum member
 
Posts: 171
Joined: Tue Sep 19, 2006 9:34 pm
Location: Texas, USA

I may need to roll back, too

Postby SolidSilver on Sun Mar 30, 2008 4:20 pm

Reboot failed after I upgraded. I hooked up to the serial console and found that the RAID array isn't starting up. I did have to replace a disk a couple of weeks ago. The RAID rebuild seemed to work fine. Now I'm getting this:

Starting up RAID devices: md1 md2 /dev/md3: No such device
/dev/md3 is not a RAID0 or LINEAR array!
md3 md5 md6


The disks look like this:

(RAID Repair) 1 # fdisk -l

Disk /dev/hda: 122.9 GB, 122942324736 bytes
16 heads, 63 sectors/track, 238216 cylinders
Units = cylinders of 1008 * 512 = 516096 bytes

Device Boot Start End Blocks Id System
/dev/hda1 1 3969 2000344+ fd Linux raid autodetect
/dev/hda2 3970 7938 2000376 fd Linux raid autodetect
/dev/hda3 7939 9923 1000440 82 Linux swap
/dev/hda4 9924 238216 115059672 5 Extended
/dev/hda5 9924 13892 2000344+ fd Linux raid autodetect
/dev/hda6 13893 238216 113059264+ fd Linux raid autodetect

Disk /dev/hdc: 122.9 GB, 122942324736 bytes
255 heads, 63 sectors/track, 14946 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/hdc1 1 250 2000344+ fd Linux raid autodetect
Partition 1 does not end on cylinder boundary.
/dev/hdc2 250 499 2000376 fd Linux raid autodetect
Partition 2 does not end on cylinder boundary.
/dev/hdc3 499 623 1000440 82 Linux swap
Partition 3 does not end on cylinder boundary.
/dev/hdc4 623 14947 115059672 5 Extended
/dev/hdc5 623 872 2000344+ fd Linux raid autodetect
/dev/hdc6 872 14947 113059264+ fd Linux raid autodetect


The drives are the same size, but looks like different geometry. Am I screwed here?

(replies to codewhacker@gmail.com, please. my RAQ4 is my mail server)
SolidSilver
Forum member
 
Posts: 4
Joined: Wed Apr 11, 2007 3:04 pm

Postby SolidSilver on Sun Mar 30, 2008 4:54 pm

FYI, after a kernel rollback my RAQ is running again. (whew!)

For those who need to roll back, become root and cd to /boot. The update script backed up your old kernel to vmlinux.bz2_{old_version}{target}. Look for the most recent backup file to the date of the new vmlinux.bz2.

Do this:
mv vmlinux.bz2 vmlinux.bz2_2.6.20.1{target}
cp vmlinux.bz2_{old_version}{target} vmlinux.bz2
reboot


To get your target, do
cat /proc/cobalt/systype

Pacifica and Carmel are i586, Alpine and Monterey are i686.

If anyone has ideas on how to solve my RAID problem under 2.6.20.1, please send them along.
SolidSilver
Forum member
 
Posts: 4
Joined: Wed Apr 11, 2007 3:04 pm

Postby KenzieD on Mon Mar 31, 2008 2:56 pm

I have updated the kernel on the following with no problems
1xRaq3 (single drive)
1xRaq4 (single drive)
2xRaq4 (raid)

Then I tried my Symantec 5310 using the XTR kernel. It rebooted, but I wasn't able to get into the unit. Everything looked fine, but no admin page, no ping, nothing. I had stuck the dual nic back into the box because I didn't want to lose it. Then I tried plugging the cat5 into the one of the extra ports. Boom. Lookie what suddenly works now...

Image

Woohoo!!!!

I will try my Qube3 when I get home tonight.

Kenzie
KenzieD
Super Forum member
 
Posts: 35
Joined: Thu Feb 01, 2007 8:44 pm

Postby jimbob on Tue Apr 01, 2008 4:50 pm

KenzieD wrote:I have updated the kernel on the following with no problems
1xRaq3 (single drive)
1xRaq4 (single drive)
2xRaq4 (raid)

Then I tried my Symantec 5310 using the XTR kernel. It rebooted, but I wasn't able to get into the unit. Everything looked fine, but no admin page, no ping, nothing. I had stuck the dual nic back into the box because I didn't want to lose it. Then I tried plugging the cat5 into the one of the extra ports. Boom. Lookie what suddenly works now...


