Analyze Tab : Connection failed, try again later…

  • 15 September 2021
  • 6 replies

  • Known Participant
  • 17 replies

Hi Team,

After fresh GDCN installation we have init the LDM with GBQ when we visit Analyze Tab  UI throwing error tost “Connection failed, try again later…” 



Plese help us ..and kindly let me know logs required from our end


Thanks and Regards,


Best answer by jacek 17 September 2021, 08:46

View original

6 replies

Userlevel 2

Hi Ashok,

I can see that you did not drag and drop anything from the left panel into MEASURES or ROWS and you still get the red bar.

Please, try to drag and drop any element from the left panel into MEASURES, save the visualization (report) and prepare a dashboard.

Sorry for this annoying bug, it relates to the data preview triggered when you hover your mouse over elements in the left panel and an element is too large (meaning byte size of all values in the element).

The bug will be fixed in the next 1.4 release.



Userlevel 2

And also sorry for the late response - we all attended GoodData offsite yesterday, where many of us actively participated e.g. with presentations, so we did not have time to track the community.

Hi @jacek  thanks for the response !,


I am unable to create any of Dashboards insights!!



Userlevel 2

Well, my assumption is that the label “Audience name” is too large (cardinality x avg length).
Because the bug I mentioned it is not possible to workaround it like you are trying to do with the filter.

Let’s proof the assumption. Please, send me logs from your installation. If it is Community edition, send me all logs from the docker container, if it is K8S deployment, send me logs from result-cache pods.

If it is K8S deployment, did you customize correspoding values.yaml file in any way?

Also, please, select COUNT DISTINCT as well as AVG LENGTH of “Audience name”, resp. of the underlying column and share it here.

If my assumption is correct, there should be a way how to scale up relevant internal services even in your version 1.3, both in Community edition and in K8S.

Hi @jacek  we have running on GDCN K8’s deployment .

PFA for pod logs of  result-cache .


So this looks like a different problem than what @jacek anticipated:

"detail":"{'ERROR',{user_fault,\"no source table covering objects: objects=[attribute/audiences.audience_id/headquarters]\"}}"

This means that you are missing Data mapping between attribute “audiences.audience_id/headquarters” and the underlying physical tables. Ideally your mapping should look like this in LDM modeler:

I assume that in your case the Source Columns are blank. You can either filled them manually or better generate them from scratch using Scan feature of LDM modeler with the “Generate datasets” option turned on like this:

Let us know if that helped.