_edited_edited.png)
In the data visualization Tableau is the leading platform which is used by several data analytics. It makes visualization easy to understand for everyone, including nontechnical background people. It provides formulas, fields, calculations and simple visualization tools and methods to provide an intuitive user interface. Many organizations use Tableau to deliver scalable solutions and make better business decisions. It allows a flexible approach to the design and build of dashboards, which is true for smaller datasets, but working with large datasets having millions of records your dashboards take too much time to render. However, many people face a problem when they start applying the more features of Tableau. We are at faster pace world where every second matters, and the delay by some seconds can cause a loss of audience. So, the performance of Tableau matters a lot.
Luckily, we have compiled some of our top tips for keeping a Tableau dashboard quick and snappy when working with large datasets. In this blog, I will share the most common best practice for you to improving performance of your dashboards in Tableau.
Before we get started, it is important to note that not all the best practices will work for the unique problems. It really depends.
Aside from fixes you can perform within Tableau to improve speed, there are further considerations that should be acknowledged while working to improve performance.one of the most important things that matter is your hardware itself. If it is slow computer, Tableau will run slow as well. Another important thing is also your network speed. If data is not on your machine and you are connecting through a server.
Start with a performance recording
the performance recording should be started from an empty workbook. so how can you make your dashboards run faster? Your first step is to find out the problem spots by running the performance recording.it is the indication of where the problems might be coming from. The performance recorder can pinpoint slow worksheets, Slow queries and long render time on dashboards.
To start a performance recording on Tableau Desktop:
Help > Settings and Performance > Start Performance Recording
You can follow following tips to sort out the performance issues.
1. Your data strategy drives performance
The tip is fairly straightforward: the more data Tableau has to go through, the longer it takes. There are few ways of minimizing the amount of data used in Tableau.
· Use the extract connection instead of live connection to your data source.
· Use the data source filter or use an extract connection to the data source.
· Minimize the number of joins and relationships. Minimize the number of fields used in dashboard. To achieve that you need to do “Hide All Unused Fields” options to remove all unused columns from the data source.
· Unused columns should be removed to reduce loading time.
· Remove hierarchies for invisible dimensions.
· Use common metadata for dashboard creation.
· Use correct data types, using incorrect data types can have an impact on performance as some comparisons are quicker than others.
· Roll up by aggregating or filtering to maintain the required data granularity.
2. Reduce the Filters
Filter feature is powerful and expressive feature of Tableau but at the same time excessively use of filters can cause performance issues.
· Use quick filters.
· Include filters are quicker than exclude filters. The exclude filter loads entire domain, while include filter do not.
· Use parameters and action filters. These reduce the query load.
· Use continuous date filter, it can take advantage of the indexing properties of your database.
3. Optimize calculations
Wherever possible you need to perform the calculations in the database to increase the performance speed. . Use aggregate calculations for calculated fields. Also performing low level calculations in the database helps to achieve your goal.
· MIN or MAX functions help quicken the processing time rather than AVG function.
· Another great tip to use more Booleans, Integer and float calculations instead of string calculations. Computers can process integers and Booleans much faster than strings.
· Limit the usage of marks on view, row level calculations, and blending calculations.
· When testing for multiple conditions in a calculated field, using ELSEIF instead of ELSE IF will keep the evaluation in the same statement instead of moving to the nested statement. Nested statement will come with a performance cost.
. Do not reference the same calculated field in another calculation more than necessary. for example, if calculated field is referred three times in the same calculation, then running the calculation will require the calculated field to be calculated three times overall.
. Distinct counting (COUNTD) is one of the slowest aggregations in Tableau. it should be avoided if possible. Using distinct counts is often important to the analysis, so completely avoiding any counts is not feasible. There are some other multiple alternatives to COUNTD that could be used.
4. Custom SQL
· You must avoid using custom SQL, rather use live views to implement custom SQL and connect Tableau to the view.
· Tableau creates subqueries for custom SQL which is difficult for database to handle.
· Try using views in the database or use multiple joins instead of using custom SQL.
5. Rendering your dashboard
· The “Automatic configuration for the size of your dashboard might contribute to performance issue too. Try to use fixed size for the dashboards.
· High usage of marks counts will increase the rendering time of dashboards.
· Do not use large file size images. This will affect the loading time.
· Avoid using more no of custom shapes in the dashboard.
6. Dashboard layout
· Limit the number of worksheets in a dashboard. Each worksheet will have to run its own queries, so the more worksheets in a single dashboard, the more queries Tableau will have to render.
· Limit the number of worksheets in a dashboard. Each worksheet will have to run its own queries, so the more worksheets in a single dashboard, the more queries Tableau will have to render.
7. Publishing
After publishing a dashboard to the Tableau server, the local computation May impact the server. so, it's best to limit the use of groups, hierarchies, blending, calculations and reference lines.
8. Clean up your workbooks
Multiple number of worksheets can lower Tableau performance in a big way, Delete or consolidate unused worksheets and data sources to improve the speed of your dashboard. A clean workbook is easier to work, and it provides the results instantly.
Hopefully, the tips I shared in my blog can help you optimize the speed of your dashboards. A Tableau server has a lot of capabilities that govern the data sources, database connections, workbooks and more, but these capabilities can only be utilized when the Tableau is optimized for that kind of performance. In addition to all the best practices that are discussed in this blog, there are of course many other tricks that you can perform to speed up your dashboard performance.