How to setup cassandra cluster with four nodes.

[sancho sebastine <sa...@gmail.com>]

Hi,

I am having trouble in setting up cassandra cluster with multiple nodes. It
would be great if you an shed some light on what I am doing wrong.  Here are
the steps that I did.

1) Setup first node cassandra1(ip: 10.224.54.1) with the following
properties
auto_bootstrap: false
listen_address: 10.224.54.1
rpc_address: 10.224.54.1
seeds:
    - 10.224.54.1
//This node starts up great and i can add data and retreive from it.

2) Setup the second node cassandra2(ip: 10.224.54.2) with the following
properties
auto_bootstrap: true
listen_address: 10.224.54.2
rpc_address: 10.224.54.2
seeds:
    - 10.224.54.1

//The node starts up fine and migrates data. But the problem is second node
cassandra2 takes up 100% of the data. And the nodetool does not recognize
two nodes, it just says the cluster has only one node which is the last one
i added.

ubuntu@domU-12-31-38-00-E0-12:~$ cassandra-0.7/bin/nodetool
-h 10.224.54.1 ring
Address         Status State   Load            Owns    Token

1010.224.54.2  Up     Normal  540.63 KB       100.00%
95245447804760228247104454664152650725

3) When is setup the third node the same thing happens. All the data is
migrated to the third node and nodetool ring shows only the third node.

Please let me know what am i doing wrong. ?
Should i be setting auto_bootstrap=false in the second node once it has
started up. ?
Should i be setting the third node's seed as the second node or the
first(The documentation says i need to set it as the first)?
And finally, What is the difference between a seed node and a non seed node?
Does seed node store data as well?

I really appreciate any help, as i have followed all the online tutorials
that i could find, but still couldn't get the cluster started up.

Thanks
Sancho Sebastine