[opensuse] Disks not starting...
Hi all, I'm having a bit of a problem with the newer 3.11.10-11 and 3.11.10-17 kernels. When I try to boot to either kernel, for some reason my disk partitions are not starting. I get a message indicating that there's a start job running for my disk partitions (I have two partitions, /boot and /), which eventually times out and the boot fails. When I boot using the 3.11.10-7 kernel, everything starts with no problems. I am using LVM, both partitions are on the same physical disk. Not sure what additional information would help debug the problem. Any suggestions as to how to resolve this would be greatly appreciated! I've tried fsk'ing the partitions, but that seems to have no effect. Brendan -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Hi all,
I'm having a bit of a problem with the newer 3.11.10-11 and 3.11.10-17 kernels. When I try to boot to either kernel, for some reason my disk partitions are not starting. I get a message indicating that there's a start job running for my disk partitions (I have two partitions, /boot and /), which eventually times out and the boot fails.
When I boot using the 3.11.10-7 kernel, everything starts with no
Op woensdag 9 juli 2014 20:41:17 schreef Brendan McKenna: problems.
I am using LVM, both partitions are on the same physical disk. Not sure what additional information would help debug the problem.
Any suggestions as to how to resolve this would be greatly
appreciated!
I've tried fsk'ing the partitions, but that seems to have no effect.
Brendan
I am not sure it is the same kernel, but I use the kernel from the factory repository and have a similar problem. However after a few restarts the system mounts all my partitions and works OK. I opened a bug report https://bugzilla.novell.com/show_bug.cgi?id=885986 -- fr.gr. Freek de Kruijf -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Hi Freek, Based on what you've put into your bug report, I don't think that this is the same problem. You indicate that it's happening with openSUSE 13.2, but I am (at least, according to /etc/issue) using 13.1. In my case, the problem comes up almost as soon as one of the newer kernel versions tries to start one of my hard disk partitions. The boot process doesn't appear to end at that point -- it continues on until it eventually reaches a state where I have a message indicating that there are still two start jobs running (they're trying to start the partitions -- each of them has the LVM name for a hard disk partition associated with it). Eventually, the wait for the start jobs times out and the boot fails at that point. As I indicated, I have fsk'ed both partitions and there don't appear to be any problems with the filesystems. I'd be happy to investigate further, if needed. Just need a pointer on where to look. Brendan On 10/07/14 12:51, Freek de Kruijf wrote:
Hi all,
I'm having a bit of a problem with the newer 3.11.10-11 and 3.11.10-17 kernels. When I try to boot to either kernel, for some reason my disk partitions are not starting. I get a message indicating that there's a start job running for my disk partitions (I have two partitions, /boot and /), which eventually times out and the boot fails.
When I boot using the 3.11.10-7 kernel, everything starts with no
Op woensdag 9 juli 2014 20:41:17 schreef Brendan McKenna: problems.
I am using LVM, both partitions are on the same physical disk. Not sure what additional information would help debug the problem.
Any suggestions as to how to resolve this would be greatly
appreciated!
I've tried fsk'ing the partitions, but that seems to have no effect.
Brendan
I am not sure it is the same kernel, but I use the kernel from the factory repository and have a similar problem. However after a few restarts the system mounts all my partitions and works OK. I opened a bug report https://bugzilla.novell.com/show_bug.cgi?id=885986
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 07/12/2014 07:34 AM, Brendan McKenna wrote:
In my case, the problem comes up almost as soon as one of the newer kernel versions tries to start one of my hard disk partitions. The boot process doesn't appear to end at that point -- it continues on until it eventually reaches a state where I have a message indicating that there are still two start jobs running (they're trying to start the partitions -- each of them has the LVM name for a hard disk partition associated with it).