Hello KenzieD, Glad to hear that you have not had any problems updating your systems.
Your additional NIC card must have been an e100?
The new kernel has support built in for all Cobalt hardware for ease of maintenance.
When adding additional network cards to Linux systems, be aware that additional configuration is required to enure that the present NIC's do not get re-assigned to new device names.

Please let me know if you encounter any problems with your new network configuration.

Jimbob
Strongbolt - new life for old raqs
jimbob
Site Admin
 
Posts: 347
Joined: Sat Apr 08, 2006 9:09 am
Location: Devon, UK

Posting information about problems

Postby jimbob on Tue Apr 01, 2008 5:00 pm

Hello All,

Please remember to post as much relevant information as possible when posting about a problem. This will aid in identifying the issue more quickly and more effectively.

Also, please email details to support@osoffice.co.uk

When reporting a fault, please remember to include details of :

1. What kind of server is it?
2. Number of drives and drive configuration (RAID?)
3. Do you have any additional PCI cards in the server?
4. What is the server used for?
5. Does it sustain a high load?
6. How old are the fans?
7. Whats the average temperature the server runs at?
8. Has the server got a history of crashing?
9. Have you required any custom kernel modules?

Thanks,

Jimbob
Strongbolt - new life for old raqs
jimbob
Site Admin
 
Posts: 347
Joined: Sat Apr 08, 2006 9:09 am
Location: Devon, UK

Re: I may need to roll back, too

Postby jimbob on Tue Apr 01, 2008 5:22 pm

Hello Solidsilver,

Disk /dev/hda: 122.9 GB, 122942324736 bytes
16 heads, 63 sectors/track, 238216 cylinders
Units = cylinders of 1008 * 512 = 516096 bytes

Disk /dev/hdc: 122.9 GB, 122942324736 bytes
255 heads, 63 sectors/track, 14946 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes



To prove that this is happening consistently, please try the following:
Reboot 3 times to the old Kernel and confirm correct operation (no RAID issues)
Reboot 3 times to the new kernel and confirm if it fails.

Since Linux 2.5.51, the IDE driver no longer uses BIOS info on the first two disks.
It is curious that you have experienced this when moving from one 2.6 kernel to another 2.6 kernel. I will try and offer some suggestions to fix this.

If anyone else experiences this, please let me know.

Thanks

Jimbob
:)
Strongbolt - new life for old raqs
jimbob
Site Admin
 
Posts: 347
Joined: Sat Apr 08, 2006 9:09 am
Location: Devon, UK

Postby KenzieD on Thu Apr 03, 2008 4:36 pm

jimbob wrote:Please let me know if you encounter any problems with your new network configuration.
Jimbob


Hi Jimbob. Not really a problem, but something that confused me at first. When I look at the TCP/IP settings in the GUI, it shows me the same MAC address for Eth0 & Eth2, and then the same for Eth1 & Eth3. However, it shows me everything correctly if i do ifconfig on the console (4 different mac addresses).
Just to see what it would do, I pluged in a 4 port nic, and I got the same MAC on 0, 2, & 4, and then the same on 1, 3, & 5. But again they were all different using ifconfig.
It does seem to be working properly, but just reports the wrong information to the GUI.

Take care,
Kenzie
KenzieD
Super Forum member
 
Posts: 35
Joined: Thu Feb 01, 2007 8:44 pm

Kernel 2.6.20.1 HIGHMEM

Postby raq-recl on Sun Apr 13, 2008 5:52 pm

Last week I upgrade to kernel 2.6.20.1 on a RaQ4 I use to host small websites.
After rebooting I noticed that the new kernel didn't recognize the 1GB RAM
on the machine, but just 896 MB.
I've experienced the same problem with the 2.6.16 supplied w/ SB 1.06-02, which I corrected myself by recompiling the kernel with CONFIG_HIGHMEM4G enabled and did the same with the new 2.6.20.1.

