Hello! I'm having a problem regarding schema updates on Metabase. I have parquet files in AWS S3 that are fed into Metabase so they can be displayed as a table. However, whenever I add a new column in the middle of the schema the data is not displayed in their right positions. Suppose I have the following data on S3:
| a | b | c |
|---|---|---|
| car | 40 | true |
then, I add a new column as follows:
| a | d | b | c |
|---|---|---|---|
| car | 40 | true | |
| bus | 6.0 | 7 | false |
Instead of displaying the data correctly with a null value for the d column in the first row, the data is shown as follows:
| a | d | b | c |
|---|---|---|---|
| car | 40 | true | |
| bus | 6.0 | 7 | false |
I think MB tries to read the parquet files by Index instead of column names. Is it possible to force MB to fetch data by column? Otherwise, I will have to add new columns only at the end of the schema everytime and that's very annoying.
Obs.: The data is stored as parquet in S3.