Tableau
Intermediate 13 min read Tableau

Tableau Server, Cloud and Publishing

The governance plane under the dashboards

Tableau Server/Cloud publishes workbooks with data sources, schedules refreshes, enforces permissions and row-level security, and runs monitoring/usage analytics — the governance plane of Tableau.

Tableau ServerTableau Cloudpublishingrefresh schedulepermissionsRLS
Prerequisites: Tableau Dashboards
Before we start

What you will learn

  • Publish workbooks and data sources with refresh schedules
  • Model permissions and core-based row-level security
  • Operate with monitoring and usage analytics
Introduction

What is this?

Tableau Server (self-hosted) or Tableau Cloud (SaaS) is where workbooks and data sources are published for distribution. Published items carry metadata, refresh schedules (extract data sources), permission rules (site/group/project), and row-level security via user filters. Admins monitor data quality, adoption and performance through maintenance views and usage analytics.

Motivation

Why do you need it?

  • Publishing turns a lone viz into an operational, governed product.
  • A refresh schedule & permissions make the dashboard trustworthy and safe.
  • RLS user filters deliver one published source to many audiences.
  • Operations (monitoring, backup, performance) is the difference between a toy and a platform.
Think about it

Mental model

Imagine

A retail pipeline: warehouse, delivery, checkout

Think of

Server publishing

The workshop (Desktop) makes pieces; the warehouse (Server) stores and ships them; the delivery van (refresh schedule) restocks on time; and the checkout (permissions) decides who's allowed in the aisle. Automated stock checks (monitoring) keep it from running out mid-shift.

Mechanics

How it works

  • Publish from Desktop: choose server/cloud, project, data source + extract, and set permissions.
  • Extract data sources refresh on named schedules; live sources connect through on-premises bridges for Tableau Cloud.
  • Projects and groups define permissions; content inherits from the project unless overridden.
  • User filters map logged-in users to row sets (RLS) for shared data sources.
  • Usage analytics (Tableau 2021+, Cloud included) show who views what, load times and data quality signals.

Publish → refresh → secure → monitor.

Publish
Refresh
Secure
Share
Monitor
Hands-on

Examples

The user-filter RLS portfolio

The user-filter RLS portfolio
User filter: [E-Mail Address] IN
  { FIXED [Region]: 
      LOOKUP(MIN([user-to-region][Region]), 0) }
 
Published as:
  A data source with an extract schedule (every 6h)
  plus a workbook whose audience = the group "Region Managers".
 
Result: one source, per-region rows on refresh, zero per-user extracts.

A user filter restricts each viewer to their mapped region at query time — the published-data-source equivalent of DAX row-level security.

Real world

In production

  • A data team publishes a certified sales data source; 12 teams build sheets on top without engineering each copy.
  • An extract refreshing at 6am powers an executive view that must never show yesterday's stale numbers.
  • Admins run usage analytics to retire the 30 dashboards nobody opens each quarter.
Guidelines

Best practices

  • Publish the data source once, revoke per-workbook re-hosting, keep one version.
  • Schedule refreshes with sensible staggering; fail loudly on missed runs.
  • Model permissions by project/group — the fewest exceptions wins.
  • Audit lineage and usage quarterly; remove unused sources and dashboards.
Watch out

Common mistakes

  • Publishing a fresh workbook with embedded data nobody can refresh.
  • Permission exceptions strewn across workbooks instead of project structure.
  • Live data sources with no bridge on Cloud, silently stale.
  • Skipping monitoring — then a quietly-broken extract breaks the executive prep.
When things break

Debugging tips

  • Refresh failed → open the data-source's refresh log; check credentials and schedule target.
  • User sees no rows → the user filter maps them to nothing; check membership + filter logic.
  • Dashboard lag → inspect the extract size and any live connections at peak.
Interviews

Interview checkpoint

Test yourself

Quick quiz

0/2
1

What repopulates an extract with fresh data on a schedule?

2

User filters let one published data source serve many audiences with different row visibility.

Go deeper

Official documentation

Tableau Server data management
docs
User filters overview
docs