Connector Improvement: Allowing facebook connector to handle large initial syncs
The request: Break up the facebook api requests to avoid pagination limit.
For the facebook connector it seems to have a pagination limit of about 40k records per ad account. I'm assuming fivetran currently runs one request per ad account. If historically you have more than 40k records, some will be dropped. (usually at random too)
In order to not have to join with historical data in our data warehouse. We would ideally like the data to be handle completely by the fivetran facebook connector during the historical sync, so it is all in one place.
Note that this was only observed with the historical sync, but not sure if fivetran will be able to tell if past records have been updated beyond this limit either. So this could potentially be affecting other customers who have more than 40k records per ad account. Meaning they are not getting synced data if they have more than 40k records in their ad account.
Facebook api mentions they have limits on pagination, although they cannot publicly disclose the limit: https://developers.facebook.com/docs/graph-api/results
Please sign in to leave a comment.
Comments
0 comments