[Gluster-users] How to resynchronize geo-replication slave after crash

Kotresh Hiremath Ravishankar khiremat at redhat.com
Thu Aug 27 07:14:56 UTC 2015


Hi Milos,

Geo-replication expects gfids of files between master and slaves to be same for the data syncing.
Interfering the geo-rep sync by any kind of manual methods that could change the gfids is not
recommended and data for those files will not be synced.

To bring back to the sane state, those file which are manually synced through rsync which has different
gfids must be removed.

Method to Initiate a Full sync depends on version being used.

Which gluster version are you using?

Thanks and Regards,
Kotresh H R

----- Original Message -----
> From: "Milos Cuculovic - MDPI" <cuculovic at mdpi.com>
> To: gluster-users at gluster.org
> Sent: Thursday, August 27, 2015 12:30:55 PM
> Subject: [Gluster-users] How to resynchronize geo-replication slave after	crash
> 
> Hi All,
> 
> I have a geo-replication master - slave.
> The slave had some troubles and the replication was not working for a
> certain time. What I did:
> 
> * Manual rsync to synchronize master and slave
> * Start geo-replication
> 
> It seems the new files are not synchronized correctly.
> Is there a procedure how to do a full sync with GlusterFs and continue
> with replication?
> 
> Thank you.
> 
> Milos Cuculovic
> 
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://www.gluster.org/mailman/listinfo/gluster-users
> 


More information about the Gluster-users mailing list