QuickSight vs database: same query, different result by 1

Q&A
,
1

Hi QuickSight Community,

I’m seeing a mismatch between results from my database and Amazon QuickSight when running the same query. For one specific column, QuickSight’s value is consistently off by 1 compared to the database result.

I’ve already confirmed the column data type matches in the database and in QuickSight, but the discrepancy remains.

2

Hi @abhjaiswal ,

Without having further information to identify the root of the issue, here are some potential theories:

  1. In addition to confirming the data type matches, the issue may be due to rounding differences between your data source and your calculation in Quick Sight
  2. Are you ingesting your data into SPICE, or is it a direct query? This should not cause the error, however you can test switching between these 2 modes to further narrow down the root cause of the issue
  3. Check values such as nulls as they may be interpreted differently between your data source and the calculation you are doing in Quick Sight.

Please verify the above, and if you determine this is a bug please raise a support ticket.

Thanks

3

Hi @oyin,

Thanks for your response, I had gone through all 3 points you have mentioned so since data type is of Integer type there is no scope for round off logic error however, I still try to change the format to include decimal values upto 3 decimal place but the issue still persist.
Changing dataset ingestion from SPICE to direct query doesn’t seems to be helping either, and there were no null values in that column.
So I think I need to raise a support ticket for this, Could you please guide where do we need to raise the ticket is this to be done on my AWS account support or there is some other platform specifically for AWS Quick

4

If your data source is a database, you can use query monitoring tools to inspect the SQL that QuickSight generates. Keep in mind that, in addition to the base query defined in your dataset, any transformations and filters applied at the analysis or visual level are translated into additional query logic and executed against the database. By tracking this final query, you can compare it with your original query to identify any differences

5

Hi @abhjaiswal,

If you are stilling encountering this same issue, here is the documentation that showcases how to create an AWS Support ticket. Feel free as well to update the thread if your support case leads to a solution in case other people are running into similar behavior!

Thank you!