![]() ![]() Run the query from the same machine as Tableau Desktop through a SQL querying tool.Have your DBA run the exact same query from Tableau directly on the database (or as close to directly as the can).Good examples of these processes are in the Best Practices For Designing Efficient Workbooks whitepaper or Interwork’s Performance Checklist If the limited amount returns quickly, then that indicates the dashboards and sheets need to be optimized to return less data until a user has filtered down. Most SQL querying tools limit the actual number of results they return back. How many results come back (you want a full row count, usually displayed somewhere). ![]() To investigate #2: Run the Tableau query in your direct querying tool. If they have reached a state where they feel like things are as optimized as possible, but Tableau is still doing some things that will necessarily take a long time, consider altering the worksheet where the inefficient query is running to see if Tableau will write a simpler query. They should be able to make changes or recommendations that will optimize the query at the database level. The query is neither complex nor results in much data, but still takes a long time to return.įor #1: Send the query to your DBA and have them run an EXPLAIN to understand why the query might be taking a long time to run.The query isn’t difficult to compute, but returns an extremely large amount of data.The query itself is complex and the database is not optimized to answer it in the form Tableau is sending.“Executing Query”Ī query typically takes long time for one of these reasons (it can be a combination though): The Performance Recording will be a new workbook which will show which steps are taking the most time. Once it has reloaded, click around or change filters (whatever actions have been slow on the Server). If you already have the workbook open, you can go up to the Data menu at the top, find the data source name and then choose Refresh. The best instructions are from the Tableau Online Help. Some databases have resource limits in place for certain users, so at this point we want to be testing with the exact same credentials. These should be the same credentials used on Tableau Server. You will need the database credentials to you can log into the database once it the source is local.If there is a Published Data Source (the Tableau icon is next to the data source rather than a database cylinder icon), follow this KB article to make it a local data source.Open the TWB or TWBX file in Tableau Desktop.Download the workbook file from Tableau Server to the local machine.To get all the processing moved to Desktop, follow these steps: If you are using a workbook that has Published Data Sources, your queries will still be going through Tableau Server. This means that the first testing stage is always in Tableau Desktop. You might alternatively think of this in terms of “The maximum load speed of a given viz on Server is less than or equal to the load speed of the same workbook in Tableau Desktop”. The main principle we use to isolate the issue is “ If it is slow in Tableau Desktop, it will be slow in Tableau Server“. If it is slow in Tableau Desktop, it will be slow in Tableau Server Test the workbook during a time of high usage on the Tableau Server (either natural or do load testing using TabJolt).Did it perform relatively the same as the test in Step 2 (within 1-3 seconds)? If yes: Publish the workbook to Tableau Server, and find a time when there is low-to-no usage on the Tableau Server.Does it perform the same as it did on the previous machine? If yes: Test the workbook in Tableau Desktop on the Tableau Server machine.In this post, I’ll be describing a set of steps to follow to isolate the causes of performance issues on Tableau Server.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |