Hi, has anyone tried an afr using one sparc and one x86 node? I am using two x86 systems an one old sun ultra 5 for client side replicating a single volume over tcp-ip. the files replicate perfectly. now the problem: when i stop all gluster-server of an architecture, i cant "ls" or "find" any files anymore in the gluster-mountpoint. removing, creating and so on works and replicates fine. any help would be nice Io Noci ultra5:/srv/gluster-afr# ls -l total 2964 -rw-r--r-- 1 root root 1468016 2009-03-01 18:22 file1 -rw-r--r-- 1 root root 1548272 2009-03-01 18:22 file2 ultra5:/srv/gluster-afr# stat file1 File: `file1' Size: 1468016 Blocks: 2888 IO Block: 8192 regular file Device: 12h/18d Inode: 504931 Links: 1 Access: (0644/-rw-r--r--) Uid: ( 0/ root) Gid: ( 0/ root) Access: 2009-03-01 18:22:33.000000000 +0100 Modify: 2009-03-01 18:22:34.000000000 +0100 Change: 2009-03-01 18:22:47.000000000 +0100 ultra5:/srv/gluster-afr# /etc/init.d/glusterfs-server stop Stopping glusterfs server: glusterfsd. ultra5:/srv/gluster-afr# ls -l ls: reading directory .: Input/output error total 0 ultra5:/srv/gluster-afr# stat file1 File: `file1' Size: 1468016 Blocks: 2888 IO Block: 8192 regular file Device: 12h/18d Inode: 504931 Links: 1 Access: (0644/-rw-r--r--) Uid: ( 0/ root) Gid: ( 0/ root) Access: 2009-03-01 18:22:33.000000000 +0100 Modify: 2009-03-01 18:22:34.000000000 +0100 Change: 2009-03-01 18:22:47.000000000 +0100 ultra5:/srv/gluster-afr# rm file1 ultra5:/srv/gluster-afr# ls - ls: reading directory .: Input/output error ultra5:/srv/gluster-afr# /etc/init.d/glusterfs-server start Starting glusterfs server: glusterfsd. ultra5:/srv/gluster-afr# ls -l total 1520 -rw-r--r-- 1 root root 1548272 2009-03-01 18:22 file2 ultra5:/srv/gluster-afr# my config is: *server: posix-> posix-locks-> server *client: three client brick -> afr 3 client bricks(option read-subvolume localbrick) debian 5.0 glusterfs 1.3.12 (trying 2.0.0rc2 later) fuse-2.7.4-1.1 kernel 2.6.26
