Open Stack

Hi,

could you be more specific what is "not too fast" for you? I don't  
really have too much information from my side, but I can describe how  
we do it.
We use Ceph as back end for all services (nova, glance, cinder), and  
the most important machines are backed up by our backup server  
directly via rbd commands:
  - We create a snapshot of the running instances and export the  
snapshot to an external drive, this is a full backup.
In addition to that many of our machines have their home or working  
directories mounted from CephFS which we also backup as a tar ball to  
an external drive once a week (also full backup).
And then we have also a "real" backup solution in place (bacula) where  
we store incremental as well as full backups from individually  
configured resources for different intervals.
All these different approaches have different runtimes, of course.  
Just as an example, one 40 GB VM (rbd image) which has areound 24 GB  
in-use takes around 6 minutes for the full backup.
Although we also have the cinder-backup service up and running nobody  
is using it because the important volumes are attached to instances  
which we backup by our rbd solution, so there's no real need for that.

Regards,
Eugen

Zitat von Nguyễn Hữu Khôi <nguyenhuukhoinw at gmail.com>:

> Hello guys.
> I am looking for instance backup solution. I am using Cinder backup with
> nfs backup but it looks not too fast. I am using a 10Gbps network. I would
> like to know experience for best practice for instance backup solutions on
> Openstack.
> Thank you.
> Nguyen Huu Khoi