---On Sat, 2006-09-02 at 07:42 -0500, Gerald Waugh wrote:
We have a problem with upgrade to 4.4 When we run 'yum update' everything appears to be OK But, if we reboot the server with the new kernel, we get "kernel panic" when the new kernel attempts to load.
The CentOS splash screen shows briefly, but not long enough to select the old kernel
Gerald
(let's try it like this ... johnny :)
On Sat, 2006-09-02 at 10:50 -0500, Johnny Hughes wrote:
---On Sat, 2006-09-02 at 07:42 -0500, Gerald Waugh wrote:
We have a problem with upgrade to 4.4 When we run 'yum update' everything appears to be OK But, if we reboot the server with the new kernel, we get "kernel panic" when the new kernel attempts to load.
The CentOS splash screen shows briefly, but not long enough to select the old kernel
Hmmm ...
You should get a screen that has a kernel name and is blue, at that screen, you should be able to press any key and then select a kernel before the kernel actually starts booting.
Does that work?
The other issue sounds like maybe you have a custom built kernel driver that you need to build for your new kernel before you boot (maybe a SCSI/RAID driver)?
On Sat, 2006-09-02 at 10:54, Johnny Hughes wrote:
We have a problem with upgrade to 4.4 When we run 'yum update' everything appears to be OK But, if we reboot the server with the new kernel, we get "kernel panic" when the new kernel attempts to load.
The CentOS splash screen shows briefly, but not long enough to select the old kernel
Hmmm ...
You should get a screen that has a kernel name and is blue, at that screen, you should be able to press any key and then select a kernel before the kernel actually starts booting.
Does that work?
If that doesn't work, boot the install CD with 'linux rescue' at the boot prompt. It should auto-detect your installed system, mount it, and suggest that you chroot /mnt/sysinstall do that, edit /etc/grub.conf to default to your old kernel, then exit twice to reboot.
You should get a screen that has a kernel name and is blue, at that screen, you should be able to press any key and then select a kernel before the kernel actually starts booting.
Does that work?
The screen comes up, and both the 34 and 42 kernels are displayed, but only for an instant. Is there a way to make this screen delay longer?
The other issue sounds like maybe you have a custom built kernel driver that you need to build for your new kernel before you boot (maybe a SCSI/RAID driver)?
No, using the standard kernel that was in 4.3, and it works.
Gerald
On Sat, 2006-09-02 at 11:57 -0500, Gerald Waugh wrote:
You should get a screen that has a kernel name and is blue, at that screen, you should be able to press any key and then select a kernel before the kernel actually starts booting.
Does that work?
The screen comes up, and both the 34 and 42 kernels are displayed, but only for an instant. Is there a way to make this screen delay longer?
Boot linux rescue and edit grub.conf to give a longer delay? If you see the screen at all, it is finding the grub stuf and displaying. I bet the delay was reduced for production considerations.
The other issue sounds like maybe you have a custom built kernel driver that you need to build for your new kernel before you boot (maybe a SCSI/RAID driver)?
No, using the standard kernel that was in 4.3, and it works.
Gerald
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Johnny Hughes wrote:
On Sat, 2006-09-02 at 10:50 -0500, Johnny Hughes wrote:
---On Sat, 2006-09-02 at 07:42 -0500, Gerald Waugh wrote:
We have a problem with upgrade to 4.4 When we run 'yum update' everything appears to be OK But, if we reboot the server with the new kernel, we get "kernel panic" when the new kernel attempts to load.
The CentOS splash screen shows briefly, but not long enough to select the old kernel
Hmmm ...
You should get a screen that has a kernel name and is blue, at that screen, you should be able to press any key and then select a kernel before the kernel actually starts booting.
Does that work?
The other issue sounds like maybe you have a custom built kernel driver that you need to build for your new kernel before you boot (maybe a SCSI/RAID driver)?
Since I'm still happily running CentOS 4.3, I should probably keep my thoughts to myself, but that immediate kernel panic sure is a reminder of attempting to boot an i686 kernel with an i586 processor. I'd try to duplicate the reported symptom but my AMD K6-2 box has become an "organ transplant" donor strictly because of kernel requirements.
Since I'm still happily running CentOS 4.3, I should probably keep my thoughts to myself, but that immediate kernel panic sure is a reminder of attempting to boot an i686 kernel with an i586 processor. I'd try to duplicate the reported symptom but my AMD K6-2 box has become an "organ transplant" donor strictly because of kernel requirements.
