<@U04S8PLJWMT> why is my connection retrying 20 ti...
# replication-ask-ai
l
@kapa.ai why is my connection retrying 20 times even when I specified max 1 retry in the connection builder?
k
Your connection is likely retrying for two different reasons, controlled in two different places: 1. Connector-level HTTP retries (Connector Builder setting) - In the Connector Builder, the error handler’s “Max retry count” controls how many times a single HTTP request from the connector is retried (for responses marked as
RETRY
or
RATE_LIMITED
). Default is 5, but you can set it to 1. [[Advanced config](https://docs.airbyte.com/platform/connector-development/connector-builder-ui/error-handling#advanced-error-handling-configuration)] - This only affects API request retries inside the connector, not how many times the whole sync job is re-run. 2. Platform-level job retries (Airbyte job engine) - Separately, Airbyte’s job engine will retry the entire sync job on failure, with its own limits: - Up to 5 subsequent attempts where no data was synchronized - Up to 10 total attempts where no data was synchronized - Up to 20 total attempts where some data was synchronized [[Sync jobs](https://docs.airbyte.com/platform/understanding-airbyte/jobs#sync-jobs)] - These job-level retries are independent of the Connector Builder’s max retry count. So even if you set max retry = 1 in the Connector Builder, the job itself can still be retried many times by the platform, which is why you may see up to ~20 attempts overall. If you need fewer total retries, you must: - Configure max retry count in the connector (for HTTP retries), and - Adjust the job retry settings in your Airbyte deployment (OSS: via the “jobs” configuration) to reduce the allowed attempts per job. [[Sync jobs](https://docs.airbyte.com/platform/understanding-airbyte/jobs#sync-jobs)]