Graphed LogoGraphed
Graphed LogoGraphed
How It WorksAgentsIntegrationsPricing

How to See Backend Query in Tableau Server

Cody Schneider••8 min read

Ever had a Tableau dashboard that takes forever to load a simple filter, leaving your users frustrated? When performance grinds to a halt, the answer often lies in the SQL queries Tableau Server is sending to your database. This guide will walk you through the practical methods you can use to see those exact backend queries, so you can diagnose bottlenecks, validate your data, and optimize your dashboards like a pro.

GraphedGraphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Watch Graphed demo video

Why Would You Want to See a Tableau Query?

Before jumping into the "how," it's helpful to understand the "why." Peeking behind the curtain at a dashboard’s SQL is not just for developers, it's a critical skill for anyone serious about building efficient, trustworthy analytics. Most people do it for one of a few key reasons:

  • Troubleshooting Slow Dashboards: This is the most common motivation. When a view or filter is slow, the generated SQL is the prime suspect. An inefficient join, a poorly written calculation, or a massive subquery can turn a simple action into a major performance drag.
  • Validating Data Accuracy: If a KPI on your dashboard doesn't look right, examining the query helps you confirm whether Tableau is pulling the correct data and applying the right logic (filters, aggregations, etc.). You can run the query directly against your database to see the raw results.
  • Learning and Understanding: Seeing how Tableau translates your drag-and-drop actions into complex SQL is a great way to deepen your understanding of the tool and database interactions. It reveals how calculated fields, table calculations, and level-of-detail expressions are processed.
  • Optimizing Database Performance: By identifying frequently run, heavy queries, you can work to optimize them on the database side - perhaps by adding an index, creating a materialized view, or refining the table structure.

Method 1: Using the Performance Recorder

The Performance Recorder is your best first stop. It's built right into Tableau Desktop and provides a detailed breakdown of every event that happens when you interact with a workbook, from executing queries to rendering visuals. While it’s primarily a Desktop feature, you can use it to analyze performance against a published data source on Tableau Server.

This is the ideal method for dashboard authors and developers trying to optimize a specific workbook before or during its deployment on Server.

Step-by-Step Instructions:

  1. Start Recording: Open your workbook in Tableau Desktop. If you're investigating a dashboard on Server, ensure you connect to the published Tableau Server data source directly from a new workbook. Go to the top menu and select Help > Settings and Performance > Start Performance Recording.
  2. Reproduce the Action: Now, perform the specific action you want to investigate. This could be loading the dashboard, applying a filter, clicking on a chart, or switching between tabs. Let the action complete fully.
  3. Stop Recording: Once the slow action is finished, go back to the menu and select Help > Settings and Performance > Stop Performance Recording.
  4. Analyze the Results: A new Tableau workbook will open automatically, displaying a Gantt chart of the events that just occurred. The "Timeline" view shows each event, its duration, and its execution order.
  5. Find the Query: Look for events colored in green, labeled "Executing Query." Click on one of these green bars. In the details table below the timeline, you will see the exact SQL query that Tableau sent to your database. You can copy this query directly from a cell in that 'Details' table to analyze it further.

This approach gives you a targeted view of the queries generated from your direct interactions, making it perfect for pinpointing which specific filter or chart is causing the slowdown.

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Method 2: Inspecting the Tableau Server Repository

For a server-wide, historical view of performance, you need to go deeper. Tableau Server uses an internal PostgreSQL database, called the "repository," to store all its metadata - information about users, projects, workbooks, data sources, and performance history.

With administrator access, you can query this repository directly to find historical query data, identify the slowest dashboards across your entire server, and see what users are doing. This is the powerhouse method for server admins tasked with monitoring server health and governance.

Step 1: Enable Repository Access

By default, external access to the repository is disabled for security reasons. A Tableau Server Administrator needs to enable it via the Tableau Services Manager (TSM) command line on the server machine.

  1. Open Command Prompt or PowerShell as an administrator on your Tableau Server machine.
  2. Run the following TSM command to enable a read-only user for the repository. Be sure to use a strong password:

tsm data-access repository-access enable --repository-username readonly --repository-password YourStrongPasswordHere

  1. After the command runs, apply the pending changes:

tsm pending-changes apply

This command will briefly restart Tableau Server, so it's best to do this during a maintenance window.

Step 2: Connect to the PostgreSQL Database

With access enabled, you can now connect to the repository using any standard SQL client that supports PostgreSQL, like DBeaver, pgAdmin, or even Tableau Desktop itself.