It's a Tyan GS14 Celeron D 2.53GHz My console output reads (in-part)
ata1: SATA max UDMA/133 cmd 0x30D0 ctl 0x30C6 bmdma 0x3090 irq 10 ata2: SATA max UDMA/133 cmd 0x30C8 ctl 0x30C2 bmdma 0x3098 irq 10 ata1: dev 0 ATA-7, max UDMA/133, 234441648 sectors: LBA48 ata1: qc timeout (cmd 0xef) ata1: failed to set xfermode (err_mask=0x4) scsi0 : ata_piix ata2: SATA port has no device. scsi1 : ata_piix Loading dm-mod.ko module device-mapper: 4.5.0-ioctl (2005-10-04) initialised: dm-devel@redhat.com Loading jbd.ko module Loading ext3.ko module Loading dm-mirror.ko module Loading dm-zero.ko module Loading dm-snapshot.ko module Making device-mapper control node Scanning logical volumes RKernel panic - not syncing: Attempted to kill init! eading all physi cal volumes. This may take a while... No volume groups found Activating logical volumes Volume group "VolGroup00" not found ERROR: /bin/lvm exited abnormally! (pid 318)
Gerald
On Sat, 2006-09-02 at 12:08 -0500, Gerald Waugh wrote:
<snip>
It's a Tyan GS14 Celeron D 2.53GHz My console output reads (in-part)
ata1: SATA max UDMA/133 cmd 0x30D0 ctl 0x30C6 bmdma 0x3090 irq 10 ata2: SATA max UDMA/133 cmd 0x30C8 ctl 0x30C2 bmdma 0x3098 irq 10 ata1: dev 0 ATA-7, max UDMA/133, 234441648 sectors: LBA48 ata1: qc timeout (cmd 0xef) ata1: failed to set xfermode (err_mask=0x4) scsi0 : ata_piix ata2: SATA port has no device. scsi1 : ata_piix Loading dm-mod.ko module device-mapper: 4.5.0-ioctl (2005-10-04) initialised: dm-devel@redhat.com Loading jbd.ko module Loading ext3.ko module Loading dm-mirror.ko module Loading dm-zero.ko module Loading dm-snapshot.ko module Making device-mapper control node Scanning logical volumes RKernel panic - not syncing: Attempted to kill init!
I've seen this before... while developing my fallback procedure. If my brain was younger...
eading all physi cal volumes. This may take a while... No volume groups found Activating logical volumes Volume group "VolGroup00" not found ERROR: /bin/lvm exited abnormally! (pid 318)
I don't use sata (yet) or raid or jbod but this sure sounds like the initrd problem I encountered. Boot into rescue mode and do a pvscan -- verbose.
In *my* case, I had to make my fallback volume a different volume group. Trying to boot, it seems there is a hard-coded VolGroup00/LogVol00 in the init file in the initrd. It is on an ignorelockingfailure. I edited that file remade my initrd, did grub-install and all worked.
Does you /etc/fstab show the full /VolGorup or /dev/mapper by any chance? Maybe it's not VolGroup00?
Anyway, I don't know why yours would be different, but I hoped this might be a clue.
Gerald
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
On Sat, 2006-09-02 at 12:08 -0500, Gerald Waugh wrote:
<snip>
In the forum here
http://www.centos.org/modules/newbb/viewtopic.php?topic_id=5087&forum=27
Same complaint essentially. Believes nothing changed but the kernel (installed the .42 a few days ago) and gets a panic trying to kill init. Boots old kernel and Nirvana is claimed.
Hoped it might give a clue.
Bill
The same problem happened with a server after I upgrade the kernel to 2.6.9-42.0.2.
Another server, with similar configuration, boots normally on new kernel.
Both uses software RAID and LVM2, (/boot on md0 and Logical Volume on md1). Only differences between the servers is HD, on first have two Seagate (ST380013AS Rev: 3.18) 80 GB, another have two Western Digital (WDC WD2000JD-55H Rev: 08.0) 120 GB.
As it is a production server, I didn't make any test to investigate the problem, just boot on previous kernel.
