How to create view of hdfs dataset

what type of a view file should be created, so that tableau can query a set of parquet files using that view?"

Like in Apache drill a view file can be created “ABC.view.drill”, which can be easily queried by tableau.

Is there any similar kind of view file for dremio?

Hi,

You can promote a folder that contains the parquet files and create a dataset from that folder. For folders, if you hover over the action column you will see a format folder button.

You then can reference that folder (hdfs.data.dremio in this case) in Tableau.

I used to create drill file programatically, but how we can promote a folder which contains parquet files programatically?
Programatically as in, by use of scala programs.
Basically Drill file used to be like
{
View Name: ABC
Command: Select * from HDFS Location


}

Here in above view, command contains HDFS Location , which is a location where parquet files are present.

The option you suggested will create/export the data in another file format, which will be copying the data from one formate to another,which is memory consuming process.

You can use our REST API (http://docs.dremio.com/rest-api/catalog/post-catalog-id.html) to promote programmatically. For folders of parquet, you should be able to also just run SELECT * FROM hdfs.data.dremio and it will auto-promote for you.

Actually, I need a file which will act as a view. So that using that view file tableau can query the parquet file data which is on HDFS location.
And this needs to be done via dremio. I don’t think tableau is able to query via API.