On Thu, May 1, 2008 at 3:36 PM, Brandon Lamb <brandonlamb@xxxxxxxxx> wrote: > http://www.gluster.org/docs/index.php/Setting_up_AFR_on_two_servers_with_server_side_replication > > Most importantly please check out the "Breaking things" section I > added. I will have to go look through the mailing list for > conversations regarding this. > > There is no way around losing the cluster by doing server side afr is > there? If you lose a server, the whole thing is down. > > Does this change if you have more than 2 servers? So far my testing > and instructions have only been dealing with two servers and one or > more clients. http://www.gluster.org/docs/index.php/Setting_up_AFR_on_two_servers_with_server_side_replication Hm ok this is kind of strange actually. So in the example, we have 2 servers doing server side afr to each other. Then we have two clients, client1 connects to server1 and client2 connects to server2. Now if you kill glusterfsd on server2, client2 no longer works, I would expect this. But why does client1 no longer work? If you do afr client side and a server goes down, you can continue working because the client still has access to the other server, why does this not hold true when doing server side afr? Shouldnt client1 that is connected to server1 still be working? server1 should still have access to the volume since it has a local copy even though server2 is down shouldnt it? What is going on here that is so different? client2 cant talk to server2 because its down, ok. client1 can talk to server1 because it is up, but it doesnt work because server1 cant talk to server2? phewwww!
