[sheepdog-users] node recovery performance

[Liu Yuan]

On Tue, Nov 26, 2013 at 02:20:52PM +0100, Valerio Pachera wrote:
> 2013/11/26 Gerald Richter - ECOS <richter at ecos.de>
> 
> > [GR] The first server starts recovery. When now the second server reboots,
> > both server are trying to make a recovery from each other. What is
> > happening inthis case? Which data will be used (assuming that some data
> > might be modified)?

Latest data in either nodes will be used.

> What is the output of 'dog node recovery'?
> 
> Call the nodes A and B.
> If you turn off B and later A, you have to restart node A first.
> In this case, I think you should only see node B by 'dog node recovery'.
> 
> If you turn on B, sheep is probably not going to allow you to join the
> cluster with node A because of the epoch date that is more recent.

This was true for prevoius release. after v0.7 (inclusive), there is no order
imposed to restart the cluster. You can restart the nodes in any sequence and
sheepdog will make sure all the data in all the nodes are latest and consistent.

Thanks
Yuan