[Gluster-devel] AFR translator

Gerry Reno greno at verizon.net
Tue Feb 27 21:39:27 UTC 2007

Trying to understand how the AFR translator could function:
In a scenario where I have storage configured on every cluster node and 
I use AFR to mirror all the nodes so they are all identical.  So I guess 
in this case each node would be both server and client.  If a node were 
to suffer a hardware failure on it's local storage device, would that 
client node then still remain up but just pull it's file/data access 
across the network from the other nodes?  Or would it drop out of the 


More information about the Gluster-devel mailing list