lukas
2015-May-21 16:56 UTC
[Gluster-users] replicated volume - blocked access after on brick fails
Hi, I have a replicated gluster volume consisting of two bricks Volume Name: gvolrep Type: Replicate Volume ID: 8c5dfd6e-775d-44fd-9c72-c37314428c0d Status: Started Number of Bricks: 1 x 2 = 2 Transport-type: tcp Bricks: Brick1: vm-1:/srv/brick/gluster_node1 Brick2: vm-2:/srv/brick/gluster_node2 The problem is that after I turn off , let's say node2, glusterfs is blocking access for almost 1min on node1. Is this to be expected and why? Thank you in advance!
Joe Julian
2015-May-21 17:16 UTC
[Gluster-users] replicated volume - blocked access after on brick fails
The reason for the long (42 second) ping-timeout is because re-establishing fd's and locks can be a very expensive operation. Allowing a longer time to reestablish connections is logical, unless you have servers that frequently die. On 05/21/2015 09:56 AM, lukas wrote:> Hi, > > I have a replicated gluster volume consisting of two bricks > Volume Name: gvolrep > Type: Replicate > Volume ID: 8c5dfd6e-775d-44fd-9c72-c37314428c0d > Status: Started > Number of Bricks: 1 x 2 = 2 > Transport-type: tcp > Bricks: > Brick1: vm-1:/srv/brick/gluster_node1 > Brick2: vm-2:/srv/brick/gluster_node2 > > The problem is that after I turn off , let's say node2, glusterfs is > blocking access for almost 1min on node1. Is this to be expected and why? > > Thank you in advance! > _______________________________________________ > Gluster-users mailing list > Gluster-users at gluster.org > http://www.gluster.org/mailman/listinfo/gluster-users