Stuck In Refresh

Question

So, I found that on the refresh page you can find the number of rows generated by a dataset.  This help to determine if the Dataset build was at least reasonable.  But what do you do when it gets caught in 30+ min refresh loop. It would help if the standard D2L datasets shows the number of rows so I could set an upper limit for the number data rows that could or might be generated.  Is there an option beyond duplicating it and redesign last dataset added?

I found that opening two windows in Analytics Builder I can see the dataset creation and monitor the output in the related analysis. The problem is that once it gets in the refresh loop, I get old or no data as it constantly searches for data.

At this point I have no better solution than to open an analysis for each D2L datasets and then go back and forth as I try to figure out when and what row counts on data collections are probable and which ones show failed designs.

John.G.31 · Answer

For my peace of mind, a data set is a D2L construct, and a data collection is user-built construct.

When creating a data collection in the Analytics Builder (AB), once the data collection is created and you select it for editing. You will see Summary - Refresh - Permissions - Usage pages listed under the data collection name on the next page. Refresh shows, for lack of a better term, Compile date and compile process status.  It will also show number of data rows generated.  Where DOMO would show on one page the 7 datasets and the number of data rows generated for each data set listed under each utilized data set.

I have not found the equivalent in AB.  However, for a constructed data collection in AB you can find how many rows are generated.  If I was expecting 20K-30K and if 1M showed up I knew the Join was not adequate.  In AB I have to go to the refresh page or go the analysis and run a pivot analysis or export the CSV or run a pivot table in Excel to generate total counts on a page to summarize the total rows generated.

Simple concept but if no data set has over 11k rows no relational data collection would have over 11K rows. Unless the related join is allowing a multiple association that should not exist.

If anyone finds a way to not go back and forth to analysis to determine the effectiveness of a join, please let me know. Two monitors and two instances (web pages) work, but I still have to remember to Save and Publish and refresh the Analysis windows.

Roqeeb.O.922 · Answer

Hi @John.G.31,

We love that you’re reaching out to the Brightspace Community with this question!

At the moment, all Brightspace Data Sets adhere to the default system limit to return 150 million rows of the most recent data.

Are you able to provide more information about the refresh page so that we can better assist you regarding this?

Thank you,

Jennifer.W.973 · Answer

@John.G.31 Hi John, what is the refresh page? I'm also trying to figure out the best way to get basic data profiling (like in Power BI) that will provide important numbers like count of rows before having to create an analysis and build a table that matches what I already built in the dataset… but it does not seem like Quicksight includes any of that.