[Gluster-users] Gluster - data migration.

Kushnir, Michael (NIH/NLM/LHC) [C] michael.kushnir at nih.gov
Thu Feb 14 03:38:16 UTC 2013

Hi Rafal,

I just ran into a similar situation. 

You can leave slave bricks empty and create a volume. As long as, in each replicated set of bricks, each brick with data is on the left side of the pair, gluster will replicate your data. 


gluster volume create my-volume replica 2 transport tcp master:/brick1 slave:/brick1 master:/brick2 slave:/brick2

In my case, that replicated the data from the server with data to the server without data. 



From: Rafał Radecki [radecki.rafal at gmail.com]
Sent: Thursday, February 07, 2013 8:57 AM
To: gluster-users at gluster.org
Subject: [Gluster-users] Gluster - data migration.

Hi All.

I have two servers ("master" and "slave") with a replicated gluster
volume. Recently I've had a problem with slave and gluster does not
work on it now.
So I would like to:
- stop and remove current volume on master (on slave it is not accessible);
- stop gluster software on master (already stopped on slave);
- remove gluster software on master and slave (previous administrator
used own built rpms, I would like to use
- clean old information:
setfattr -x trusted.glusterfs.volume-id /gluster
setfattr -x trusted.gfid /gluster
rm -rf /gluster/.glusterfs
- rsync data from master to slave;
- start gluster and create a volume with data in rsynced /gluster directory.
Are there any pitfalls I should know about?

Best regards,
Rafal Radecki.
Gluster-users mailing list
Gluster-users at gluster.org

More information about the Gluster-users mailing list