Best way to configure multinode cassandra cluster deployment

Question

Although I have worked with cassandra at application level, I do not have deployment experience. For my grad project, I need to deploy cassandra to multiple nodes in futuregrid.org just wondering if there are any tools to deploy cassandra just by giving hostnames and few conf options.

Otherwise I have wrote a small deployment script but I am not able to figure out how to create single cassandra.yaml file that can be used to instantiate all instances of cassandra by ssh-ing to those nodes. This page seems to hint that we might need to create separate configurations per node and then instantiate them, but I might be missing something.

Some background regarding my setup:

1. I am using resources from futuregrid.org, you can assume that each node is similar to other, so I strictly do not need separate conf.
2. I am using DSE 4 version.

Please advise. Any pointers on either of these would be great.

Answer 1

After going through some iterations on this in the comments, in most cases you won't need to do node specific customizations of cassandra.yaml. The one I tend to set (because I have multiple interfaces in many of my machines) is listen_address, but as https://stackoverflow.com/users/56076/don-branson points out, if you leave that blank the system will sort that one out for you.

Check http://www.datastax.com/documentation/datastax_enterprise/4.0/datastax_enterprise/deploy/deploySingleDC.html for more current information just in case.