Hi,
I have an Iceberg Table with a column defined as of timestamp
type.
I’m experimenting with Athena v3 Datasources in QuickSight.
Creating a Dataset from the Athena v3 Datasource, using a custom SQL to retrieve all columns, it works if the dataset is leaved in Direct Mode, and I’m able to build an analysis from that data, using the timestamp column as well.
But, If I choose to change the Dataset to SPICE, then all rows are skipped due to a MALFORMED_DATE
error for the timestamp column.
In the Athena query editor I’m able to query the Iceberg table without issues and I can see that the column shows timestamps as 2023-01-23 12:20:00.000000 UTC
.
The same steps as above, using the same Iceberg table, and SPICE, but using Athena v2 works without any issue.
Any suggestion?
Thank you.
UPDATE: in my Athena v3 based dataset, if I edit the query to exclude the timestamp column and to include a new one as CAST(CAST(myTimestamp as varchar) as timestamp) as convTimestamp
it works, the data is correctly ingested to SPICE and I can build my analysis. Not the ideal way…