Also, disperse is no where near as powerful as crush mapping which can
be configured to be rack aware. Would be nice to have that level of
configuration eventually.
On 06/27/2016 12:25 PM, Joe Julian wrote:
In large clusters you're deploying them using automation tools. It's
not that hard. That being said, the disperse translator can manage
redundancy over a non-replicated volume which would give you that
feature.
On 06/27/2016 12:18 PM, Gandalf Corvotempesta wrote:
If I remember properly, the brick order when creating a replicated
volume is important, as we have to write bricks in order to preserve
redudancy
Any plans to fix this allowing a 'random' order like ceph? In ceph
there is no need to create volums by setting bricks in proper order.
On large clusters, this could be very tricky.
_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://www.gluster.org/mailman/listinfo/gluster-users
_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://www.gluster.org/mailman/listinfo/gluster-users
_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://www.gluster.org/mailman/listinfo/gluster-users
