[sheepdog-users] Testing snapshot-object-reclaim

[Hitoshi Mitake]

At Tue, 25 Feb 2014 08:50:23 +0100,
Valerio Pachera wrote:
> 
> Hi, I've been doing some 'qemu-img convert' tests so I had many new vdi
> 
> dog vdi list
>   Name        Id    Size    Used  Shared    Creation time   VDI id  Copies  Tag
>   testraw      0   10 GB  2.3 GB  0.0 MB 2014-02-25 08:17   12a991
> 2:1
>   wheezy_template     0   10 GB  5.9 GB  312 MB 2014-02-23 20:59
> 5ddf8b      2
>   test         0   10 GB  2.3 GB  0.0 MB 2014-02-25 07:58   7c2b25
> 2:1
>   testraw2c     0   10 GB  2.3 GB  0.0 MB 2014-02-25 08:21   ae12ae
>   2
>   testc2       0   10 GB  2.3 GB  0.0 MB 2014-02-25 08:03   cdab40
>  2
>   test1        0   10 GB  2.3 GB  0.0 MB 2014-02-25 08:00   fd32fc    2:1
> 
> The I removed all of them but 'test1'.
> 
> dog vdi list
>   Name        Id    Size    Used  Shared    Creation time   VDI id  Copies  Tag
>   test         0   10 GB  2.3 GB  0.0 MB 2014-02-25 07:58   7c2b25    2:1
> 
> I would expect to have 2.3GB*2=4.6GB in the cluster used space, but I see 15GB.
> 
> dog node info
> Id      Size    Used    Avail   Use%
>  0      233 GB  8.0 GB  225 GB    3%
>  1      452 GB  7.0 GB  445 GB    1%
> Total   685 GB  15 GB   670 GB    2%
> 
> Total virtual image size        10 GB
> 
> 
> What do you think about that?

Can I see an output of "tree <path of sheepdog store>" on both of the nodes?

Thanks,
Hitoshi