I run into a oops when mirgate /vz directory to ext4 lvm under gentoo + openvz.
ipv6 ~ # df -T
Filesystem Type Size Used Avail Use% Mounted on
/dev/sda2 ext3 9.9G 4.3G 5.1G 46% /
udev tmpfs 10M 112K 9.9M 2% /dev
/dev/sda1 ext3 198M 34M 155M 18% /boot
/dev/mapper/gwbn-lvz
ext4 40G 2.2G 36G 6% /vz
shm tmpfs 3.9G 0 3.9G 0% /dev/shm
And when this happens, vzctl will hang up; mount and unmount of /vz hangs too.
The word "ipv6" in console-output is just a hostname.
On Thu, Jul 22, 2010 at 05:39:16PM +0800, Drunkard Zhang wrote:
> I run into a oops when mirgate /vz directory to ext4 lvm under gentoo + openvz.
What were you doing precisely at the time of the oops? The oops logs
indicates it was a cp, but a cp of what? And is this somethign you
can easily reproduce?
- Ted
2010/7/22 Ted Ts'o <[email protected]>:
> On Thu, Jul 22, 2010 at 05:39:16PM +0800, Drunkard Zhang wrote:
>> I run into a oops when mirgate /vz directory to ext4 lvm under gentoo + openvz.
>
> What were you doing precisely at the time of the oops? The oops logs
> indicates it was a cp, but a cp of what? And is this somethign you
> can easily reproduce?
>
I did: "/etc/init.d/vz start" with vpsid 101 and 109, which from:
http://bb.xnull.de/projects/gentoo/amd64/stage4-amd64-current.tar.bz2
following the guide from:
http://www.gentoo.org/proj/en/vps/openvz-howto.xml
Didn't change the content of stage4-amd64-current.tar.bz2 yet.
It looks like when "/etc/init.d/vz start", it will cp all files from
/vz/private/101/ to /vz/root/root/101/
The oops allways reproduced yesterday between reboot and kernel
remake, but this morning when I trying to finger out the oops point
using strace, I can't get the oops anymore. Sorry for interrupt,:-(
Maybe distcc caused problem? I used to make kernel... Sorry again.
Drunkard Zhang <[email protected]> writes:
> 2010/7/22 Ted Ts'o <[email protected]>:
>> On Thu, Jul 22, 2010 at 05:39:16PM +0800, Drunkard Zhang wrote:
>>> I run into a oops when mirgate /vz directory to ext4 lvm under gentoo + openvz.
>>
>> What were you doing precisely at the time of the oops? The oops logs
>> indicates it was a cp, but a cp of what? And is this somethign you
>> can easily reproduce?
>>
Wow. Sorry almost missed that thread.
2.6.32-openvz-budarin.1 does not contains any specific patches for ext4.
So it is probably memory corruption caused by ovz stuff (2.6.32 is still
experimental), or you caught unknown bug, the later it very unlikely because
"cp" is very popular cmd :)
I'll try to reproduce your issue.
> I did: "/etc/init.d/vz start" with vpsid 101 and 109, which from:
> http://bb.xnull.de/projects/gentoo/amd64/stage4-amd64-current.tar.bz2
> following the guide from:
> http://www.gentoo.org/proj/en/vps/openvz-howto.xml
>
> Didn't change the content of stage4-amd64-current.tar.bz2 yet.
> It looks like when "/etc/init.d/vz start", it will cp all files from
> /vz/private/101/ to /vz/root/root/101/
That's not true, root/$veid is just stacked filesystem on top
of private/$veid.
>
> The oops allways reproduced yesterday between reboot and kernel
> remake, but this morning when I trying to finger out the oops point
> using strace, I can't get the oops anymore. Sorry for interrupt,:-(
>
> Maybe distcc caused problem? I used to make kernel... Sorry again.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html