Recommended Initial Cluster Configuration

I need to create a cluster to support approximately 3 million 12KBish docs for a total of about 32GB worth of data. We are a small development team of about 10-15 people and won't be executing a lot of concurrent queries, but we will be doing aggregations, full-text searching, etc on the data and want the queries to be performant. Can someone recommend a good starting point as far as the number of nodes, amount of RAM, required disk space, etc to start with? Thanks.

Kevin

If this is production, then 3 nodes will do.
Hardware wise, depends on your budget.

@warkolm Thank you very much for the response and information!