@kapa.ai Are there any problems with having 2 connections write data to the same schema?

Thinking completed Yes, there are real problems with having 2 connections write to the same schema (namespace) and table names. Here's what can go wrong: Concurrency Issues with Temporary Tables When Airbyte syncs data, it creates temporary (

tmp

) tables as part of the process. If two connections share the same destination schema and the same stream/table names, one connection can delete or overwrite the

tmp

tables created by the other, causing data corruption or sync failures. [Forum Post] Stream Uniqueness Requirement Airbyte explicitly requires that stream identifiers be unique across connections that share the same destination. Configuring more than one connection to sync to the same destination stream can cause: • Job failures • Data loss (as multiple connections clash while writing to the same table) [Configuring Schemas] Recommended Solutions To avoid these issues, you have a few options: 1. Use unique namespaces per connection — Write each connection to a separate schema/dataset, then combine the data downstream using a view or query. [Namespaces Docs] 2. Use a Stream Prefix — Add a unique prefix to each connection's streams so table names don't collide within the same namespace. [Configuring Schemas] 3. Combine data downstream — Let Airbyte sync to separate locations, then merge the data in your destination (e.g., via a BigQuery view or dbt model). > Note: As a last resort workaround (not recommended for production), you could limit concurrent syncs by setting

MAX_SYNC_WORKERS

to

1

, but this reduces throughput and doesn't fully solve the underlying problem. [Forum Post] 👍 0 button 👎 0 button ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.