Hide Forgot
I followed the following guide to create a 2 server server-side replication cluster with 1 client. http://www.gluster.com/community/documentation/index.php/High-availability_storage_using_server-side_AFR After starting the glusterfs daemon on both servers and attaching the volume on the client using 'glusterfs -f /etc/glusterfs/glusterfs-client.vol /mnt/myshare', any file I create on the client shows the following error on the client immediately. root@storageclient01:/mnt/share# touch from_client01 touch: closing `from_client01': Bad file descriptor This is the output in the client gluster log: [2010-03-11 16:52:54] N [glusterfsd.c:1396:main] glusterfs: Successfully started [2010-03-11 16:52:54] N [client-protocol.c:6228:client_setvolume_cbk] cluster: Connected to 10.194.221.142:6996, attached to remote volume 'gfs'. [2010-03-11 16:52:54] N [fuse-bridge.c:2942:fuse_init] glusterfs-fuse: FUSE inited with protocol versions: glusterfs 7.13 kernel 7.12 [2010-03-11 16:52:54] N [client-protocol.c:6228:client_setvolume_cbk] cluster: Connected to 10.242.143.223:6996, attached to remote volume 'gfs'. [2010-03-11 16:53:11] W [fuse-bridge.c:1174:fuse_err_cbk] glusterfs-fuse: 15: FLUSH() ERR => -1 (Bad file descriptor) The file is replicated to both servers and I don't see any error in the server logs. The problem seems to go away if I restart the glusterfs daemon on both servers. I have done this twice from scratch and had the same results. This is on all ec2 ubuntu 9.10 32bit instances using the latest official amazon ubuntu ami.
Avati, can you see if these bugs are valid anymore? if not valid.. please close.
Server side replication is not supported. If you intend to use server side replication, please use gnfs in 3.1.0.