Re: Linux 4.0.0-rc5 - [ 1851.858652] BTRFS: error (device sda1) in cleanup_transaction:1686: errno=-17 Object already exists

[Torbjørn <lists@skagestad.org>]

On 03/31/2015 08:12 PM, Chris Mason wrote:
>
>
> On Mon, Mar 30, 2015 at 1:42 PM, Torbjørn <lists@skagestad.org> wrote:
>> Hi,
>>
>> Just a follow up on this report.
>>
>> The file system in question is a raid1 across 2x320G old Western 
>> Digital WD3200KS.
>> I janked them out of the server to run a fsck on another computer 
>> (after a proper shutdown).
>>
>> One of the disks did not get properly detected on the secondary 
>> computer.
>> Hopefully the fsck of the single disk is still of some value to you.
>>
>> As you can see, there are several issues with the fs.
>> The system has occasionally had hard reboots.
>>
>> The fs does not have any real value for me. Everything worth anything 
>> is backed up.
>> I'll keep the drive around in case it's of any value for some devs.
>>
>> As noted before: this (corrupted) fs only get errors when booting 
>> into 4.0-rc5. With 4.0-rc4 or earlier it works as if nothing is wrong.
>
> This is really strange because we also have reports from v3.19 stable 
> kernels, but none of the btrfs patches between rc4 and rc5 were tagged 
> for stable.
>
> Can I convince you to hammer a bit more on rc4?  I'd like to make sure 
> it really was a regression introdcued in rc5.
>
> -chris
>
>
>
>
Perhaps I was a bit unclear. The error is triggered when booting into 
rc5. If I reset and try to start up rc4 or earlier, it is still there. 
After zero-log I can boot into rc4.

I already did a fresh reinstall to alternative drives. I did not want to 
have the corrupted fs as root.

I'm not sure I can get the second disk in the raid1 to work. It was head 
crashing when trying to attach it for fsck. If I somehow get it back 
online I can do some more testing. Anything in particular?

--
Torbjørn