Tableau view fails to get loaded

Why Tableau views do not load or scheduled reports to fail
Written by Carsten
Updated 2 years ago

Whenever Kiwi is requesting a resource from Tableau a "job" is created to keep track of the request and potentially identify issues that may happen.
We are constantly monitoring these jobs and try to make sure that any error that comes up is being fixed on our site or with the help of the user. 

As an admin user, you can check your own jobs and see if everything works fine or if there are errors by going to

Following we have listed some typical errors we see:

Internal Error 500 

Internal Error 500 is by far the most frequent error we see looking into our job data.
In 90 % of all cases, this is due to the fact that the data source behind the Tableau View has no credentials embedded. So when Kiwi requests it the login to the data source fails and we cannot refresh or load the chart. However, there is an easy fix: republish the workbook to your Tableau Server with embedded credentials. For detailed guidance on setting embedded credentials, you can read the Tableau help page.

Tableau timeouts

We all know it. Sometimes, refreshing data in Tableau can take a while. There is no difference for Kiwi. The default timeout for our jobs is 2 min. If your data source takes longer to refresh the job fails. There are three solutions:

  1. We increase the job timeout to max. 4 min. Just tell us what views are affected or if we should increase timeouts for all views.

  2. Automated caching of specific views. With our latest release, we can automate a Tableau refresh every x min and keep it cached. So, whenever a view is requested there is < x min old version of that view that can be used without a refresh. This is especially helpful for the Slack BI Self-Service because users do not have to wait. Besides that, we have a default caching time of 2 hours. So if you request a view twice within 2 hours the second time it comes from the cache.

  3. Just wait :) Any failed job is automatically retried after 5 min. If there was a timeout in the first place the chances are high that after 5 min the job has been refreshed and can be delivered without any errors.

Your Issue is not listed?

If you face any other issues or you need further technical support / info from us do not hesitate to contact us. In order to provide fast help please add the jobId of the failed job. 

Did this answer your question?