There are multiple ways you can publish
.hyper files to Tableau Server, the options you have depend upon whether your file contains a single-table or multiple-tables. Programmatically publishing a
.hyper files, using the REST API or Tableau Server Client library, has special requirements.
In this section
.hyper file has a single table, you have several options for publishing your extract to Tableau Server.
Option 1: Use Tableau Desktop to publish. You can publish the
.hyper file as part of a workbook (embedded data source
.twbx file), or as a separate packaged data source (
.tdsx) file. See Publish a Data Source.
Option 2: Use tabcmd to publish the
.hyper file (without
.tdsx). Use the Tableau REST API or Tableau Server Client Library (Python) to publish the
.hyper file. If you publish the extract this way, the single table needs to be named
Extract and the table must be in the
Extract schema (or namespace). Note that the names are case-sensitive and they must match exactly.
Option 3: (Not Supported) Create a Tableau data source (
.tds) file and package it with the
.hyper file into a single zipped file. Rename the
.zip file as a
.tdsx file. This is the safest way to publish a file, however, we don’t provide an automated way to create
.tds files. The only way to do this automatically involves modifying the XML-based
.tds file. For example, use Tableau Desktop to save the
.hyper file. From the Data menu, select the extract, and then select Add to Saved Data Sources. Then you need to modify the
.tds file (this is not supported). A
.tds file is a Tableau data source (description) file, which is an XML file that describes the connections, the joins, etc., within the
.hyper file has multiple tables, you can publish your extract to Tableau Server using Option 1 or Option 3, as described for single-table Hyper files. However, you can not use Option 2 to programmatically publish the Hyper file with tabcmd, the Tableau REST API, or the Tableau Server Client Library (Python). This is because that option requires that the
.hyper file to have a single table named
Extract that uses the
Extract schema (