An Azure service for ingesting, preparing, and transforming data at scale.
The first run might be happening during a time when the data in Snowflake is still being updated or not fully committed, which causes the Lookup activity to fail. The second run works because the data becomes available.
I recommend that you implement a retry mechanism in your pipeline using an "Until" activity or a retry policy for the Lookup activity, allowing it to reattempt fetching the data until it succeeds.
You could also try downgrading to the previous version temporarily to see if it resolves the issue.