Thre was a thread about this a couple or so of months ago, and I've been hit by it too with a recent kernel update. It has to do with the new way LVM is being handled. 'lvm metadata handler' was introduced. There is a setting in /etc/lvm/lvm.conf # Whether to use (trust) a running instance of lvmetad. If this is set to # *and* when lvmetad is running (it is not auto-started), the volume group # metadata and PV state flags are obtained from the lvmetad instance and no # scanning is done by the individual commands. In a setup with lvmetad, # lvmetad udev rules *must* be set up for LVM to work correctly. Without # If lvmetad has been running while use_lvmetad was 0, it MUST be stopped # before changing use_lvmetad to 1 and started again afterwards. It has been set to "1" by some upgrade and unless you have the lvmeta daemon running when you activate the lvm dire things happen. Go to that config and set: use_lvmetad = 0 I got bricked by an upgrade because my root is on lvm. I am very glad, normally, that my root is on lvm, but this came unexpectedly. -- /"\ \ / ASCII Ribbon Campaign X Against HTML Mail / \ -- "That's right; the upper-case shift works fine on the screen, but they're not coming out on the damn printer... Hold? Sure, I'll hold." -- e.e. cummings last service call -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Op zaterdag 12 juli 2014 12:34:27 schreef Brendan McKenna:
Hi Freek,
Based on what you've put into your bug report, I don't think that this is the same problem. You indicate that it's happening with openSUSE 13.2, but I am (at least, according to /etc/issue) using 13.1.
In my case, the problem comes up almost as soon as one of the newer kernel versions tries to start one of my hard disk partitions. The boot process doesn't appear to end at that point -- it continues on until it eventually reaches a state where I have a message indicating that there are still two start jobs running (they're trying to start the partitions -- each of them has the LVM name for a hard disk partition associated with it).
Eventually, the wait for the start jobs times out and the boot fails at that point. As I indicated, I have fsk'ed both partitions and there don't appear to be any problems with the filesystems.
I'd be happy to investigate further, if needed. Just need a pointer on where to look.
Brendan
On 10/07/14 12:51, Freek de Kruijf wrote:
Op woensdag 9 juli 2014 20:41:17 schreef Brendan McKenna:
Hi all,
I'm having a bit of a problem with the newer 3.11.10-11 and 3.11.10-17
kernels. When I try to boot to either kernel, for some reason my disk partitions are not starting. I get a message indicating that there's a start job running for my disk partitions (I have two partitions, /boot and /), which eventually times out and the boot fails.
When I boot using the 3.11.10-7 kernel, everything starts with no
problems.
I am using LVM, both partitions are on the same physical disk. Not
sure what additional information would help debug the problem.
Any suggestions as to how to resolve this would be greatly
appreciated!
I've tried fsk'ing the partitions, but that seems to have no effect.
Brendan
I am not sure it is the same kernel, but I use the kernel from the factory repository and have a similar problem. However after a few restarts the system mounts all my partitions and works OK. I opened a bug report https://bugzilla.novell.com/show_bug.cgi?id=885986
I don't think the bug has to do with plymouth. This morning I did not press Esc and still not all partitions got mounted. 13.2 uses a new kernel like your 13.1 system. When I press Esc I see my system hanging performing something with lvm and later, after a timeout, I see messages about not having mounted all the partitions. The suggestion to set use_lvmetad = 0 in lvm.conf does not apply in my case, it is already set at that value. -- fr.gr. member openSUSE Freek de Kruijf -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Hi Freek, After some additional prompting from Anton and some further investigation, I was able to resolve the problem. It turns out that there is a second copy of the lvm.conf file in the initrd for that kernel in /boot. I had to edit that copy of the file -- setting use_lvmetad = 0 in that copy of the file -- and then re-building the initrd. Once I did that I was able to boot the -17 kernel with no problems. Brendan On 12/07/14 23:16, Freek de Kruijf wrote:
Op zaterdag 12 juli 2014 12:34:27 schreef Brendan McKenna:
Hi Freek,
Based on what you've put into your bug report, I don't think that this is the same problem. You indicate that it's happening with openSUSE 13.2, but I am (at least, according to /etc/issue) using 13.1.
In my case, the problem comes up almost as soon as one of the newer kernel versions tries to start one of my hard disk partitions. The boot process doesn't appear to end at that point -- it continues on until it eventually reaches a state where I have a message indicating that there are still two start jobs running (they're trying to start the partitions -- each of them has the LVM name for a hard disk partition associated with it).
