Leveraging standard elasticsearch nodes

[johnament]

In your docs, you write

In case, you are worried, that you need to include the 20MB elasticsearch dependency in your project, you do not need to be.

The problem is that I'm already including this 20 MB dependency in my code. In fact, its better for us this way for a number of reasons:

• AWS Node discovery works well. No need to configure hosts.
• We can run locally (dev mode) with just an embedded node, bigger environments with bigger nodes.

So I have to wonder, can you consider supporting the standard node configuration out of the box, e.g. in your builder add

withClient(Client)

Or even

withSettings(Settings)

To allow a fully configurable connection, without the need to already use ElasticSearch (honestly though, the reason this reporter interests me is because I already use ElasticSearch in my app, so it makes sense)

[spinscale]

As the reporter is using HTTP instead of the transport protocol, I would need to write an additional reporter using the client. You could potential use the guice injector in your own code to get some of the information that you need.