Anna Bogo
11/26/2025, 12:51 PMkapa.ai
11/26/2025, 12:51 PMAnna Bogo
11/26/2025, 12:55 PMkapa.ai
11/26/2025, 12:55 PMkapa.ai
11/26/2025, 12:55 PM"Some streams either received an INCOMPLETE stream status, or did not receive a stream status at all: service_retool_2.dhl_api_documents\nio.airbyte.commons.exceptions.TransientErrorException: Some streams were unsuccessful due to a source error. See logs for details." - Source error: "Source process exited with non-zero exit code 3" "Something went wrong within the source connector" - No records committed: The sync emitted 10 records but committed 0 to Snowflake. ### What does this mean? - The error is classified as a transient error on the destination side, but the root cause appears to be a source connector failure (exit code 3). - The destination (Snowflake) is reporting that it did not receive a complete or valid stream from the source, which matches the source process exiting unexpectedly. - This pattern is similar to other issues where the source connector fails, causing the destination to report incomplete or missing data, and ultimately no records are written [source-jira sync hangs forever](https://github.com/airbytehq/airbyte/issues/58082), [source-zendesk-support ticket_comments not syncing](https://github.com/airbytehq/airbyte/issues/52133). ### What should you do next? 1. Check the Source Connector Logs: The most actionable step is to look at the detailed logs for the source connector (Retool in your case). The exit code 3 is a generic failure, and the root cause (such as a schema issue, API error, or data problem) should be visible in the source logs. 2. Review Data and Schema: Since the same tables work in Redshift but not in Snowflake, check for: - Data types or values in the failing stream that might not be compatible with Snowflake. - Schema mismatches or reserved keywords in Snowflake. 3. Connector Version and Updates: Ensure both your source and destination connectors are up to date, as some transient errors and stream handling bugs have been fixed in recent versions [Data not showing in snowflake from redshift](https://discuss.airbyte.io/t/2119). 4. Resource Constraints: While your sync is small (10 records), sometimes resource issues or temporary network/API failures can cause transient errors. If this is a recurring pattern, try restarting the sync or the Airbyte instance. 5. Retry and Split Streams: If the issue is isolated to a single stream, try syncing only that stream in a new connection to isolate the problem further (which you mentioned you already tried). ### Summary The failure is due to the source connector exiting with an error, which then causes the destination (Snowflake) to report an incomplete stream and commit no records. The next step is to review the detailed source connector logs for the specific error message that caused the exit code 3. If the logs do not provide enough information, consider updating your connectors and Airbyte version, and double-check the schema and data for compatibility issues. If you can provide the relevant section of the source connector logs, I can help analyze the root cause further.Anna Bogo
11/26/2025, 12:57 PMkapa.ai
11/26/2025, 12:57 PMkapa.ai
11/26/2025, 12:58 PM/recent endpoint limited results to 10,000 records, and the connector logic could cause it to always fetch the same subset if state handling or endpoint selection is incorrect. A similar issue could be happening with your source, especially if the API or connector was updated recently or if the endpoint being used is not paginating correctly. Review the connector’s logic and the API documentation to ensure the correct endpoint and pagination are being used [Hubspot engagements stream -- /recent vs /paged endpoint](https://discuss.airbyte.io/t/3718). 2. State Handling or Cursor Issue If the connector is not updating its state or cursor correctly, it may repeatedly fetch the same records. This can happen if the cursor field is not being updated, or if the API does not support proper incremental sync and the connector is not handling pagination or state checkpointing as expected. This is a common cause for only seeing the same records on each sync [Incremental sync requirements](https://docs.airbyte.com/platform/1.8/connector-development/connector-builder-ui/incremental-sync). 3. Source API Data Change If the source data itself has changed (e.g., only 10 records now match the query or filter), Airbyte will only sync those. Double-check the source system to ensure there are more records available and that no filters or permissions have changed. 4. Connector or API Error If the connector or API is returning errors or incomplete data, Airbyte may only process a partial result set. Review the Airbyte logs for any warnings or errors from the source connector. Recommended Actions: - Check if the source API or connector was updated or changed in the last few days. - Review the connector configuration, especially pagination and cursor settings. - Inspect the Airbyte logs for the source connector for any errors or warnings. - Confirm that the source API is returning more than 10 records when queried directly (e.g., via Postman or curl). - If the connector uses state, try resetting the connection state in Airbyte and running a full refresh. If you provide the connector type and any recent changes, I can help with more targeted troubleshooting steps. If the issue persists and none of the above applies, the knowledge sources do not provide a more specific solution for this exact scenario.Anna Bogo
11/26/2025, 1:01 PMkapa.ai
11/26/2025, 1:01 PMkapa.ai
11/26/2025, 1:02 PMSUPER object, and fields exceeding this limit are dropped before insertion, which can result in missing or incomplete records in the destination. Similar issues have been reported with other connectors when handling large fields, leading to sync failures or only partial data being extracted and committed [Issue with long JSON field and mysql source](https://discuss.airbyte.io/t/1566). If you encounter this again, consider truncating or excluding excessively long fields in your source schema, or adjusting the connector configuration to handle such cases. This approach can help ensure successful extraction and syncing of your data.