Document how to install multiple graphryder-api instances on one server

[tanius]

Since one Graphryder API instance can only deal with one ethnographic corpus, installing several instances is so far the only way to manage multiple corpora in parallel. For that, we want a way to install these multiple instances on one single server rather than requiring a new server or virtual machine for each.

The only issue with this seems to be how to have the data separated inside the Neo4j database. Should be solvable, though – perhaps as simple as adding a new user.