sometimes the data been indexed doesn't contain all the information needed for visualisation, we'll need to add in additional information, which mostly are some configurations resides in database/file, to help us to create more meaningful charts/dashboards for users. For example, for the server performance data collected through ELK, most likely we would want to add in server vendor, model, physical location, batch, owner, purpose etc, so that when we create visualisation, we can do it, e.g. by purpose, or by location, or by batch etc.. Is there a way I can link these reference data to server performance data?
Not natively.
The best option is to index it with the data, or build something that can pull in that info from other indices.
thanks for your reply...seems like not so straight forward....
Is there a place where we can see what features elastic is going to add to the product? And is there a way we can enter a feature request? I mean a suggestion 
You can discuss for feature requests here and then if confirmed you can add it to elasticsearch github repository where you will also find pending feature requests.
I doubt that this specific feature request could be added in elasticsearch itself.
I'd suggest that you use the new Ingest feature coming in 5.0 to implement your enrichment logic if you wish.
So basically write your own ingest processor in Java as a plugin.