add details about data type mismatch #4031
Conversation
|
|
||
| In cases where a data type is determined incorrectly, the support team can help you update the data type. As an example, if a field can include float values as well as integers, but the first value we received was an integer, we will set the data type of the field to integer, resulting in a loss of precision. | ||
|
|
||
| To update the data type, the support team will update the internal schema that Segment uses to infer your warehouse schema. We will start syncing the data with the correct data type after the change is made. However, if you want to backfill all historical data correctly, it will be required to drop the impacted tables on your end so Segment can recreate them in the correct datatype, and then backfill those tables. |
There was a problem hiding this comment.
Hi @muhammadatifjav, I had a quick question about this part - after a customer drops the impacted tables, would Segment automatically recreate the tables and backfill them, or would a customer need to reach out to facilitate the backfill?
There was a problem hiding this comment.
After customer drops the impacted table, Segment will automatically recreate the table and start syncing new data automatically. However, if they want historical data to be corrected, then they will need to drop the table, and then confirm it to us and we will then run the backfill. I've made another commit to make it more clearer, lmk if that is better
|
Thank you for your contribution! Your pull request is merged, but may take a day or two to appear on the site. |
Proposed changes
Warehouse data type mismatch is a common issue that the customers face. We have macros/internal docs to explain the issue, and offer possible resolution, but it can be helpful if we can explain why this happens and quicker to resolve if the customer uses the template when reaching out to resolve this data type mismatch.
The changes proposed include:
Merge timing
Related issues (optional)