Of course the above is not a complaint or something like that, but just a suggestion to Jimbob to include 1GB> support in the next updates or perhaps patch the current ( 2.6.20.x) kernel .
raq-recl
Forum member
 
Posts: 2
Joined: Sat Apr 12, 2008 8:55 pm

Re: Kernel 2.6.20.1 HIGHMEM

Postby TimSB on Mon Apr 14, 2008 8:38 am

raq-recl wrote:Last week I upgrade to kernel 2.6.20.1 on a RaQ4 I use to host small websites.
After rebooting I noticed that the new kernel didn't recognize the 1GB RAM
on the machine, but just 896 MB.
I've experienced the same problem with the 2.6.16 supplied w/ SB 1.06-02, which I corrected myself by recompiling the kernel with CONFIG_HIGHMEM4G enabled and did the same with the new 2.6.20.1.

Of course the above is not a complaint or something like that, but just a suggestion to Jimbob to include 1GB> support in the next updates or perhaps patch the current ( 2.6.20.x) kernel .


IIRC, Tim Hockin advised me that the memory limit is a hardware limitation, rather than something that can be fixed in software.

I think the number of available physical address lines "runs out", so the memory is limited to something like 880Mb, which in practical terms means you can either fit a 512Mb+256Mb stick or use 2x 512Mb sticks, but find that only 880Mb is "available".

So, whether you used the default current Strongbolt kernel, or upgraded to the new one, the same problem will occur.

I seem to recall a thread on the forum already about this...try this one:

http://www.osoffice.co.uk/forum/viewforum.php?f=15

regards

Tim
Last edited by TimSB on Tue Apr 15, 2008 10:37 pm, edited 1 time in total.
Any advice I may give is given in good faith but may be incorrect so listen to what other people have to say as well and don't blame me when it all goes tits up.
TimSB
Forum Admin
 
Posts: 561
Joined: Fri Jun 16, 2006 12:30 pm
Location: north of London, UK

IPTables support

Postby plodder on Tue Apr 15, 2008 7:22 pm

Hi.

I'm trying to understand the level of iptables support included in the latest and previous versions of the kernel (i.e. 2.6.20.1 & 2.6.16). I have a mix of RaQ 4 and RaQ 550 servers and am trying to get APF (Advanced Policy Firewall) - as supplied as part of Solarspeed's Security package (v5 beta) - working. What I've found is that it starts on 2.6.16 but there are some iptables messages and I think it isn't able to add new rules as it should. I'm told the command "lsmod | grep iptable" should give a response something like the one below...

Code: Select all
iptable_mangle          2881  1
iptable_filter          3009  1
ip_tables              17600  8
ipt_state,ipt_TOS,iptable_mangle,ipt_REJECT,ipt_LOG,ipt_limit,ipt_multiport,iptable_filter


indicating that iptables is as needed.

On my servers I get just...

Code: Select all
iptable_filter          1440  1
ip_tables              10328  1 iptable_filter


As I say, that's all using the 2.6.16 kernel. If I try 2.6.20.1 then APF doesn't even start (I think because there's even less iptables support).

I wish I knew more about compiling kernels etc. to include everything needed for iptables support, but I don't! Does this information help to explain my problem and, hopefully, allow for a new kernel to be compiled and distributed to solve it?

Thanks!
Servers:
4x RaQ 550 (1.26GHz / 2GB / 2x 500GB)
1x RaQ 4i (450MHz / 512MB / 2x 80GB)
1x RaQ 3 (300MHz / 512MB / 1x 120GB)

Disks:
Hitachi HDP725050GLAT80 500GB PATA
Hitachi HDS721680PLAT80 80GB PATA
Seagate ST3120022A 120GB PATA
plodder
Forum member
 
Posts: 14
Joined: Wed Feb 20, 2008 12:48 am

i2c modules missing

Postby atorox on Sun Apr 20, 2008 3:25 pm

Just installed new kernel on RAQ.

Processors 1
Model AMD-K6(tm)-III Processor
CPU Speed 547.81 MHz (400MHz overclocked)
IDE Devices hda: Maxtor 6E040L0 (Capacity: 38.29 GB)
hdc: Maxtor 6E040L0 (Capacity: 38.29 GB)
USB Devices Linux 2.6.20.1i586 ohci_hcd OHCI Host Controller

Tried to run sensors to see cpu temperature but there are no i2c modules.
atorox
Forum member
 
Posts: 3
Joined: Fri Jan 25, 2008 9:16 am

Postby macwunder on Mon Apr 21, 2008 3:55 pm

Code: Select all
cat /proc/cobalt/sensors/thermal


Works for me...

