I just saw that the "failed to connect to 'x.x.x.x:7199' : connection refused" error for nodetool that I am facing used to exist earlier also, case in point https://issues.apache.org/jira/browse/CASSANDRA-4494 I can see this problem resurfacing in Cassandra 1.1.9 on my system. I am using RHEL 6.0 and 7199 can be seen bound by Cassandra upon netstat. When I do "telnet x.x.x.x 7199", that works too.
Any pointers please. 2013/3/8 amulya rattan <talk2amu...@gmail.com> > I have started cassandra on three nodes, however opscenter only shows two > of them. The missing one is starting properly, with gossip and everything > else showing no errors. It also has the other nodes as seeds in > cassandra.yaml, so I am amazed why it doesn't show as part of the ring. I > have tried restarting it a bunch of times but neither does it throw any > error, nor it shows as part of the ring > > Another strange thing is that nodetool won't start on any of the nodes. It > keeps throwing connection refused error. It's not the same usual error one > faces that starts with "Error connection to remote JMX agent!". It simply > says connection refused on x.x.x.x:7199. However, I can see the JMX port is > bound by cassandra on 0.0.0.0. I can use cassandra-cli on 9160 and > opscenter on 8888, but not jmx port. As last resort, I turned off firewall > completely, but to no avail. > > I never faced any such errors setting cassandra cluster up in the past, so > this is completely flabbergasting to me given that the port is bound and > firewall is off. > > Anybody faced anything similar in the past? Or anybody knows how to solve > it? All responses appreciated. > > ~Amulya >
