Lucas Segers
12/15/2025, 4:25 PMkapa.ai
12/15/2025, 4:25 PMHTTP_IDLE_TIMEOUT and READ_TIMEOUT) to handle long-running discovery operations, not per-call logic in the connector itself.[\[Scaling Airbyte\]](https://docs.airbyte.com/platform/enterprise-setup/scaling-airbyte#schema-discovery-timeouts) - Sync heartbeats and destination timeouts are controlled globally via flags.yaml (e.g., heartbeat-max-seconds-between-messages, destination-timeout-max-seconds), again at platform level, not at connector-definition level.[\[Heartbeats\]](https://docs.airbyte.com/platform/understanding-airbyte/heartbeats#technical-details) Individual connectors sometimes expose connection/read timeout parameters in their JDBC URL / advanced settings (e.g., Postgres connectTimeout, Oracle oracle.net.CONNECT_TIMEOUT), but those are connector-specific configuration fields, not a general low-code “call timeout” feature.[\[Postgres destination\]](https://docs.airbyte.com/integrations/destinations/postgres#naming-conventions)[\[Oracle source\]](https://docs.airbyte.com/integrations/sources/oracle#getting-started-airbyte-open-source) So based on the available documentation, there is no documented way to define a generic API-call timeout directly in the low-code connection builder. You’d need to rely on platform-level timeouts/heartbeats or connector-specific options where they exist.