Re: ext3 filesystem corruption on md RAID1 device

[Dmitry Monakhov <dmonakhov@openvz.org>]

"Buehl, Reiner" <reiner.buehl@hp.com> writes:

> Hi Dimitry,
>
>> How does it happen? Do you remember any power failures happened
>> previously?
>
> I had no power failures. The first such error appears always approx. 2
> minutes after booting the system. This behavior started after I moved
> the md RAID fro
Hm.. So may be this is not a barrier related.
>m two smaller Seagate disks to the currently used 1TB disks. There for my first thought was a failed disk or memory but they seem to be ok based on the SMART and memtest results.
>
You may possibly force this to happen sooner like follows:
find /your-mnt/ -ls > /dev/null #fore fstat for all inodes
>> What mount options do you used? Probably you use default options
>> which means that fs was mounted w/o barrier support. Even if was
>> mounted
>> with barriers, your raid driver may simply ignore it.
>
> I use default options and software raid (mdadm). Does mdadm handle barriers correct?
>
>> Can you please post following info:
>> 1) mount options  and cat /proc/mount
>
> /dev/md1        /               ext3            defaults                        01
>
> bilbo:~# cat /proc/mounts 
> rootfs / rootfs rw 0 0
> none /sys sysfs rw,nosuid,nodev,noexec,relatime 0 0
> none /proc proc rw,nosuid,nodev,noexec,relatime 0 0
> udev /dev tmpfs rw,relatime,size=10240k,mode=755 0 0
> /dev/disk/by-uuid/a059fdf2-4ff6-4c30-ba7f-77c85e7f5d1b / ext3 rw,relatime,errors=continue,data=ordered 0 0
> tmpfs /lib/init/rw tmpfs rw,nosuid,relatime,mode=755 0 0
> usbfs /proc/bus/usb usbfs rw,nosuid,nodev,noexec,relatime 0 0
> tmpfs /dev/shm tmpfs rw,nosuid,nodev,relatime 0 0
> devpts /dev/pts devpts rw,nosuid,noexec,relatime,gid=5,mode=620,ptmxmode=000 0 0
> /dev/mapper/vg_data-lv_video /video ext3 rw,relatime,errors=continue,data=ordered 0 0
> /dev/mapper/vg_data-lv_mpg /video/film/mpg ext3 rw,relatime,errors=continue,data=ordered 0 0
> rpc_pipefs /var/lib/nfs/rpc_pipefs rpc_pipefs rw,relatime 0 0
> automount(pid3007) /mnt/auto autofs rw,relatime,fd=4,pgrp=3007,timeout=120,minproto=2,maxproto=4,indirect 0 0
> automount(pid2964) /mnt/usb autofs rw,relatime,fd=4,pgrp=2964,timeout=2,minproto=2,maxproto=4,indirect 0 0
> automount(pid3062) /net autofs rw,relatime,fd=4,pgrp=3062,timeout=300,minproto=2,maxproto=4,indirect 0 0
> nfsd /proc/fs/nfsd nfsd rw,relatime 0 0
>
>
>> 2) Write something to your fs and sync;
>>    like follows "echo test > /path_to_your_mnt/test ; sync"
>
> Done.
>
>> 3) dmesg log after stage(2)
>
> Nothing got written to dmesg during step 2.
This test was necessary to check that your md has barriers support
but this may happen only if you passed '-obarrier=1' explicitly
Since you use default mount options, then you use ext3 without barriers
so this test is not necessary
>
> I currently have 2 of the EXT3-fs errors in dmesg but writing and syncing did not produce another one. 
>
> Best regards,
> Reiner.