What is the difference between capturing changes and writing a copy?
A capture reads the source into a collection. A separate step writes the collection to a destination. A database connector can use change data capture, which reads changes as they happen from a supported database rather than taking a whole new copy each time.
Check the specific source requirements before choosing that method. A connector that reads a change log can need setup on the original database.
Will removing a source row remove the copied row?
That depends on the destination connector and its settings. Estuary’s PostgreSQL destination can mark the row as deleted or remove it if hard deletes are enabled. Check the chosen source sends the deletion and the destination behaves the way your report expects.
Why can the bill count the same data twice?
Estuary counts data when it is captured and again when it is delivered to a destination. It also counts active source and destination connector instances. The Developer plan currently lists up to 10 GB per month and two connector instances.
If one source feeds two destinations, include both deliveries in the estimate.
What should you document if another team takes over?
Keep the copy in a destination your team controls. Document the source capture, collection, destination mapping and the settings that affect updates and deletions. Estuary documents editing flow specifications outside the dashboard. Keep credentials separately.
How would you check that a change reaches a report?
Draw a fictional source, collection and destination. Use no credentials, database or network connection.
- Pick a test eventWrite a made-up order ID and then change its amount.
- Trace each stopMark where the update is read, held and written. Draw how you would spot a destination still showing the old amount.
- Check a deletionDraw the expected destination after a marked deletion and after a removal. Name the rule the report should use.