Heitor A. M. Cardozo
----- Original Message ----- From: "Gerald Waugh" gwaugh@frontstreetnetworks.com To: "'CentOS mailing list'" centos@centos.org Sent: Saturday, September 02, 2006 2:08 PM Subject: RE: [CentOS] Kernel-2.6.9-42.0.2EL fails to boot after upgradetoCentOS 4.4
Since I'm still happily running CentOS 4.3, I should probably keep my thoughts to myself, but that immediate kernel panic sure is a reminder of attempting to boot an i686 kernel with an i586 processor. I'd try to duplicate the reported symptom but my AMD K6-2 box has become an "organ transplant" donor strictly because of kernel requirements.
It's a Tyan GS14 Celeron D 2.53GHz My console output reads (in-part)
ata1: SATA max UDMA/133 cmd 0x30D0 ctl 0x30C6 bmdma 0x3090 irq 10 ata2: SATA max UDMA/133 cmd 0x30C8 ctl 0x30C2 bmdma 0x3098 irq 10 ata1: dev 0 ATA-7, max UDMA/133, 234441648 sectors: LBA48 ata1: qc timeout (cmd 0xef) ata1: failed to set xfermode (err_mask=0x4) scsi0 : ata_piix ata2: SATA port has no device. scsi1 : ata_piix Loading dm-mod.ko module device-mapper: 4.5.0-ioctl (2005-10-04) initialised: dm-devel@redhat.com Loading jbd.ko module Loading ext3.ko module Loading dm-mirror.ko module Loading dm-zero.ko module Loading dm-snapshot.ko module Making device-mapper control node Scanning logical volumes RKernel panic - not syncing: Attempted to kill init! eading all physi cal volumes. This may take a while... No volume groups found Activating logical volumes Volume group "VolGroup00" not found ERROR: /bin/lvm exited abnormally! (pid 318)
Gerald
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
The same problem happened with a server after I upgrade the kernel to 2.6.9-42.0.2.
Another server, with similar configuration, boots normally on new kernel.
Both uses software RAID and LVM2, (/boot on md0 and Logical Volume on md1). Only differences between the servers is HD, on first have two Seagate (ST380013AS Rev: 3.18) 80 GB, another have two Western Digital (WDC WD2000JD-55H Rev: 08.0) 120 GB.
As it is a production server, I didn't make any test to investigate the problem, just boot on previous kernel.
Our server is a test (sandbox) server, we use LILO, we increase the timeout to 120, And this gives us time to select the old kernel.
Thanks to all who responded to the problem with the new kernel, we appreciate your input.
The server runs fine with the update on the old kernel.... That's OK, but why won't it run on the -42.0.2 kernel???
Gerald
Heitor Augusto M. Cardozo schrieb:
The same problem happened with a server after I upgrade the kernel to 2.6.9-42.0.2.
Another server, with similar configuration, boots normally on new kernel.
Both uses software RAID and LVM2, (/boot on md0 and Logical Volume on md1). Only differences between the servers is HD, on first have two Seagate (ST380013AS Rev: 3.18) 80 GB, another have two Western Digital (WDC WD2000JD-55H Rev: 08.0) 120 GB.
As it is a production server, I didn't make any test to investigate the problem, just boot on previous kernel.
Heitor A. M. Cardozo
Just a short note: the kernel 2.6.9-42.0.2.EL boots fine now on my RAID1 machine. I´m gonna test this days the one from the plus repo.
Kind regards,
Chris
A little birdy told me that Christian H�gel said:
] Heitor Augusto M. Cardozo schrieb: ] > The same problem happened with a server after I upgrade the kernel ] > to 2.6.9-42.0.2. ] > ] > Another server, with similar configuration, boots normally on new ] > kernel. ] > ] > Both uses software RAID and LVM2, (/boot on md0 and Logical Volume ] > on md1). Only differences between the servers is HD, on first have ] > two Seagate (ST380013AS Rev: 3.18) 80 GB, another have two ] > Western Digital (WDC WD2000JD-55H Rev: 08.0) 120 GB. ] > ] > As it is a production server, I didn't make any test to investigate ] > the problem, just boot on previous kernel. ] > ] > Heitor A. M. Cardozo ] > ] > ] ] Just a short note: the kernel 2.6.9-42.0.2.EL boots fine now on my RAID1 ] machine. ] I�m gonna test this days the one from the plus repo.
just to chime in... in case it helps anyone identify a problem... but it wasn't the 2.6.9-42.0.2.EL kernel which broke my LILO boots on RAIDed drives... i think it was more likely something like "udev"... as i was running the 2.6.9-42.0.2.EL fine before the larger v4.4 update came out... and LILO did not demonstrate the problems i mentioned... in any case... i wasn't keen on retrograding other packages to stabilize the problem... instead i chose to jump LILO ahead (the new LILO has so many nice features... even just the "-T" switches made it worthwhile)... the "grubby" problem was something else altogether... that's purely a RH package that has been let slip in its maintenance... i wasn't heavily interested in all the maintenance that it needed... but i'm a stickler about making sure things "work" the way they should...
