<div dir="ltr">Hello.                                                                                                                  <br>I have installation with 2 servers and one volume with type &quot;Replicate&quot;.                                                <br>Volume mounted on these 2 server too.                                                                                   <br>If we turn off one server, another will be work and mounted volume will be use without problem.                         <br>But if we rebooted our another server, when first was turned off (or gluster was stopped on this server), our volume cann&#39;t mount (glusterd started).<br>We will be get error &quot;Mount failed. Please check the log file for more details.&quot;                                        <br>                                                                                                                        <br>In log for this mount point we will be see:                                                                             <br>[2015-06-17 12:13:53.207698] E [client-handshake.c:1496:client_query_portmap_cbk] 0-repofiles-client-0: failed to get the port number for remote subvolume. Please run &#39;gluster volume status&#39; on server to see if brick process is running.<br>[2015-06-17 12:13:53.207760] I [client.c:2215:client_rpc_notify] 0-repofiles-client-0: disconnected from repofiles-client-0. Client process will keep trying to connect to glusterd until brick&#39;s port is available<br>[2015-06-17 12:14:00.213279] E [socket.c:2267:socket_connect_finish] 0-repofiles-client-1: connection to xxx.xxx.xxx.xxx:24007 failed (Connection timed out)<br>[2015-06-17 12:14:00.213336] E [MSGID: 108006] [afr-common.c:3591:afr_notify] 0-repofiles-replicate-0: All subvolumes are down. Going offline until atleast one of them comes back up.<br>[2015-06-17 12:14:00.217300] I [fuse-bridge.c:5080:fuse_graph_setup] 0-fuse: switched to graph 0                        <br>[2015-06-17 12:14:00.217518] I [fuse-bridge.c:4009:fuse_init] 0-glusterfs-fuse: FUSE inited with protocol versions: glusterfs 7.22 kernel 7.22<br>[2015-06-17 12:14:00.217620] I [afr-common.c:3722:afr_local_init] 0-repofiles-replicate-0: no subvolumes up             <br>[2015-06-17 12:14:00.217746] I [afr-common.c:3722:afr_local_init] 0-repofiles-replicate-0: no subvolumes up             <br>[2015-06-17 12:14:00.217771] W [fuse-bridge.c:779:fuse_attr_cbk] 0-glusterfs-fuse: 2: LOOKUP() / =&gt; -1 (Transport endpoint is not connected)<br>[2015-06-17 12:14:00.219877] I [fuse-bridge.c:4921:fuse_thread_proc] 0-fuse: unmounting /repo                           <br>[2015-06-17 12:14:00.220171] W [glusterfsd.c:1194:cleanup_and_exit] (--&gt; 0-: received signum (15), shutting down        <br>[2015-06-17 12:14:00.220189] I [fuse-bridge.c:5599:fini] 0-fuse: Unmounting &#39;/repo&#39;.                                    <br>                                                                                                                        <br>In /etc/fstab mount has written as:                                                                                     <br>127.0.0.1:/files                            /repo           glusterfs       rw,_netdev              0 0                 <br>                                                                                                                        <br>I found recomendation:                                                                                                  <br>1. gluster peer probe on localhost by his hostname (or ip). But I get &quot;peer probe: success. Probe on localhost not needed&quot; and problem preserved.<br>2. mount by file like                                                                                                   <br>mount -t glusterfs /var/lib/glusterd/vols/files/trusted-files.tcp-fuse.vol /repo                                        <br>or                                                                                                                      <br>mount -t glusterfs /var/lib/glusterd/vols/files/files.tcp-fuse.vol /repo                                                <br>But it doesn&#39;t work too.                                                                                                <br>                                                                                                                        <br>Version of gluster is 3.6.2-3.                                                                                                                                                                               <br>                                                                                                                        <br>How I can get configuration file for mount share/volume from localhost if other nodes in replica set have down?    <br></div>