> > Now, one of the nodes dies, and when I bring it back up, it does'nt join > the cluster again, but becomes it own node/cluster. I can't get it to join > the cluster again, even after doing 'removenode' and clearing all data. >
That obviously should not have happened. That being said we have a few know issue in 1.2.0 that we've fixed for 1.2.1 that while they don't match exactly this description might be related. In other words, that is a bug, but it might be worth checking against the current 1.2 branch (or waiting for 1.2.1) to see if you can reproduce. > Then I start to look into the system.peers table and something doesn't > seem right. > > Node A has the other 2 nodes listed, but no tokens. > Node B has the other 2 nodes listed, but only tokens on one of them. > Node C has the other 2 nodes listed, and tokens for both of them. > https://issues.apache.org/jira/browse/CASSANDRA-5121 > Furthermore, after I replace the failed node, it still remains in the > system.peers table, with no tokens. > That might well be a bug we haven't fixed yet. Would you mind opening a ticket for that in jira? Otherwise, and to sum up, I don't think you've done anything wrong but 1.2.0 has some bugs. We're planning on releasing 1.2.1 soonish, which probably means sometimes around the middle of next week. -- Sylvain