B. Karhan simon@pop.psu.edu PRI/SSRI Unix Administrator
The screen comes up, and both the 34 and 42 kernels are displayed, but only for an instant. Is there a way to make this screen delay longer?
Boot to the install cd's, do linux rescue, and chroot /mnt/sysimage. Then edit your /boot/grub/grub.conf, setting the line timeout=something to timeout=15. Then see if you cant boot to the old kernel.
It's a Tyan GS14 Celeron D 2.53GHz My console output reads (in-part)
ata1: SATA max UDMA/133 cmd 0x30D0 ctl 0x30C6 bmdma 0x3090 irq 10 ata2: SATA max UDMA/133 cmd 0x30C8 ctl 0x30C2 bmdma 0x3098 irq 10 ata1: dev 0 ATA-7, max UDMA/133, 234441648 sectors: LBA48 ata1: qc timeout (cmd 0xef) ata1: failed to set xfermode (err_mask=0x4) scsi0 : ata_piix ata2: SATA port has no device. scsi1 : ata_piix Loading dm-mod.ko module device-mapper: 4.5.0-ioctl (2005-10-04) initialised: dm-devel@redhat.com Loading jbd.ko module Loading ext3.ko module Loading dm-mirror.ko module Loading dm-zero.ko module Loading dm-snapshot.ko module Making device-mapper control node Scanning logical volumes RKernel panic - not syncing: Attempted to kill init! eading all physi cal volumes. This may take a while... No volume groups found Activating logical volumes Volume group "VolGroup00" not found ERROR: /bin/lvm exited abnormally! (pid 318)
It looks like to me that Johnny was right:
The other issue sounds like maybe you have a custom built kernel driver that you need to build for your new kernel before you boot (maybe a SCSI/RAID driver)?
On Sat, 2006-09-02 at 11:19 -0600, Joshua Gimer wrote:
It's a Tyan GS14 Celeron D 2.53GHz My console output reads (in-part)
<snip>
It looks like to me that Johnny was right:
The other issue sounds like maybe you have a custom built kernel
driver
that you need to build for your new kernel before you boot (maybe a SCSI/RAID driver)?
Or just a custom initrd that loads a driver module? Maybe no custom kernel is needed? Unzip and extract the two initrds will tell.
BTW, space? Did the map file get installed properly?
-- Thx Joshua Gimer _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
On Sat, 2006-09-02 at 13:46 -0400, William L. Maltby wrote:
On Sat, 2006-09-02 at 11:19 -0600, Joshua Gimer wrote:
It's a Tyan GS14 Celeron D 2.53GHz My console output reads (in-part)
<snip>
It looks like to me that Johnny was right:
The other issue sounds like maybe you have a custom built kernel
driver
that you need to build for your new kernel before you boot (maybe a SCSI/RAID driver)?
If you get your old kernel booted Gerald, "uname -a" should convince the skeptics. ;-)
Or just a custom initrd that loads a driver module? Maybe no custom kernel is needed? Unzip and extract the two initrds will tell.
... uh after a diff -u type compare! If there are changes that are not expected, like an insmod or modprope or unexpected diff in the modprobe.conf ...
BTW, space? Did the map file get installed properly?
-- Thx Joshua Gimer _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Johnny Hughes schrieb:
---On Sat, 2006-09-02 at 07:42 -0500, Gerald Waugh wrote:
We have a problem with upgrade to 4.4 When we run 'yum update' everything appears to be OK But, if we reboot the server with the new kernel, we get "kernel panic" when the new kernel attempts to load.
The CentOS splash screen shows briefly, but not long enough to select the old kernel
Gerald
(let's try it like this ... johnny :)
What kernel is that?The one from centos-plus?
Chris
-
Benjamin Karhan -
Christian Hügel -
Gerald Waugh -
Heitor Augusto M. Cardozo -
Johnny Hughes -
Joshua Gimer -
Les Mikesell -
Robert -
William L. Maltby