  • Host: Your Tableau Server URL or IP address
  • Port: 8060
  • Database: workgroup
  • Username: readonly (or the username you set)
  • Password: The password you set

You may also need to configure SSL, depending on your server's security settings.

Step 3: Query the Performance Tables

The repository contains dozens of tables, but for performance analysis, you'll want to focus on a few key ones:

  • public.historical_events: Records every event that happens on the server, like view loads, data source queries, etc.
  • public.historical_event_types: A lookup table to decode the names of the events in historical_events.
  • public.hist_workbooks, public.hist_views, public.hist_users: Contain historical information about workbooks, views, and users, allowing you to connect event IDs to human-readable names.

Here’s an example query you could run to find the top 10 longest "Execute Query" events in the last 7 days, along with the view and workbook they belong to:

SELECT
  he.created_at,
  hv.name AS view_name,
  hw.name AS workbook_name,
  he.duration_in_ms,
  he.details  -- The details column often contains the full query
FROM historical_events AS he
JOIN historical_event_types AS het ON he.historical_event_type_id = het.type_id
JOIN hist_views AS hv ON he.hist_view_id = hv.id
JOIN hist_workbooks AS hw ON he.hist_workbook_id = hw.id
WHERE het.name = 'Execute Query'
AND he.created_at >= 'now'::date - '7 days'::interval
ORDER BY he.duration_in_ms DESC
LIMIT 10,

The answer and the full SQL query text are often found in the details column of the historical_events table for this event type. Be warned, parsing this text can sometimes be tricky as it might be in JSON format or contain escaped characters, but it's all there for you to find.

GraphedGraphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Watch Graphed demo video

Method 3: Searching Through Tableau Server Log Files

If the above methods don’t yield the answers you need, or if you need to debug a problem in real-time, the final place to look is the server log files. This is the most granular, technical, and often messiest method, but it captures everything.

The relevant logs are typically generated by the VizQL Server process (vizqlserver), which is responsible for turning dashboard interactions into database queries.

Steps to Analyze Logs:

  1. Collect the Logs: The easiest way to get the logs is to generate a log archive. As a server admin, run the following TSM command:

tsm maintenance ziplogs -f my-logs.zip

This collects all log files from all nodes into a single zip archive. You can then download my-logs.zip and extract it to your local machine for analysis without slowing down the server.

  1. Locate the VizQL Log Files: After unzipping, navigate to the vizqlserver folder. You will find files named something like vizqlserver_*.log. These are the files you need to search.
  2. Search for begin-query and end-query tags: Open one of the log files in a text editor that can handle large files (like Sublime Text, VS Code, or Notepad++). Tableau often wraps the full SQL query between clear XML-style begin-query and end-query tags in the log. Use your editor's search function to find these queries:
--- begin-query
SELECT "Products"."Category" AS "Category",
SUM("Sales"."Sales") AS "sum_Sales"
FROM "dbo"."sales_and_products_table"
GROUP BY "Products"."Category"
--- end-query

The log will also contain other contextual events to see the details around the query, performance statistics, and performance issues that occurred. Searching through the logs is often difficult, verbose, and challenging but offers the maximum detail to determine the performance bottlenecks.

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Choosing the Right Method for You

Not sure which path to take? Here’s a quick-reference guide:

  • For Dashboard Creators/Developers: Start with the Performance Recorder. It’s intuitive, fast, and gives you direct feedback on building more efficient workbooks.
  • For Server Administrators: Your home base is working within the Tableau Server Repository. Using direct queries enables you to perform trend analysis and identify server-wide analytics performance opportunities before users experience issues.
  • For Deep, Technical Troubleshooting: Search in the Tableau Server logs. This powerful debugging resource is ideal for system-level issues that affect overall server performance.

Final Thoughts

In this guide, you now know three powerful methods for examining backend queries in Tableau. From using the Performance Recorder to querying the Tableau Server Repository and analyzing server log files, these methods provide comprehensive ways to troubleshoot and optimize your Tableau dashboards.

Understanding and using these tools require some technical knowledge, but they empower your analytics team to be proactive and efficient in maintaining high-performance dashboards. By enabling your team to work directly with data, you can make faster decisions and maintain a competitive edge. Incorporating solutions like Graphed allows marketers to connect data in a single system easily and build professional real-time analyses. This integration supports automated reporting processes and ensures your analytics are always current, enabling your team to focus on generating insights rather than fixing performance issues.

Related Articles