I'm attempting to do a replace-brick as some of you helpfully pointed me at, but I've run into a problem. I have existing nodes server1 and server2 (Ubuntu Lucid 64 running gluster 3.1), and I want to add a new node (server3) and migrate the contents of server1 to it. On server2 I did: gluster probe server3 At the time I did this, the firewall wasn't quite configured right - server3 received the probe request, but couldn't connect back. I reconfigured the fw and that appeared to help, but it's now in an odd state. On server2, peers are now listed thus: Number of Peers: 2 Hostname: server1 Uuid: bc37e1d7-153b-41c3-814a-391cbc70283f State: Peer in Cluster (Connected) Hostname: server3 Uuid: 9b5284f2-e00c-46db-a3d2-defd82cf96d8 State: Probe Sent to Peer (Connected) On server3: Number of Peers: 2 Hostname: server2 Uuid: fc25b480-17b7-4b8c-b447-3287f46e59bc State: Connected to Peer (Connected) Hostname: server1 Uuid: bc37e1d7-153b-41c3-814a-391cbc70283f State: Peer in Cluster (Connected) If I attempt a replace-brick from server2: gluster volume replace-brick sharedcontent server1:/var/sharedcontent server3:/var/sharedcontent start server3, is not befriended at the moment 'befriended' does not appear in the docs or wiki (though 'friend' turns up a lot of wiki spam). What have I missed? How can I fix this? I also noticed that the 3.1 docs say there is no way to remove a node from the peer list! How am I supposed to take down server1 cleanly? Marcus -- Marcus Bointon Synchromedia Limited: Creators of http://www.smartmessages.net/ UK info at hand CRM solutions marcus at synchromedia.co.uk | http://www.synchromedia.co.uk/