Code: Select all
0 [CPU]: 34
[root@raq ~]#
Jeff -- Yellow RaQ XTR, 1x (so far) 1GHz P-III, 2GB RAM, 4 x 80GB RAID 1+0, SB1
PowerMac G3 500 MHz, 512 MB RAM, 40 GB, Ubuntu
HP DC7100 3 GHz P4 HT, 2.5 GB RAM, 4 x 320 GB RAID 1+0, BlueOnyx
HP has 12,000 BOGOMIPS Baby!!!
User avatar
macwunder
Legendary Forum member
 
Posts: 171
Joined: Tue Sep 19, 2006 9:34 pm
Location: Texas, USA

ipv6?

Postby KabEe on Thu Jul 10, 2008 9:57 am

hi there,
the new kernel doesn't support ipv6?

FATAL: Module ipv6 not found.
CRITICAL : [ipv6_test] Kernel is not compiled with IPv6 support
KabEe
Forum member
 
Posts: 2
Joined: Wed Jul 09, 2008 1:18 pm

Postby jimbob on Fri Jul 11, 2008 3:53 pm

Hello,

The old kernel did not support ipv6 either.
BlueQuartz has not been ported to IPv6 yet, and the interface would break.

Jimbob
Strongbolt - new life for old raqs
jimbob
Site Admin
 
Posts: 347
Joined: Sat Apr 08, 2006 9:09 am
Location: Devon, UK

Re: Kernel update to 2.6.20.1

Postby mthielemann on Sun Apr 25, 2010 5:15 am

Unfortunately this kernel seems not to be available for download anymore. Could anyone please send me the kernel for Qube 3 or tell me how to get it? Thanks for your help.

Regards,
Marc
mthielemann
Forum member
 
Posts: 24
Joined: Sat Aug 26, 2006 9:39 pm
Location: Berlin, Germany

Re: Kernel update to 2.6.20.1

Postby Mad Professor on Sun Apr 25, 2010 9:31 am

mthielemann: osOffice have now password protected the folders that where once open access.

If you send me a PM with the file(s) you are after and an e-mail address, I will send the files to you asap.

Best Regards.
Mad Professor
Super Forum member
 
Posts: 39
Joined: Tue Dec 16, 2008 6:50 pm

Re: Kernel update to 2.6.20.1

Postby mthielemann on Sun Apr 25, 2010 2:43 pm

Thanks for your help. Unfortunately this kernel seems only to work with SB1 but I use SB2 ... :(

Regards,
Marc
mthielemann
Forum member
 
Posts: 24
Joined: Sat Aug 26, 2006 9:39 pm
Location: Berlin, Germany

Re: Kernel update to 2.6.20.1

Postby tonj on Sat Jul 31, 2010 11:02 am

@plodder:
regarding those Hitachi HDP725050GLAT80 500GB drives...
do they actually fit properly into a raq550 or did you have to mod the fitting in any way?
tonj
Super Forum member
 
Posts: 34
Joined: Thu Nov 26, 2009 3:52 pm

Re: Kernel update to 2.6.20.1

Postby indy on Tue Feb 01, 2011 3:02 am

I do the kernel update to 2.6.20.1 on RaQ550.
It's works fine.

But, my addon PCI serial card is still not work....

Checking the /boot/config-2.6.20.1i686

#
# Serial drivers
#
CONFIG_SERIAL_8250=y
CONFIG_SERIAL_8250_CONSOLE=y
CONFIG_SERIAL_8250_PCI=y
CONFIG_SERIAL_8250_NR_UARTS=2
CONFIG_SERIAL_8250_RUNTIME_UARTS=2
# CONFIG_SERIAL_8250_EXTENDED is not se

So, only 2 UARTS ports activate.....

I'd like to recompiling 2.6.20.1i686 kernel.
Please let me know url for download 2.6.20.1i686 kernel.

Thanks.

INDY
indy
Forum member
 
Posts: 3
Joined: Thu Jan 27, 2011 2:47 am

Re: Kernel update to 2.6.20.1

Postby indy on Wed Feb 16, 2011 2:56 am

Are no one here?

Please open the Kernel 2.6.20.1 source code.

Please teach the reason it is not possible to open it to the public.

Thanks.

INDY
indy
Forum member
 
Posts: 3
Joined: Thu Jan 27, 2011 2:47 am


Return to Kernel issues

Who is online

Users browsing this forum: No registered users and 1 guest

cron