Tableau Resource Monitoring Tool Performance Charts

The Tableau Resource Monitoring Tool includes dashboards that help you monitor and analyze various performance metrics like hardware resource usage, user activity, and the status of Tableau Server processes. They can be useful in identifying performance bottlenecks and the overall health of your Tableau Server.

This topic describes the information on the charts on the Performance page. The performance page provides a dashboard view of the overall health of the Tableau Server and includes the following tabs:

  1. Environment: A dashboard of the performance metrics specific to the environment you selected.
  2. Servers: Focuses on the hardware resource metrics.
  3. Insights: Highlights the slowest views and the longest extract refreshes.
  4. Status: Tableau Server processes status. Information about each process on the nodes whether they are active, busy, or down. This is similar to what you might see on the Tableau Server Settings page.

Environment Tab

The charts on the environment tab give you an all-up view of the health of the Tableau Server.

  • The left side of the page has resource-related metrics: Performance, Tableau Processes, and Background Tasks.
  • The right side charts focus on the user activity and the impact: Concurrent Users, Slow View Load Requests, and Total View Load Requests.

Navigation Tips:
At the top left corner of the page, you can select a time line that is applied to all the charts on this page. You can also select a time range by selecting a part of a specific chart (range selection).
- When you make a range selection in the hardware performance and background task charts, the time line acts as a filter for all the charts on this tab.
- When you make a range selection in the slow view load, and user activity charts, you will be automatically be navigated to the Activity page which provides more details for the selected time range.

Performance Chart

The information on this chart shows you the overall health and usage of the hardware resources for each Tableau Server node. The hardware resources included in this chart are CPU, memory, disk queue, and network.

Note: Network information is not available and not currently supported for Tableau Server running on Linux.

Tableau Processes Chart

Use this chart to get more details about a specific node and the processes running on that node. Starting in 2021.4, almost all Tableau Server processes are tracked with a few exceptions like cluster controller and process that are configured external to Tableau Server cluster, like the External Repository, External File Store, and Independent Gateway. This chart allows you to identify the processes that are driving CPU or memory utilization on a particular node. The chart is dynamically updated to highlight the top ten processes that are using the most resources on a node for the selected time period.

Note: Some Tableau Server sub-processes like tabprotosrv, postgres, and gateway, are not capturedby RMT Server and can cause apparent discrepancies between the Performance chart and the sum of all the values from the Tableau Processes chart.

Background Tasks Chart

The chart is an overview of the volume of background tasks, categorized by the task type. Use the toggle button to see the total number of tasks, the median duration of those tasks, and a histogram that shows you the run-time for the selected time period. This can be a quick way to drill down to see which tasks are taking a long time to run.

Concurrent Users Chart

This chart shows you the number of users that have been sending requests in the time period that is selected.

Note: When you select a portion of this chart, it automatically takes you to the related activity page and shows you the information for that specific time period.

Slow View Load Requests Chart

This chart uses a baseline that is established for each view and is then uses it to compare the time it takes to render that view to determine if the view is taking longer than expected.

In version 2021.4, the baseline is established by calculating the median value of the first 10 times a specific workbook is rendered successfully.

In version 2021.4.1 and later, the baseline is established by calculating the 95th percentile of 50 times a specific workbook is rendered successfully.

Once the baseline is established, every time that same workbook is rendered in the future, the time taken for a workbook to load is then compared to its baseline. Depending on whether the time taken to render the view falls within the expected range or at varying degrees outside of the expected range, they are categorized as follows:

  • Normal: <=2x baseline
  • Long: >=2x baseline
  • Very Long: >=4x baseline
  • Failed: Failed to load or resulted in an error

This baseline and the comparisons only apply to initial rendering of the view. It does apply to subsequent actions like filter selections. Also, when you publish a new version of the workbook, it triggers a recalculation of the baseline.

Note: When you select a portion of this chart, it automatically takes you to the related activity page and shows you the information for that specific time period.

For a detailed walk through on how to use this chart and investigate performance with view rendering, see Investigating Slow View Load Requests.

Total View Load Requests Chart

This chart gives you an overall feel for how many views have been rendered over a period of time. This chart is useful to assess the scope of the impact of a particular incident.

Note: When you select a portion of this chart, it automatically takes you to the related activity page and shows you the information for that specific time period.

Server Tab

The charts on this tab provide a more detailed view of the Performance chart on the Environment tab.

Insights Tab

Slowest Views

This chart shows the slowest views ordered by average view load duration. The orange dot indicator represents that view’s longest load duration. The time period that is represented here is the full range of data that is available, default is 2 weeks.

Longest Extract Refreshes

This chart shows the slowest Refresh Extract tasks ordered by duration. Both data source extracts and workbook extracts are considered. They can be distinguished by the icon next to their name. The time period that is represented here is the full range of data that is available, default is 2 weeks.

Status Tab

This tab lists the Tableau Server processes and their status on each node of the Tableau Server Cluster. This does not include Tableau Server processes that are configured external to Tableau Server cluster like External Repository, External File Store, and Independent Gateway.

Who can do this

Any Resource Monitoring Tool user can view the charts.

Thanks for your feedback!Your feedback has been successfully submitted. Thank you!