Architecture

We have created an image to show the infrastructure of the WriteBackExtreme extension. This shows what actually happens when you call in your WriteBackExtreme extension to your dashboard.

There’s a connection to the WriteBackExtreme extension server where it will serve the necessary JavaScript, HTML and CSS to the dashboard for the extension to be rendered there (the rendering of the extension is client side). Write back transactions are handled by the WriteBackExtreme Extension server through API calls.

A repository database is required (supported database types for the repository database are listed below) and a connection to your database where you can write directly to is probably already present in your infrastructure (Write Back databases, the common data sources you use for your Tableau Reporting as well).

Whitelisting/Safelisting of the Extension Server URL needs to be done on Tableau Server/Tableau Online to allow communication with the extension server for users visiting your published dashboard. Communication with the WriteBackExtreme extension server will only take place from the clients side (whether it is Tableau Desktop or a visitor of a published workbook in the browser). As already mentioned, rendering of the extension will take place in the dashboard (client side).

Where does the write back data live?

The WriteBackExtreme extension can connect to multiple databases by setting up different connections in your Management Console. Use the plugins tab of your management console to see the available database drivers you can connect to. If you miss a specific driver please contact our support.

The writeback data lives in the tables of databases that are (already) present in your infrastructure; the data sources already available to your (Tableau) infrastructure. When creating your (DATA Helper and ADD Data) schemas in the management console, the actual tables will be added to your database automatically by the extension. Changing your schema in the Management console will also change the actual tables in the database. Because of this automation by the extension it is highly recommended not to manually alter the tables created by the extension (and do all changes through the Management console of the extension).

You can also connect to existing tables to change data that already resides in tables in your databases (using the DATA Correction module). The DATA Correction module allows you to change data that is already present in your database and therefore will not create its own table but use the existing table. Be aware that Data correction changes your actual data!

The WriteBackExtreme extension can set up connections to different databases and the data will always remain in the database you select during the creation of the schema.

Where does the data live for repository and which databases are supported?

A database will be set up when installing WriteBackExtreme for repository data. This will be a separate database from your Write Back databases, preferably installed on the same machine as the WriteBackExtreme extension server, this is to avoid lag and performance degradation.

A data repository is the logical aggregation of data items from separate databases into one centralized location for a specific purpose that cannot be achieved using the databases themselves. There is no server side processing of your data, auditing and logging information will be stored in the Repository database. The configuration and connection information of the WriteBackExtreme extension will also be stored (in an encrypted way of course) in the Repository database

Supported databases

	MySQL version 5.7, 8.0 or higher
	Microsoft SQL Server 2016 or later
	PostgreSQL version 14 or later
	Snowflake via ODBC or natively (Editing a schema is partially supported because of database limits)
	Amazon Redshift
	Oracle via OCI8 ^2.2.0 or 3.0
	Azure Synapse (Text fields are limited to 4000 characters)
	Google Big Query Keep in mind that GBQ has Quotas and limits. For example it is limited to 1500 Table modifications per day which makes it less usable for data correction.
	Teradata via ODBC
	Databricks via ODBC