Re: ceph-deploy and journal on separate disk

Alfredo Deza <alfredo.deza@xxxxxxxxxxx> · Wed, 14 Aug 2013 10:41:23 -0400

On Wed, Aug 14, 2013 at 7:41 AM, Pavel Timoschenkov <Pavel@xxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote:

>>>It looks like at some point the filesystem is not passed to the options. Would >>>you mind running the `ceph-disk-prepare` command again but with
>>>the --verbose flag?
>>>I think that from the output above (correct it if I am mistaken) that would be >>>something like:

>>>ceph-disk-prepare --verbose -- /dev/sdaa /dev/sda1

Hi. 
If I’m running:
ceph-deploy disk zap ceph001:sdaa ceph001:sda1
and

ceph-disk -v prepare /dev/sdaa /dev/sda1, get the same errors:
======================================================
root@ceph001:~# ceph-disk -v prepare /dev/sdaa /dev/sda1
DEBUG:ceph-disk:Journal /dev/sda1 is a partition
WARNING:ceph-disk:OSD will not be hot-swappable if journal is not the same device as the osd data
DEBUG:ceph-disk:Creating osd partition on /dev/sdaa
Information: Moved requested sector from 34 to 2048 in

order to align on 2048-sector boundaries.
The operation has completed successfully.
DEBUG:ceph-disk:Creating xfs fs on /dev/sdaa1
meta-data="" isize=2048   agcount=32, agsize=22892700 blks

         =                       sectsz=512   attr=2, projid32bit=0
data     =                       bsize=4096   blocks=732566385, imaxpct=5
         =                       sunit=0      swidth=0 blks
naming   =version 2              bsize=4096   ascii-ci=0
log      =internal log           bsize=4096   blocks=357698, version=2
         =                       sectsz=512   sunit=0 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0
DEBUG:ceph-disk:Mounting /dev/sdaa1 on /var/lib/ceph/tmp/mnt.EGTIq2 with options noatime
mount: /dev/sdaa1: more filesystems detected. This should not happen,

       use -t <type> to explicitly specify the filesystem type or
       use wipefs(8) to clean up the device.

mount: you must specify the filesystem type
ceph-disk: Mounting filesystem failed: Command '['mount', '-o', 'noatime', '--', '/dev/sdaa1', '/var/lib/ceph/tmp/mnt.EGTIq2']' returned non-zero exit status 32

If executed this command separately for both disks - looks like ok:

For sdaa:

root@ceph001:~# ceph-disk -v prepare /dev/sdaa
INFO:ceph-disk:Will colocate journal with data on /dev/sdaa
DEBUG:ceph-disk:Creating journal partition num 2 size 1024 on /dev/sdaa
Information: Moved requested sector from 34 to 2048 in

order to align on 2048-sector boundaries.
The operation has completed successfully.
DEBUG:ceph-disk:Journal is GPT partition /dev/disk/by-partuuid/d1389210-6e02-4460-9cb2-0e31e4b0924f

DEBUG:ceph-disk:Creating osd partition on /dev/sdaa
Information: Moved requested sector from 2097153 to 2099200 in

order to align on 2048-sector boundaries.
The operation has completed successfully.
DEBUG:ceph-disk:Creating xfs fs on /dev/sdaa1
meta-data="" isize=2048   agcount=32, agsize=22884508 blks

         =                       sectsz=512   attr=2, projid32bit=0
data     =                       bsize=4096   blocks=732304241, imaxpct=5

         =                       sunit=0      swidth=0 blks
naming   =version 2              bsize=4096   ascii-ci=0
log      =internal log           bsize=4096   blocks=357570, version=2

         =                       sectsz=512   sunit=0 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0
DEBUG:ceph-disk:Mounting /dev/sdaa1 on /var/lib/ceph/tmp/mnt.K3q9v5 with options noatime

DEBUG:ceph-disk:Preparing osd data dir /var/lib/ceph/tmp/mnt.K3q9v5
DEBUG:ceph-disk:Creating symlink /var/lib/ceph/tmp/mnt.K3q9v5/journal -> /dev/disk/by-partuuid/d1389210-6e02-4460-9cb2-0e31e4b0924f

DEBUG:ceph-disk:Unmounting /var/lib/ceph/tmp/mnt.K3q9v5
The operation has completed successfully.
DEBUG:ceph-disk:Calling partprobe on prepared device /dev/sdaa

For sda1:

root@ceph001:~# ceph-disk -v prepare /dev/sda1
DEBUG:ceph-disk:OSD data device /dev/sda1 is a partition
DEBUG:ceph-disk:Creating xfs fs on /dev/sda1
meta-data="" isize=2048   agcount=4, agsize=655360 blks

         =                       sectsz=512   attr=2, projid32bit=0
data     =                       bsize=4096   blocks=2621440, imaxpct=25

         =                       sunit=0      swidth=0 blks
naming   =version 2              bsize=4096   ascii-ci=0
log      =internal log           bsize=4096   blocks=2560, version=2

         =                       sectsz=512   sunit=0 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0
DEBUG:ceph-disk:Mounting /dev/sda1 on /var/lib/ceph/tmp/mnt.G30zPD with options noatime

DEBUG:ceph-disk:Preparing osd data dir /var/lib/ceph/tmp/mnt.G30zPD
DEBUG:ceph-disk:Unmounting /var/lib/ceph/tmp/mnt.G30zPD
DEBUG:ceph-disk:Calling partprobe on prepared device /dev/sda1

The separate commands (e.g. `ceph-disk -v prepare /dev/sda1`) works because then the journal is on the same device as the OSD data, so the execution is different to get them to a working state.

I suspect that there are left over partitions in /dev/sdaa that are causing this to fail, I *think* that we could pass the `-t` flag with the filesystem and prevent this.

Just to be sure, could you list all the partitions on /dev/sdaa (if /dev/sdaa is the whole device)?

Something like:

    sudo parted /dev/sdaa print

Or if you prefer any other way that could tell use what are all the partitions in that device.

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