On 04/07/2014 15:25, Wido den Hollander wrote:
> On 07/04/2014 03:18 PM, Loic Dachary wrote:
>> Hi,
>>
>> I extracted a disk with two partitions (journal and data) and copied its 
>> content in the hope to restart the OSD and recover its content.
>>
>>     mount /dev/sdb1 /mnt
>>     rsync -avH --numeric-ids /mnt/ /var/lib/ceph/osd/ceph-$(cat /mnt/whoami)/
> 
> I think you went wrong there, rsync man page:
> 
> -a, --archive               archive mode; equals -rlptgoD (no -H,-A,-X)
> -X, --xattrs                preserve extended attributes
> 
> So you didn't copy over the xattrs, so basically the data is lost/unusable.

Thanks ! Fortunately the original disks are still available ;-)

>>     rm /var/lib/ceph/osd/ceph-$(cat /mnt/whoami)/journal
>>     dd if=/dev/sdb2 of=/var/lib/ceph/osd/ceph-$(cat /mnt/whoami)/journal
>>
>> and then
>>
>>     start ceph-osd id=$(cat /mnt/whoami)
>>
>> It crashes on https://github.com/ceph/ceph/blob/v0.72.2/src/osd/PG.cc#L2182 
>> and before it happens there is
>>
>>     load_pgs ignoring unrecognized meta
>>
>> and the full "debug osd = 20" logs are in http://paste.ubuntu.com/7746993/ 
>> and this is
>>
>> root@bm4202:/etc/ceph# dpkg -l | grep ceph
>> ii  ceph                        0.72.2-1trusty     amd64              
>> distributed storage and file system
>> ii  ceph-common                 0.72.2-1trusty     amd64              common 
>> utilities to mount and interact with a ceph storage
>> ii  python-ceph                 0.72.2-1trusty     amd64              Python 
>> libraries for the Ceph distributed filesystem
>> root@bm4202:/etc/ceph# ceph --version
>> ceph version 0.72.2 (a913ded2ff138aefb8cb84d347d72164099cfd60)
>>
>> Cheers
>>
>>
>>
>> _______________________________________________
>> ceph-users mailing list
>> ceph-users@lists.ceph.com
>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>>
> 
> 

-- 
Loïc Dachary, Artisan Logiciel Libre

Attachment: signature.asc
Description: OpenPGP digital signature

_______________________________________________
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

Reply via email to