Eventually, the wait for the start jobs times out and the boot fails at that point. As I indicated, I have fsk'ed both partitions and there don't appear to be any problems with the filesystems.
I'd be happy to investigate further, if needed. Just need a pointer on where to look.
Brendan
On 10/07/14 12:51, Freek de Kruijf wrote:
Op woensdag 9 juli 2014 20:41:17 schreef Brendan McKenna:
Hi all,
I'm having a bit of a problem with the newer 3.11.10-11 and 3.11.10-17
kernels. When I try to boot to either kernel, for some reason my disk partitions are not starting. I get a message indicating that there's a start job running for my disk partitions (I have two partitions, /boot and /), which eventually times out and the boot fails.
When I boot using the 3.11.10-7 kernel, everything starts with no
problems.
I am using LVM, both partitions are on the same physical disk. Not
sure what additional information would help debug the problem.
Any suggestions as to how to resolve this would be greatly
appreciated!
I've tried fsk'ing the partitions, but that seems to have no effect.
Brendan
I am not sure it is the same kernel, but I use the kernel from the factory repository and have a similar problem. However after a few restarts the system mounts all my partitions and works OK. I opened a bug report https://bugzilla.novell.com/show_bug.cgi?id=885986
I don't think the bug has to do with plymouth. This morning I did not press Esc and still not all partitions got mounted. 13.2 uses a new kernel like your 13.1 system. When I press Esc I see my system hanging performing something with lvm and later, after a timeout, I see messages about not having mounted all the partitions.
The suggestion to set use_lvmetad = 0 in lvm.conf does not apply in my case, it is already set at that value.
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Brendan lvmetad has been problematic. for me, the solution was to set 'use_lvmetad = 0' in lvm.conf if you haven't seen them, these may clarify further Bug 876192 - LVM2 SEGFAULT; incorrectly reports "Incorrect metadata area header checksum" & prevents PV/VG/LV operations / upstream src build of LVM2 works without error https://bugzilla.novell.com/show_bug.cgi?id=876192 Bug 862076 - use_lvmetad = 1 in lvm.conf triggers systemd to get into emergency target on boot https://bugzilla.novell.com/show_bug.cgi?id=862076 Grant -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Hi Grant, You're right, that did fix the problem, but I also had to re-build the initrd file, as the one that's included in it is the one that's used initially. Thanks for the bug numbers, I'll keep an eye on those. Brendan On 13/07/14 14:58, grantksupport@operamail.com wrote:
Brendan
lvmetad has been problematic.
for me, the solution was to set 'use_lvmetad = 0' in lvm.conf
if you haven't seen them, these may clarify further
Bug 876192 - LVM2 SEGFAULT; incorrectly reports "Incorrect metadata area header checksum" & prevents PV/VG/LV operations / upstream src build of LVM2 works without error https://bugzilla.novell.com/show_bug.cgi?id=876192
Bug 862076 - use_lvmetad = 1 in lvm.conf triggers systemd to get into emergency target on boot https://bugzilla.novell.com/show_bug.cgi?id=862076
Grant
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Brendan, On Sun, Jul 13, 2014, at 08:49 AM, Brendan McKenna wrote:
You're right, that did fix the problem, but I also had to re-build the initrd file, as the one that's included in it is the one that's used initially.
Sure. For anyone else reading -- anytime anything 'disk' is modified (pv/vg/lv config, grub, raid, fstab, etc) -- it's a good idea to re-exec dracut/mkinitrd to make sure those changes are in your initrd. Easy enough to recover if you forget, but a hassle.
Thanks for the bug numbers, I'll keep an eye on those.
I've asked in one of the bugs 'what do you lose?' if lvmetad is changed from the default, =1->=0. No answer. I'm not clear on why the default is set to =1 in the 1st place. Afaict, =0 has had no ill effects here. Grant -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
participants (4)
-
Anton Aylward -
Brendan McKenna -
Freek de Kruijf -
grantksupport＠operamail.com