Two nodes One cup

We are trying to set up a proof of concept with two instances of ES 1.3.4 on the same red hat server and import a json file.
1 cloned the original yml
2 changed the node name left cluster name the same as the first instance.
3 elasticsearch.yml and call it elasticsearch2.yml. Then go change the node name, set node.master:false & node.data:false, change the path.conf: /path/to/conf2 & path.data: /path/to/data2
My question is how to start the 2nd one. The only way to start is the ./elasticsearch correct
The first incantation of our poc failed with some crazy lack of space error when trying to import the json file. And its a huge red hat server with 2tb 500mb of ram.

Any direction would be helpful.