Both are bad. Why -> you have an "even" number of replicas and this can cause split brain situation.
I would recommend you to consider this approach:
Server A - Replica_A1
Server B - Replica_A2,Arbiter_B
Server C - Arbiter_A,Replica_B1
Server D - Replica_B2
As the arbiter doesn't require much resources, you can use it safely.
The negative side of such deployment - you have 2 separate volumes which you have to manually balance the VM's images so both Volumes keep the same level of load.
Still, this is my own opinion and it might be a better solution.
Best Regards,
Strahil Nikolov
В петък, 6 септември 2019 г., 12:54:48 ч. Гринуич+3, Cristian Del Carlo <cristian.delcarlo@xxxxxxxxxxxxxxxxxx> написа:
Hi,
I have an environment consisting of 4 nodes ( with large disks).
I have to create a volume to contain image of virtual machines.
Is it better to set a Distributed Replicated Volumes like this?
Or a Replicated Volumes like this?
# gluster volume create test-volume replica 4 transport tcp server1:/exp1 server2:/exp2 server3:/exp3 server4:/exp4
Thanks,
--
Cristian Del Carlo
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
https://lists.gluster.org/mailman/listinfo/gluster-users
_______________________________________________ Gluster-users mailing list Gluster-users@xxxxxxxxxxx https://lists.gluster.org/mailman/listinfo/gluster-users
