This is S3 file dataset.
Initially, I created a dataset with just uploading manifest file directly from my local pc.
Now, I have updated s3 file like 2 additional columns. Refreshing dataset multiple times doesn’t do any change. So, I tried to upload manifest file to s3 bucket, and update dataset with reference to that s3 filepath.
Then, I see 2 additional columns are added into preview dataset but unfortunately it created a new dataset.
Is there a way to keep the original datasetID and just update data only?
I’m trying to do update-data-set in command line but wonder if this can be done in CONSOLE side .
Hi @tbdori - The data set id is created by AWS QuickSight from back ground, I do not think you can change it from UI layer. However you can replace the data set in your analysis by the new data set so that you can see this additional 2 columns in analysis.
Regards - Sanjeeb
yeah, the point is my original dataset has all those Physical/Logical table mapping info and I don’t want to do that on the new dataset. My goal is to simply replace underlying data itself, not changing datasetid and it will cause another change in my migration to STG.
I was able to successfully update data-source itself via command line.
I simply don’t know why new dataset is created when I update it.
when you use other than ‘existing manifest’, is it correct to create a new dataset?
Hi @tbdori - It depends upon the requirement. From UI, you can not maintain the same data set id by another data set via replace. It is only possible from Cli as data set ids are created by background process and from UI, you do not know what it is and it is unique per data set.
Regards - Sanjeeb