Hi! I'm trying to sync Hubspot as a data source and I noticed that for most of the streams, Airbyte only does full refresh instead of incremental. Is there a safeguard implemented (e.g. by setting some delays between each call to query each page) to prevent the API rate limit from being hit?

Yes, there is a backoff function and error handling in the Hubspot connector. You can check the code https://github.com/airbytehq/airbyte/blob/2e128cf982fe1757e098bd8dfe8052c2a75ce58a[…]te-integrations/connectors/source-hubspot/source_hubspot/api.py

Most cases when we don't support icnremental is because the API doesn't give a proper field to have this sync mode

Got it, thanks @[DEPRECATED] Marcos Marx! Can I double check though, if I read the code right, it seems like the built in mechanism is to back off once the rate limit is reached?

I guess where I'm coming from is if the Airbyte data sync is hitting the rate limit, it could interfere with the actual operations of the business (i.e. if our system is calling the Hubspot API to POST / PATCH records while the API rate limit has been hit, it would fail)

Got it, thanks @[DEPRECATED] Marcos Marx! Can I double check though, if I read the code right, it seems like the built in mechanism is to back off once the rate limit is reached?

yes

if possible create another token to read data will be helpful to not interfere in your other system