Hi everyone!
I'm experiencing a strange issue.
We have data coming into Elastic as a CSV file, delimited by semicolons.
When I create a table in Canvas, some rows contain data from another column. Additionally, in one of the columns, those rows include data from multiple columns, with the delimiters still intact.
However, when I check Discover, that column only contains a single type of data.
Example (Fake Data)
If I check "No. of items" in Discover, it only shows integers.
In the CSV file, each column contains only one type of data, as it should. It's as if the split function that separates the CSV data into columns didn't fully work for some rows..
Anyone who knows what might be wrong?
/Jens
Anyone?
Do you know what the problem might be? Or do you know someone I can reach out to? It seems like a bug in Canvas, but I'm not sure.
I might add that the code is super simple. Just a simple SELECT...FROM...WHERE query. No .css or anything like that.
I tried running a query in Discover to see if I could reproduce the error, but it didn’t return any rows. This confirms that the data is correct in Discover, meaning the issue must be with how the data is displayed in Canvas.
Additionally, in Canvas, I tried displaying a simple number instead of a table. I ran:
SELECT COUNT(DISTINCT columnX) FROM xxx WHERE columnX = valueThatShouldntBeInColumnX
and it returned 64.
Something is definitely wrong.
Hi.
Problem solved. It turned out that the data was wrong in the Elastic database as well. Apparently it was a parsing problem when the data was loaded into Elastic. When I checked Discover, the column had only one type of data. But someone else at the company I work for could see there was multiple data in that column.
