Hi,
Our gluster system is currently made of 4 replicated pairs of servers, holding either 2 or 4 bricks of 4 HDs in RAID 10. We have a bunch of clients which are mounting the system through the native fuse glusterfs client, more specifically there are all using the same server1 to get the config of the volume with a backupvolfile-server=server2, i.e. in /etc/fstab as: *server1*:/home /mnt/server glusterfs defaults,_netdev,use-readdirp=no,direct-io-mode=disable,backupvolfile-server=*server2*,log-level=ERROR,log-file=/var/log/gluster.log 0 0 Would it be better if my clients were using alternatively server1 to server8 in terms of network/disk loads ? Many thanks, Alessandro.
_______________________________________________ Gluster-users mailing list [email protected] http://www.gluster.org/mailman/listinfo/gluster-users
